P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
...
43 downloads
1074 Views
3MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
International Review of Industrial and Organizational Psychology 2011 Volume 26
International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
International Review of Industrial and Organizational Psychology 2011 Volume 26 Edited by
Gerard P. Hodgkinson The University of Leeds, UK and J. Kevin Ford Michigan State University, USA
A John Wiley & Sons, Ltd., Publication
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
This edition first published 2011 C 2011 John Wiley & Sons Ltd. Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientific, Technical and Medical business with Blackwell Publishing. Registered Office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom Editorial Offices The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 9600 Garsington Road, Oxford, OX4 2DQ, UK 350 Main Street, Malden, MA 02148-5020, USA For details of our global editorial offices, for customer services, and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. The right of Gerard P. Hodgkinson and J. Kevin Ford to be identified as the authors of the editorial material in this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data International review of industrial and organizational psychology. —1986—Chichester; New York; Wiley, c1986– v.: ill.; 24cm. Annual. ISSN 0886-1528 1 /4 International review of industrial and organizational psychology 1. Psychology, Industrial—Periodicals. 2. Personnel management—Periodicals. [DNLM: 1. Organization and Administration—periodicals. 2. Psychology, Industrial—periodicals. W1IN832UJ] HF5548.7.157 158.7005—dc 19 86-643874 AACR 2 MARC-S Library of Congress [8709] ISBN: 9780470971741 A catalogue record for this book is available from the British Library. This book is published in the following electronic formats: ePDFs 9781119992608; Wiley Online Library 9781119992592 Typeset in 10/12pt Plantin by Aptara Inc., New Delhi, India 1 2011
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
CONTENTS About the Editors
vii
Contributors
ix
Editorial Foreword
xi
1. Stress and Well-Being are Still Issues and Something Still Needs to be Done: Or Why Agency and Interpretation are Important for Policy and Practice Kevin Daniels 2. Brain, Emotion, and Contingency in the Explanation of Consumer Behaviour Gordon R. Foxall 3. Longitudinal Assessment of Changes in Job Performance and Work Attitudes: Conceptual and Methodological Issues David Chan 4. Estimating the Relative Importance of Variables in Multiple Regression Models Dina Krasikova, James M. LeBreton, and Scott Tonidandel 5. Employee Trust in Organizational Contexts Rosalind Searle, Antoinette Weibel, and Deanne N. Den Hartog
1
47
93
119
143
6. The Physical Environment of the Office: Contemporary and Emerging Issues Matthew C. Davis, Desmond J. Leach, and Chris W. Clegg
193
7. Deception and Applicant Faking: Putting the Pieces Together Brian H. Kim
239
v
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
vi
T1: IML 12:51
Printer Name: Yet to Come
CONTENTS
8. Actions Speak Too: Uncovering Possible Implicit and Explicit Discrimination in the Employment Interview Process Therese Macan and Stephanie Merritt
293
Index
339
Contents of Previous Volumes
349
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
ABOUT THE EDITORS Gerard P. Hodgkinson Leeds University Business School, The University of Leeds, Leeds, LS2 9JT, UK J. Kevin Ford
Department of Psychology, 315 Psychology Research Building, Michigan State University, E. Lansing, MI 48824, USA
Gerard P. Hodgkinson is Professor of Organizational Behaviour and Strategic Management and Director of the Centre for Organizational Strategy, Learning, and Change (COSLAC) at the University of Leeds, UK. He earned his BA, MSc, and PhD degrees at Wolverhampton Polytechnic and the Universities of Hull and Sheffield, respectively. He has (co-)authored three books and over 60 scholarly journal articles and chapters on topics of relevance to the field of industrial and organizational psychology. A Fellow of both the British Psychological Society and the British Academy of Management, and an Academician of the Academy of Social Sciences, his work centres on the analysis of cognitive processes in organizations and the psychology of strategic management. In recent years, his work on these topics has been taken forward through the award of a Fellowship of the Advanced Institute of Management (AIM) Research, the UK’s research initiative on management funded by the Economic and Social Research Council (ESRC) and Engineering and Physical Sciences Research Council (EPSRC). From 1999 to 2006 he was the editor-in-chief of the British Journal of Management and currently serves on the Editorial Boards of the Academy of Management Review, Journal of Management, Journal of Organizational Behavior, and Organization Science. A chartered occupational psychologist, registered with the UK Health Professions Council as a practitioner psychologist, he has conducted numerous consultancy assignments for leading private and public sector organizations. Further information about Gerard and his work can be found at the following addresses: (1) http://www.leeds.ac.uk/lubs/coslac/ (2) http://www.aimresearch.org. J. Kevin Ford is a Professor of Psychology at Michigan State University. His major research interests involve improving training effectiveness through efforts to advance our understanding of training needs assessment, design, evaluation, and transfer. Dr Ford also concentrates on understanding change dynamics in organizational development efforts and building continuous learning and improvement orientations within organizations. He has published over 50 articles and chapters and four books relevant to industrial and organizational psychology. Currently, he serves on the editorial boards of the Journal of Applied
vii
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
viii
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
A BOUT THE E DITORS
Psychology and Human Performance. He is an active consultant with private industry and the public sector on training, leadership, and organizational change issues. Kevin is a Fellow of the American Psychological Association and the Society of Industrial and Organizational Psychology. He received his BS in psychology from the University of Maryland and his MA and PhD in psychology from The Ohio State University. Further information about Kevin and his research and consulting activities can be found at http://www.io.psy.msu.edu/jkf.
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
CONTRIBUTORS David Chan
School of Social Sciences, Singapore Management University, 90 Stamford Road, 178903, Singapore
Chris W. Clegg
Leeds University Business School, University of Leeds, Leeds, LS2 9JT, UK
Kevin Daniels
School of Business and Economics, Loughborough University, Loughborough, LE11 3TU, UK
Matthew C. Davis
Leeds University Business School, University of Leeds, Leeds, LS2 9JT, UK
Deanne Den Hartog Amsterdam Business School, University of Amsterdam, Roeterseiland–Building M, Plantage Muidergracht 12, 1018 TV, Amsterdam, The Netherlands Gordon R. Foxall
Cardiff Business School, Cardiff University, Aberconway Building, Colum Drive, Cardiff, CF10 3EU, UK
Therese Macan
Department of Psychology, University of Missouri–St. Louis, One University Boulevard, St. Louis, MO 63121-4499, USA
Brian H. Kim
Psychology Department, Occidental College, 1600 Campus Road, F-11, Los Angeles, CA 90041, USA
Dina Krasikova
Department of Psychological Sciences, Purdue University, 703-Third Street, West Lafayette, IN 47907, USA
Desmond J. Leach
Leeds University Business School, University of Leeds, Leeds, LS2 9JT, UK
James M. LeBreton Department of Psychological Sciences, Purdue University, 703-Third Street, West Lafayette, IN 47907, USA Stephanie Merritt
Department of Psychology, University of Missouri–St. Louis, One University Boulevard, St. Louis, MO 63121-4499, USA
ix
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
x
T1: IML 12:51
Printer Name: Yet to Come
C ONTRIBUTORS
Rosalind Searle
Department of Psychology, The Open University, Walton Hall, Milton Keynes, MK7 6AA, UK
Scott Tonidandel
Department of Psychology, Davidson College, Box 7061, Davidson, NC 28035, USA
Antoinette Weibel
University of Konstanz, 78457 Konstanz, Germany
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
EDITORIAL FOREWORD This is the twenty-sixth volume of the International Review of Industrial and Organizational Psychology. As with previous volumes we have purposefully selected for coverage a judicious combination of well-established and emerging topics. A good example of the latter is the chapter by Matthew Davis, Desmond Leach, and Chris Clegg (‘The Physical Environment of the Office: Contemporary and Emerging Issues’), which surveys theory and research pertaining to the psychological impact of contemporary office designs on the performance and well-being of employees. Two chapters (‘Longitudinal Assessment of Changes in Job Performance and Work Attitudes: Conceptual and Methodological Issues’ by David Chan and ‘Estimating the Relative Importance of Variables in Multiple Regression Models’ by Dina Krasikova, James LeBreton, and Scott Tonidandel) cover important methodological advances at the forefront of the field. Several chapters in the present volume offer fresh theoretical perspectives on topics covered previously in the series. Gordon Foxall’s chapter (‘Brain, Emotion, and Contingency in the Explanation of Consumer Behaviour’), for example, surveys leading edge developments in the philosophy of mind and the affective sciences, but does so through the lens of radical behaviourism to advance a theoretical framework for the analysis of consumer behaviour that extends well beyond the framework he outlined in his previous chapter in this series (published in the 1997 volume). The chapter by Kevin Daniels (‘Stress and Well-being are Still Issues and Something Still Needs to be Done: Or Why Agency and Interpretation are Important for Policy and Practice’) revisits the now extensive literatures on stress and well-being in the work place to develop fresh insights that challenge the conventional orthodoxy underpinning current policy and practice in this core area of professional activity. Brian Kim (‘Deception and Applicant Faking: Putting the Pieces Together’) develops new theoretical insights in relation to another topic of immense significance to practitioners: the faking process through which applicants all too often attempt to secure employment by distorting the truth about their qualifications and other attributes. Finally, the chapters entitled ‘Actions Speak Too: Uncovering Possible Implicit and Explicit Discrimination in the Employment Interview Process’ (Therese Macan and Stephanie Merritt) and ‘Employee Trust in Organizational Contexts’ (Rosalind Searle, Antoinette Weibel, and Deanne Den Hartog), again contain a wealth of new insights for researchers, policy makers and practitioners alike. In sum, once again we have commissioned a mixture of analytical reviews and reflective essays, each of which provide authoritative, state-of-the-art overviews
xi
P1: IML/IKJ fm
P2: IML/UKS
JWST046-Hodgkinson
xii
QC: IML/UKS January 21, 2011
T1: IML 12:51
Printer Name: Yet to Come
E DITORIAL F OREWORD
and commentary on major developments at the forefront of the fields of organizational behavior and industrial and organizational psychology. Each chapter offers a comprehensive and critical survey of the chosen topic, and each is supported by a valuable bibliography. For advanced students, academics, and researchers, as well as professional psychologists and managers, this series remains the most authoritative and current guide to new developments and established knowledge pertaining to behavior in the workplace. GPH JKF September 2010
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
Chapter 1 STRESS AND WELL-BEING ARE STILL ISSUES AND SOMETHING STILL NEEDS TO BE DONE: OR WHY AGENCY AND INTERPRETATION ARE IMPORTANT FOR POLICY AND PRACTICE Kevin Daniels School of Business and Economics, Loughborough University, Leicestershire, LE11 3TU, UK Stress and well-being are well-known, well-researched, and well-theorized areas of industrial and organizational (I/O) psychology. There are numerous reviews, meta-analyses, and influential publications on methods going back decades (e.g., Cass, Faragher, & Cooper, 2002; Cooper, Dewe, & O’Driscoll, 2001; Frese & Zapf, 1988; Jackson & Schuler, 1985; Kasl, 1983; Rick, Thomson, Briner, et al., 2002; Spector, Zapf, Chen, et al., 2000; Warr, 1987; Zapf, Dormann, & Frese, 1996). The area is characterized by high levels of methodological sophistication, including intervention studies (e.g., Jackson, 1983; Wall, Kemp, Jackson, et al., 1986), diary and experience sampling methods (e.g., Totterdell, Wood, & Wall, 2006; Xanthopoulou, Bakker, Demerouti, et al., 2009), multi-method designs (e.g., Frese, 1985), large-scale longitudinal studies (e.g., Bosma, Marmot, Hemingway, et al., 1997; Vahtera, Kivim¨aki, Pentti, et al., 2000), and qualitative investigations (e.g., Dick, 2000; Hepburn & Brown, 2001). Moreover, all this scientific effort has found its way into policy and practice. For example, the World Health Organization, the International Labour Organization (ILO), and the European Union (EU) have all issued guidance emphasizing the need to assess causes of stress and take preventive action to eliminate these causes at source (ETUC, 2004; ILO, 2001; Leka & Cox, 2008; Leka, Griffiths, & Cox, 2003). Many developed countries have sophisticated International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
2
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
national surveillance schemes to monitor work-related stress and well-being in the working population (Dollard, Skinner, Tuckey, et al., 2007). One particularly good example of how the science of stress has influenced policy and practice is the adoption by the UK’s Health and Safety Executive, the UK Government’s agency with responsibility for health and safety at work, of ‘The Management Standards for Work-Related Stress’ (see Cousins, MacKay, Clarke, et al., 2004; MacKay, Cousins, Kelly, et al., 2004). These Management Standards are underpinned by the applied psychological and epidemiological sciences related to stress and well-being. The science has found its way into guidance and policy. Surely then those of us concerned with research in this area ought to find something else to do? In this chapter, I argue much still needs to be done and that recent, and not so recent but largely ignored, perspectives on stress and well-being illuminate ways in which research, policy, and practice can develop further. The rationale for making the statement that much still needs to be done can be broken down into four main areas. First, this is still a lively area of research. With at least three specialist journals that take a psychological perspective on work stress and well-being (International Journal of Stress Management, Journal of Occupational Health Psychology, Work and Stress), and with relevant work continuing to appear frequently in some of the major I/O psychology, organizational behaviour, and management journals (e.g., Human Relations, Journal of Applied Psychology, Journal of Management, Journal of Occupational and Organizational Psychology, Journal of Organizational Behavior), it is clear there is still much research being published capable of meeting stringent requirements of peer review in international journals. Secondly, the area is practically important. Clearly, worker health is important in its own right, notwithstanding healthcare costs (Manning, Jackson, & Fusilier, 1996) and potential relationships with work performance (Jex, 1998; Taris & Schreurs, 2009). A survey of 29 766 workers in 31 European countries found 22.3% claiming to be affected by stress at work (Parent-Thirion, Fern´andez Mac´ıas, Hurley, et al., 2007). In the UK, on the basis of various regular surveys (UK Labour Force Survey, The Health and Occupation Reporting network), the Health and Safety Executive claims in 2008–2009 that: 415 000 individuals in Britain suffered illnesses because of work stress; 16.7% of working Britons found their work stressful; there is an annual incidence rate of work-related stress, depression, or anxiety of 760 cases per 100 000 workers; and that 11.4 million working days were lost to stress. The interesting point about the UK picture is that the Health and Safety Executive claim that these figures have remained relatively stable over most of the past decade, in spite of increased awareness and the introduction of the Management Standards for Work-Related Stress in 2005–2006.1 1
http://www.hse.gov.uk/statistics/causdis/stress/index.htm (accessed 20 November 2009).
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
3
Thirdly, there remain questions over the ability of the dominant approach to research and policy, based on linking psychosocial job characteristics to health and well-being, to account for the experience of well-being at work (Briner, Harris, & Daniels, 2004; Hart & Cooper, 2001). If policy and practice are based on research that provides incomplete or inaccurate theoretical explanations, then it is worth asking whether policy and practice are sub-optimal and how they can be improved. Fourthly, recent theoretical developments in the allied field of job design have indicated that I/O psychologists might be changing the way they think about work. Mirrored by changes to more knowledge intensive, interdependent, and information technology mediated work, the cognitive aspects of job design are theoretically more prominent (Grant, Fried, Parker, et al., 2010; Hodgkinson & Healey, 2008; Morgeson & Humphrey, 2006; Parker & Ohly, 2008). Moreover, there is growing research interest in the ways workers shape the content of their own jobs, also known as job crafting (Berg, Wrzesniewski, & Dutton, 2010; Clegg & Spencer, 2007; Wrzesniewski & Dutton, 2001). This interest in the cognitive aspects of job design and job crafting points to the importance of workers’ interpretation of their work and agency in shaping their work for understanding work-related health and well-being. The purpose of this chapter is to examine in greater depth whether we can develop effective policy and practice based on the historically dominant approaches to job characteristics, well-being, and stress. I identify several weaknesses with this approach. I then describe alternative approaches to understanding job characteristics and their impact on well-being that place much greater emphasis on individual agency and interpretation. This also considers potentially adaptive forms of coping. I then examine the practical and methodological implications of these alternative approaches. I will start by outlining what stress and well-being are.
STRESS AND WELL-BEING AS AN AREA OF STUDY Stress and psychological or subjective well-being are best conceived as areas or an area of study rather than as one or a set of dependent variables (Diener, Suh, Lucas, et al., 1999; Lazarus & Folkman, 1984). In I/O psychology, research generally relates to the impact of the work environment, in conjunction or in isolation of other processes, on psychological and psychosomatic health (see e.g., Cooper, Dewe, & O’Driscoll, 2001; Warr, 2007). Although research into stress has dominated I/O psychology, research on the positive influence of work on health is also important (Bakker, Schaufeli, Leiter, et al., 2008; Quick, Cooper, Gibbs, et al., 2010; Macik-Frey, Quick, & Nelson, 2007). Therefore, work in the areas of stress and well-being can be seen as complementary or even synonymous at a broad level.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
4
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
One reason for treating stress and subjective well-being as areas of study rather than definable physiological or psychological states concerns specifying the exact nature of the concepts (Diener, Suh, Lucas, et al., 1999; Mason, 1975). For example, stress symptoms have been conceptualized as having physiological, behavioural, cognitive, and affective aspects (Cox, 1978). Subjective well-being is more focused on psychological states, but can include affective states, judgments of life or domain specific (e.g., job) satisfaction, cognitive states such as aspiration (Andrews & McKennell, 1980; Diener, 1984; Ryff & Keyes, 1995; Warr, 1990, 1994), and psychosomatic health (van Horn, Taris, Schaufeli, et al., 2004). However, there is growing evidence that some aspects of health typically researched under the rubrics of stress and well-being share common features, some common influences, and the potential for common policy approaches to reduce their incidence and improve well-being (Lunt, Fox, Bowen, et al., 2007). These areas of health are known as ‘common health problems’ because they have a high incidence in the adult population in developed economies but do not have an easily identifiable physical basis or a well-established dose–response relationship between an identified pathogen and the disease outcome (Lunt, Fox, Bowen, et al., 2007). Examples include anxiety, depression, muscular–skeletal problems, and cardiorespiratory illnesses. Importantly, for work stress and well-being research, there is evidence that common health problems might all be linked to psychosocial work characteristics such as job autonomy and social support (Lunt, Fox, Bowen, et al., 2007). Moreover, there is evidence that specific aspects of work-related well-being, including cognitive, affective, and psychosomatic components, are related to a single underlying latent dimension, with the affective component having the closest relationship to this single dimension (van Horn, Taris, Schaufeli, et al., 2004). There are other reasons for thinking that affect is a centrally important factor in research on stress and well-being. First, a common thread in much research in this area is that unpleasant affective reactions to work can be deleterious to psychological health, physical health, and behaviours related to performance (Danna & Griffin, 1999; Schwartz, Pickering, & Landsbergis, 1996; Spector & Goh, 2001). Secondly, affective reactions are thought to be the central components of well-being (Diener & Larsen, 1993; Warr, 1994) that mediate the influence of the environment on more cognitive and summative aspects of well-being, such as competence, aspiration, autonomy, integrative functioning, and satisfaction (Andrews & McKennell, 1980; Diener, 1984; Ryff & Keyes, 1995; Warr, 1990, 1994). Affective well-being can be defined as the relative experience of pleasant affects to unpleasant affect (Diener & Larsen, 1993). This locates well-being as a phenomenological experience. The dominant view of affect represents affective experience as located in a two-dimensional space, where the dimensions are positive and negative affect (Watson & Tellegen, 1985). Specific affects cluster within this space (Haslam, 1995), so that more specific dimensions of affective well-being can be identified that are related to affective states such as
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
5
anxiety, anger, tiredness, pleasure, and enthusiasm (Daniels, 2000). Affective well-being is domain specific (Warr, 1990). At least in relation to the work domain, the two major dimensions of affect may have different weightings in relation to affective well-being or be represented by a hierarchical structure (Cropanzano, Weiss, Hale, et al., 2003), with the more specific elements (e.g., anxiety, anger, enthusiasm) being most closely related to a superordinate dimension related closely to displeasure–pleasure (Daniels, 2000). Poor affective experience is often a diagnostic criterion for poor psychological health (Lindsay & Powell, 1994) whilst good affective experience is an indicator of good subjective well-being. Physical health and well-being are not usually defined as central aspects of subjective well-being; rather, links between affect and physical health are explained by other processes. In tripartite models, factors related to the psychosocial work environment (e.g., job characteristics such as the level of job demands) and personality factors combine to produce an emotional response, which in turn might influence the course of a specific disease, if the person has a physiological susceptibility to develop that disease (Schwartz, Pickering, & Landsergis, 1996) or a vulnerability to engage in behaviours that might influence susceptibility (Lunt, Fox, Bowen, et al., 2007). This susceptibility need not be genetic, but can be transient (e.g., salt in diet) or a combination of both (e.g., cholesterol levels). Moreover, affective reactions to the work environment might also influence how symptoms are experienced, presented to medical practitioners, and diagnosed (Daniels, Jones, Perryman, et al., 2004; Lunt, Fox, Bowen, et al., 2007). In summary, although stress and subjective well-being are complex areas of study that subsume a range of specific behavioural, physiological, cognitive, and affective elements, the centrality of affect to well-being may underpin a range of processes that suggest certain features of the work environment influence, however distally, a range of aspects of psychological and psychosomatic health and well-being.
MANAGING PSYCHOSOCIAL HAZARDS Adverse job characteristics are also know as psychosocial hazards because of the amount of research indicating they are risk factors for poor well-being and ill-health (MacKay, Cousins, Kelly, et al., 2004; Rick et al., 2002). Examples include high job demands, low job control, low support from coworkers, lack of role clarity (Cousins et al., 2004; Rick, Thomson, Briner, et al., 2002) amongst many others embedded in typologies of job characteristics (e.g., Cooper & Marshall, 1976; Karasek & Theorell, 1990; Warr, 1987). Such psychosocial hazards are often treated as if they are objective features of the work environment (MacKay, Cousins, Kelly, et al., 2004; Schaubroeck, 1999), from which it is possible to determine workers’ exposure to the hazards, and then primary prevention strategies can be developed to limit the level of exposure to the hazard (Cox, Griffiths, Barlowe, et al., 2000; Schaubroeck,
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
6
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
1999). As psychosocial hazards have been identified as risk factors, a sensible approach to policy and practice might focus on controlling these hazards.2 This approach is embedded in policy and guidance on the implementation of policy from many supra-national organizations (ETUC, 2004; ILO, 2001; Leka & Cox, 2008; Leka, Griffiths, & Cox, 2003). In the EU, there is no distinction in law between risks to workers’ physical health and risks to psychological health, and employers are required to conduct risk assessments and take preventive action to reduce exposure to health and safety risks (ETUC, 2004). One of the most sophisticated approaches to policy on work-related stress has been developed and implemented in recent years in the UK. This is the UK Health and Safety Executive’s Management Standards for WorkRelated Stress. This approach provides a validated questionnaire to enable employers to assess psychosocial hazards in their workforce (an indicator tool), a process to take action when hazards are identified, and target states to be achieved in the workforce as a whole as indexed by the indicator tool (Cousins, MacKay, Clarke, et al., 2004; MacKay, Cousins, Kelly, et al., 2004). Because the Management Standards approach is targeted at controlling psychosocial hazards, rather than their consequences, and regulating their level against predetermined targets, the approach is very strongly predicated upon changes in job and organizational processes in the effective management of stress and well-being (MacKay, Cousins, Kelly, et al., 2004). The Standards approach is concerned with regulating six psychosocial hazards: demands, (low) job control, (low) support, (poor) relationships at work, (poor) role clarity, and management of change. The approach has many positive features, not least of which include: the development of the standards from existing scientific evidence (Rick, Thomson, Briner, et al., 2002); the care, attention to detail, and involvement of multiple stakeholder group inherent in operationalizing the Standards (Cousins, MacKay, Clarke, et al., 2004); and the relative ease with which the Standards allow UK companies to comply with UK and EU Health and Safety Legislation in the area of psychological health and well-being. Examining the assumptions underpinning the Standards provides a useful test case as to whether similar approaches could be effective in other countries, as well as whether the UK’s approach could be developed to enhance policy and guidance for managers.3 The Standards approach is based on several
2
3
Note that inherent in ideas of psychosocial risk management is the idea of probability. That is, exposure to psychosocial hazards at work increases the chances of disease or attenuated wellbeing, but for any particular individual at any given time other protective factors may prevent harm. A Delphi study has also examined the strengths and weaknesses of the practical application of Management Standards (Cox, Karanika-Murray, Griffiths, et al., 2009) and Kompier (2004) stated some concerns about the operationalization of the Management Standards in practice. The point of departure for this review is that it is concerned with the theoretical basis of the Standards.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
7
assumptions. These assumptions are not all unique to the Standards and many elements are present in World Health Organization, ILO, and EU guidance (ETUC, 2004; ILO, 2001; Leka & Cox, 2008; Leka, Griffiths, & Cox, 2003). Because the Standards are based on prevailing scientific practice, the Standards also reflect assumptions that are commonly held by I/O psychologists working in the work stress and health area, or at least assumptions concerning how the results of research can be turned into practice (see e.g., Hurrell, Nelson, & Simmons, 1998; Schaubroeck, 1999). Some of the assumptions underpinning the Standards approach can be stated as follows. First, job characteristics are objective properties of jobs and relatively stable. Secondly, job characteristics have probabilistic relationships with well-being and health, which are usually assumed to be monotonic. Thirdly, organizations can take action so that relatively stable job characteristics can be changed to new, stable levels, which then will lead to improvements in health and well-being because stress-related problems will be prevented. Fourthly, because of probabilistic relationships between job characteristics and well-being and health, differences between individuals are largely irrelevant to the prevention of stress-related ill-health.4 However, each of these assumptions is problematic. First Assumption: Job Characteristics are Objective Properties of Jobs and Relatively Stable This assumption has two parts. First, job characteristics are objective. Put another way, job characteristics are properties of jobs and independent of the person doing the job – two people doing exactly the same job are assumed to be exposed to the same level of hazards. Secondly, job characteristics are relatively stable – over time, people more or less experience the same level of job characteristics. Without either part of this assumption, the idea that organizational action can lead to lasting improvements in jobs across people is not strictly tenable. The assumption of objectivity is easy to question (Daniels, 2006). An overview of some of the main journals in this area indicates that the most prevalent means of assessing job characteristics is by questionnaire, and the use of a questionnaire to assess psychosocial hazards is embedded in the Management Standards approach (Cousins, MacKay, Clarke, et al., 2004; MacKay, Cousins, Kelly, et al., 2004). Many critiques of self-report measures have been published that highlight limitations such as self-reports are subject to numerous biases, including those linked to trait affect (Brief, Burke, George, et al., 1988; Spector, 1992) and social interaction (Salancik & Pfeffer, 1978).
4
The Management Standards do indicate that systems need to be in place to respond to individual concerns but do not incorporate individual factors either in any detail or explicitly.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
8
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
However, other methods used to assess job characteristics do not return objective assessments either (see Daniels, 2006). Using reports from line managers or researchers’ observations of work merely reflects someone else’s perception of a job and cannot access cognitive activities that typify many jobs in advanced knowledge-based economies. Aggregating reports from all job holders with the same job title reflects shared perceptions rather than necessarily any objective reality. Even where self-reports from geographically separated jobs are aggregated, common institutional practices, such as socialization, training, or industrial relations, may condition perceptions (cf. Scott, 1995). Coding and analysing jobs based on nationally representative databases is also problematic, as such codes represent occupations rather than the local conditions of specific jobs in specific organizations. A commonly suggested solution to the problem of imperfect measures of objective job characteristics is triangulation of results across several methods, on the assumption that common patterns of relationships indicate something about the objective nature of certain job characteristics. However, triangulation does not always work: job characteristics assessed by different methods do not always correlate sufficiently well to indicate the different methods are assessing the same thing, and different methods can produce associations with different indices of health or make independent contributions to predicting the same indicators of health (Daniels, 2006). These problems with attempting to assess ‘objective’ job characteristics led to the conclusion that a better strategy might be to abandon the idea that job characteristics somehow reflect unitary concepts, but that different methods reflect different facets of job characteristics (Daniels, 2006). For example, a job characteristic like job control does not exist per se; rather, different methods for assessing job control assess different facets of job control that may be interdependent but are nonetheless different from each other. In this scheme, self-reports of job characteristics assess perceptions of work, nothing more. The assumption of stability is also questionable. Treating job characteristics as stable does not reflect the dynamics of organizational life (Peterson, 1998; Weiss & Cropanzano, 1996), that proximal events can have a stronger influence on well-being than more distal events (Pillow, Zautra, & Sandler, 1996), or empirical evidence that levels of job characteristics can change from week to week (e.g., Totterdell, Wood, & Wall, 2006), from day to day (e.g., Daniels & Harris, 2005), and even within the same day (e.g., Daniels, Boocock, Glover, et al., 2009). Moreover, changes in the nature of work, with a shift to more complex and interdependent jobs, have also questioned the assumption of treating jobs characteristics as stable (Grant & Parker, 2009). In the 27 European countries in 2005, including former Warsaw Pact transition economies, 81% of workers reported solving unforeseen problems of work, 69% reported learning new things at work, and 72% reported assessing the quality of their own work (Parent-Thirion, Fern´andez Mac´ıas, Hurley, et al., 2007). Such is the nature of modern working environments that Grant and Parker stated
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
9
that changes in the nature of work ‘challenge fundamental assumption that employees passively carry out static jobs and tasks assigned by managers’ (2009: 342). Second Assumption: Job Characteristics Have Probabilistic Relationships with Well-Being and Health The number of reviews and meta-analyses synthesizing evidence indicates that there are reliable linear associations between indicators of job characteristics and indicators of well-being and health (e.g., Cass, Faragher, & Cooper, 2002; de Lange, Taris, Kompier, et al., 2003; Jackson & Schuler, 1985; Rick, Thomson, Briner, et al., 2002). The problem here is not so much whether the assumption is tenable, rather whether the probabilities are high enough and there is enough specificity in the nature of the relationships to be able to develop organizational practices that will improve health and well-being (Rick & Briner, 2000). The problem of poor-specificity and poorly specified dose–response relationships is well known in the area (de Jonge & Dormann, 2006; Lunt, Fox, Bowen, et al., 2007). Most of the policy guidance in the area is typified by a tacit assumption of linear relationships between job characteristics and health, although more complex relationships are entirely feasible (Karanika-Murray, Antoniou, Michaelides, et al., 2009). One problem with poor specificity is that in any one instance of organizational change, it will not be possible to specify a priori what will be the best health and well-being outcomes to measure in order to determine whether the changes actually had any impact on well-being. To some extent, this can be overcome in assessment studies by taking a broad range of measures and including affective well-being amongst this range, because of its centrality to well-being. A more important problem is that even if the idea is tenable that changing job characteristics can improve well-being at a population level, because of poorly specified and stochastic relationships, managers might be required by health and safety law to implement costly organizational change to redesign jobs to protect health with no guarantee of any local health benefits, yet the changes may be sub-optimal for operational and financial performance. Third Assumption: Organizations can Take Action which then Will Lead to Improvements in Health and Well-Being Perhaps reflecting problems with the appropriate level of specificity to take organizational action to effect beneficial changes in job characteristics, there are questions over the extent to which organizational changes can influence wellbeing. In a recent review of 90 intervention studies, LaMontagne, Keegel, Louie, et al. (2007) concluded both organizationally focused and individually focused interventions (e.g. coping skills training) had positive effects on employee health. In a systematic review of interventions focused on task
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
10
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
restructuring, Bambra, Egan, Thomas, et al. (2007) concluded interventions focused on autonomy had the strongest effects on health. However, in their review of 19 interventions, the strongest evidence came from task restructuring that reduced job autonomy and was accompanied by subsequent deterioration in health rather than from interventions designed to improve job autonomy. Moreover, across all studies, less than 30% of health outcome measures changed in the predicted direction. Two studies found health outcomes changed in the direction opposite to that predicted. Egan, Bambra, Petticrew, et al. (2007) reviewed 18 intervention studies and concluded interventions with a focus on improving workers’ opportunities to exercise control and participate in decisions improve health. Two studies found evidence of deterioration in health subsequent to the intervention, which Egan et al. attributed to confounding due to organizational downsizing. In the other studies, less than 45% of the health outcomes assessed changed in the predicted direction. All of these reviews included studies with either no control groups or non-equivalent control groups. Randomized control trials reflect the most rigorous designs to make causal inferences concerning the effectiveness of interventions. In a meta-analysis of randomized controlled trials of interventions with working adults from healthy populations, Richardson and Rothstein (2008) found that organizationallevel interventions did not have a statistically reliable association with outcome measures (d = 0.14, ns), whereas individually focused interventions tended to have statistically reliable associations with outcome measures. Cognitive–behavioural interventions demonstrated the strongest impact (d = 1.16, p <0.01). Although Richardson and Rothstein found only five studies of organizational interventions that met the criteria for inclusion in the study, their results do raise questions concerning the effectiveness of organizational level interventions given null results under the most rigorous conditions of random allocation to treatment and control groups. Given the findings of the other recent reviews (Bambra, Egan, Thomas, et al., 2007; Egan, Bambra, Petticrew, et al., 2007; LaMontagne, Keegel, Louie, et al., 2007), the best that can be said is that interventions aimed at changing work and organizational structures, practices, and processes might have a subsequent beneficial effect on some health outcomes, but these effects might be specific to interventions that target only certain job characteristics rather than a broad range, and there is a small chance of interventions producing harmful effects. Fourth Assumption: Differences between Individuals are Largely Irrelevant to the Prevention of Stress-Related Ill-Health A major issue in developing policy on work-related stress is the extent to which self-reports of health and job characteristics reflect real differences in health and jobs rather than differences between individuals in their personality,
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
11
attitudes, self-concept, etc. (Daniels, Jones, Perryman, et al., 2004). One reason for this is because self-reports are one of the most efficient means of collecting data from the population in order to monitor the impact of policies developed to reduce stress and increase well-being. Indeed, one major debate has centred on negatively oriented personality traits, and whether failure to control for trait negative affect leads to inflated correlations between selfreports of job characteristics and health and well-being (e.g., Spector, Zapf, Chen, et al., 2000). However, it is also likely that differences between individuals could have a substantive role in the relationship between work and well-being (Spector, Zapf, Chen, et al., 2000). The question for policy then, is not so much whether individual differences influence the development of stress-related ill-health, but whether work has a sufficiently large influence on stress-related ill-health, after taking into account differences between individuals, to warrant interventions targeted at jobs and organizations. A recent meta-analysis examined 19 longitudinal studies that assessed reports of psychological and physical symptoms, adverse job characteristics and negative oriented personality traits (Ferguson, Daniels, & Jones, 2006). Negatively oriented personality refers to individual differences such as neuroticism, negative affectivity, and low self-esteem. In meta-regression analyses controlling for baseline measures of symptom reports, baseline reports of adverse job characteristics did not have a significant relationship with subsequent assessment of physical symptoms (β = 0.01, ns) but did have a relationship with subsequent reports of psychological symptoms (β = 0.05, p <0.05). In contrast, baseline measures of negatively oriented personality had relationships with subsequent reports of both physical symptoms (β = 0.05, p <0.01) and psychological symptoms (β = 0.09, p <0.01). In all cases, there were relationships between baseline assessments of symptoms and adverse job characteristics, after controlling for negative oriented personality, but the relationships between baseline symptoms and negatively oriented personality were higher. The results from this meta-analysis indicate that the strongest relationships were between negatively oriented personality and symptom reports and that adverse job characteristics did not have a predictive relationship with physical symptoms. Another issue tied up with the Management Standards and other approaches based on preventing ill-health by changing job characteristics is that they do not address the issue of people who may be in work but are suffering from some form of illness. The issue is significant. For example, over 30% of workers were estimated to have a health-related work limitation in one study in the USA (Lerner, Amick, Malspeis, et al., 2000). Further, in any one workplace, the levels of people suffering from chronic illnesses may be underestimated as not everyone discloses their illness to their line manager (Munir, Leka, & Griffiths, 2005). Moreover, it cannot be assumed that interventions designed to protect health will have a curative effect for those that have already developed health problems.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
12
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Workers’ illnesses need not have their generation in work. However, work may affect levels of well-being for workers with illnesses. Munir, Yarker, Haslam, et al. (2007) found that after taking into account illness type and severity, distress in those at work with chronic illness was related to restrictions in ability to perform work due to illness, presenteeism (i.e., attending work when feeling unwell), poor management of illness at work, and low workplace support. Reasons for higher levels of presenteeism amongst those with a chronic illness could be attributed, at least partly, to people trying to avoid disciplinary action for absences (Munir, Yarker, & Haslam, 2008). Workers may be more likely to disclose illness if they feel they may receive support (Munir, Leka, & Griffiths, 2005), although those suffering from anxiety and depression are less likely to receive support in the form of adjustments to their work tasks to take into account their illness (Munir, Jones, Leka, & Griffiths, 2005). Summary Approaches to enhancing well-being and reducing stress-related ill-health based on the management of psychosocial hazards or job characteristics may be sub-optimal because some of the underpinning assumptions are questionable, at least to some extent. First, job characteristics may not necessarily be objective and stable properties of jobs that are independent of the person performing the job; secondly, the low specificity and stochastic nature of dose–response relationships between psychosocial hazards and specific outcomes means that attempts to alter stable and objective properties of jobs are not guaranteed to be successful; thirdly, the evidence from intervention studies is that changing psychosocial job characteristics may have limited success in protecting health and well-being; fourthly, differences between individuals, including in their health status, are important factors in preventing stress-related ill-health and enhancing well-being. In the following sections, perspectives will be introduced that go someway to addressing these problems. These perspectives emphasize the agency of workers in shaping their jobs, the interpretation of work events and adaptive adjustment to adverse work conditions, and the role of workers, individually and collectively, in helping or hindering the effectiveness of interventions designed to protect health and improve well-being through altering work characteristics.
REVISIONING JOB CHARACTERISTICS If job characteristics are neither necessarily stable nor objective, then there are questions as to what measures of job characteristics actually assess. To address this issue, Daniels (2006) suggested abandoning the idea of a unitary concept of job characteristics, instead breaking them down into different, interdependent
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
13
facets that correspond to what is actually being measured by different methods. For example, self-reports of job autonomy assessed by questionnaire provide information on one facet of job autonomy, but managers’ reports of subordinates’ autonomy provide information on another facet of job autonomy. Daniels suggested three facets of job characteristics: latent, which reflects facets of job characteristics embedded in organizational and technological processes; perceived, which can be disaggregated into a person’s perception of their own job and a person’s perception of someone’s else job; and enacted job characteristics. Enacted job characteristics are jobs as they come to be shaped by job incumbents and others, and in Daniels’ (2006) scheme, reflect how interdependencies with others and proactive behaviour come to shape jobs (cf. Grant & Parker, 2009). In the following sections, each facet of job characteristics identified by Daniels (2006) is described in more detail. Latent Job Characteristics Latent job characteristics are the institutional and technological processes that influence work tasks (Daniels, 2006). Examples might be contractual hours as indicators of quantitative workload or temporary versus permanent employment contracts as an indicator of job security. In some ways, these might be considered to be the most ‘objective’ indicators of job characteristics, as they can be assessed with techniques that do not rely on perceptual data. However, it is unlikely that assessing latent job characteristics provides a sound basis for finding close relationships between work and well-being or health. This is because they neither reflect what people do, how they perceive what they do, nor what happens to them, all of which could have closer relationships with well-being and health. For example, taking contractual work hours as an index of quantitative workload does not reflect how many hours many people work and whether they volunteer or feel coerced into working longer hours. The choice to work longer hours itself may alter the perceptions of working longer hours, thus possibly influencing well-being and health (Lazarus, 1999). However, latent job characteristics may provide constraints within which people act, but may themselves be subject to change as individuals, alone or collectively, formally renegotiate contractual terms, organizational processes, and how technologies are used (Grant & Parker, 2009). Latent job characteristics may also provide information that influences perceptions of job characteristics (Daniels, 2006). Perceived Job Characteristics Perceived job characteristics reflect generalized perceptions of a person’s job, and are assessed by reports of job characteristics often collected by structured questionnaire measures that ask observers to rate jobs as they are perceived to be over an extended or unspecified period (Daniels, 2006). Potential perceivers
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
14
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
include the job incumbent, co-workers, managers, customers, and researchers. Although researchers’ perceptions are often treated as if they reflect an objective assessment of a job, they are still perceptual data and are likely to provide only a partial picture of what it is people do in their day-to-day work. While researchers’ perceptions may not influence incumbents’, managers’, co-workers’, or customers’ perceptions, they are likely to be mutual influences between the perceptions of those people that interact with the job incumbent in the course of performing their job. Partly this might because they are all observing the same phenomena; but also because there may be a shared awareness of latent job characteristics; because of social information exchange within an organization (Salancik & Pfeffer, 1978); because of shared socialization, training, and other processes that may be common to a job within an organization, sector, or profession (DiMaggio & Powell, 1983); and/or partly because powerful groups of workers are able to claim and then protect the right to certain job characteristics, such as skill use, regardless of job descriptions, technology, and other factors related to latent job characteristics (Noon & Blyton, 1997). Research indicates that self-reports of job characteristics may have a predictive relationship with indicators of psychological health at least (Ferguson, Daniels, & Jones, 2006). However, how job incumbents generally perceive their work need not be the process through which work influences health and well-being. It is possible that incumbents’ and others’ perceptions influence what happens on a moment-by-moment basis, within weak or strong constraints of latent job characteristics, and these dynamic processes as jobs are enacted are the appropriate units of analysis for studying the psychological and physiological processes linking work to well-being and health. Enacted Job Characteristics Enacted job characteristics are events that reflect jobs as they happen (Daniels, 2006). The idea of enacted job characteristics is strongly influenced by the notion of job crafting (Wrzesniewski & Dutton, 2001) and the idea that dynamic events rather than static conditions offer more potential for explaining the experience of work (Peterson, 1998; Weiss & Cropanzano, 1996). Enacted job characteristics are viewed as dynamic and emergent from how an individual does a job, rather than as static and determined by organizational and technological processes. They can reflect behavioural, discursive, or cognitive processes. Daniels (2006) considers that enacted job characteristics are influenced by the job incumbent’s own perceptions of what the job is or should be, what others think the incumbent’s job is or should be, and latent job characteristics. The incumbent’s own perception of his or her job is an important influence on enacted job characteristics because individuals are active in interpreting and shaping their jobs through their behaviour (Wrzesniewski & Dutton, 2001). For
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
15
example, a person may decide to leave work early because he or she believes that he or she has autonomy over his or her work schedules. Co-workers can behave in ways that influence enacted job characteristics (Grant & Parker, 2009), for example by providing ambiguous information where the incumbent’s tasks are dependent upon clear communication from co-workers. Moreover, as well as acting as individuals, workers may collectively enact certain job characteristics (Grant & Parker, 2009). Managers are likely to be particularly influential in respect of enacted job characteristics, because of normative perceptions concerning influence over others (cf. Clegg & Spencer, 2007). Latent job characteristics may act as constraints or facilitators on what can be enacted. Some latent job features, relating to jobs that require high levels of training, skill use, and discretion, may stimulate individuals to enact specific job characteristics to shape what they do and the physical environment in which they work (Grant & Parker, 2009; Pierce, Jussila, & Cummings, 2009). In turn, enacting beneficial job characteristics may lead to circular and dynamic changes in work, so that the potential to enact more beneficial job characteristics is increased (Clegg & Spencer, 2007). As enacted job characteristics are behaviours as they occur, they may be seen as discrete events or episodes that influence affective experience at work (Weiss & Cropanzano, 1996) and that emerge and change over the course of a working day (Daniels, 2006). Support for giving greater causal credence to recent, on-going events comes from studies of well-being that show more recent events have a stronger effect than events in the more distant past (Suh, Diener, & Fujita, 1996) and that show the impact of major life events on well-being is mediated by daily events (Pillow, Zautra, & Sandler, 1996). Enacted job characteristics as events or episodes can be thought of in two ways (Daniels, 2006; Daniels, Harris, & Briner, 2004): first, as the locus of cognitive processes that shape affective reactions to events and episodes (Lazarus, 1999); secondly, individuals might enact job characteristics for specific purposes. The first view reflects the many cognitive theories of affect in which the appraisal of how events alter progress towards personal goals is a key determinant of affective reactions to those events (Power & Dalgleish, 2008). These cognitive processes are one way in which enacted job characteristics might influence health through an impact on the cognitive experience of affect and attendant changes in physiological processes and behaviours. The second view posits that individuals enact job characteristics to achieve specific goals or meet particular needs, such as to exercise control, find meaning in work, develop a positive self-image, affiliate with others, or to make the tasks of work consonant with the incumbent’s self-identity (Pierce, Jussila, & Cummings, 2009; Wrzesniewski & Dutton, 2001). Enacting job characteristics for the purposes of achieving specific goals or meeting particular needs echoes goal based theories of affect. However, job characteristics might be enacted by job incumbents specifically to cope with the other enacted job characteristics (Daniels, Harris & Briner, 2004). For example, a person many enact job control to take
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
16
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
a break from working on a complex and demanding task that a manager has requested be completed. One important differentiation between the notion of enacted job characteristics and more traditional approaches that underpin much science and policy is that enacted job characteristics relate to a person doing a job at a specific point in time. In contrast, the dominant approach treats the person and the job as separate entities. This dualist notion of job characteristics and workers inherent in the dominant view is perhaps both artificial and too far from reality to be practically tenable. The notion of enacted job characteristics, as they apply to understanding stress and well-being, gives central importance to workers as active sense makers of their work as it happens and as agents that can shape their work. This need not imply deliberate and conscious choices on the part of workers; interpretation and agency can also reflect unconscious and habitual processes. In the following sections, cognitive processes, both conscious and deliberate and unconscious and automatic, will be outlined. These processes may help explain both how enacted job characteristics are interpreted and the choices that lead to enacting job characteristics to cope with potentially harmful events and episodes as they occur at work.
REVISIONING THE WORKER AS AN ACTIVE INTERPRETER The idea that stress and well-being are influenced by how we interpret events and episodes and not the events and episodes themselves has a long history in psychology (for reviews see e.g., Lazarus, 1999; Power & Dalgleish, 2008). Central to many interpretive approaches is the idea that how individuals appraise episodes and events as relevant to their goals influences their affective experience, in turn influencing their well-being (for a detailed review see Power & Dalgleish, 2008). Inherent in these models is the idea that affect signals whether things are going well, and therefore cognitive resources can be directed elsewhere, or whether things are going not so well and something needs to be done (e.g., Frijda, 1986; Oatley & Johnson-Laird, 1987). Perceived rate of progress toward a goal might influence the level of positive affect and perceived impediments might influence negative affect (Carver & Scheier, 1990). There is evidence to indicate that appraisals of work-related goal progress are associated with well-being (e.g., ter Doest, Maes, Gebhardt, et al., 2006), although these relationships may be moderated by a number of factors. Workrelated well-being seems to be associated most closely with progress towards goals that are considered to be more important (Harris, Daniels, & Briner, 2003) or consonant with other goals (Kehr, 2003), including higher order goals close to the self-concept (Maier & Brunstein, 2001). Similarly,
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
17
well-being seems to be more closely associated with goal progress when progress exceeds expectations, because goals are appraised as difficult (Wiese & Freund, 2005) or because initial appraisals of the likelihood of attaining the goals were low (Pomaki, Karoly, & Maes, 2009). However, appraisal theories have largely been ignored, or downplayed, by I/O psychologists (see, e.g., Cooper, Dewe, & O’Driscoll, 2001; Dewe & Cooper, 2007). There may be two main reasons for this. First, appraisal-based theories give primacy to individual interpretation. For some I/O psychologists and stakeholders, such as trades union officials, focusing on individual processes may seem like ‘blaming the victim’ for exposure to adverse work conditions (cf. Murphy, 1988). Secondly, assessing job characteristics that most people would ‘appraise’ in the same way might lead to the development of interventions based on changing those job characteristics that most people would appraise in the same way. One basis for this assertion is that work occupies a unique niche because of the economic transaction that takes place, rendering it a situation ripe for common interpretation (Brief & George, 1991). The problem with this perspective is that work might mean different things to different people, with the economic motive being one of only several reasons why people apply for different jobs and pursue different careers (cf. Holland, 1985). This does not mean that common patterns of appraisals within work groups are unimportant, but rather these patterns might be localized to specific work groups within organizations (Harris, 2002). Moreover, examining individual patterns of appraisals may be one way to incorporate differences between individuals, whether these differences are due to personality traits or health status, more fully into research, policy, and practice in work stress and well-being. In considering appraisal based theories, four questions are pertinent: (i) What is being appraised? (ii) How do appraisals influence well-being? (iii) How do appraisals come to be? (iv) What happens after appraisal? A fifth question, how people decide to cope with what is being appraised, is dealt with in another section. What Is Being Appraised? In most appraisal theories, discrete events or episodes can form the basis of appraisals (e.g., Lazarus, 1999; Power & Dalgleish, 2008). This fits well with notions of enacted job characteristics, which are conceptualized as dynamic. The idea that discrete workplace events or episodes form the basis of affective reactions gained a lot of momentum with the publication of Affective Events Theory (AET, Weiss & Cropanzano, 1996) and subsequent support and extensions of AET (e.g., Fisher, 2000, 2002; Fuller, Stanton, Fisher, et al., 2003). AET suggests that work-related attitudes and behaviours are influenced by work-related affect and that work events mediate the impact of more stable
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
18
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
on-going work conditions on affect. AET positions affect as the central reactive phenomenon to work events, which is consistent with viewing affective experience as central to well-being (cf. van Horn, Taris, Schaufeli, et al., 2004) and appraisal theories that are concerned with explaining affective reactions (Power & Dalgleish, 2008).
How Do Appraisals Influence Well-being? Goal-based appraisal theories indicate that events appraised as influencing goals influence affect (e.g., Lazarus, 1999). In recent dual-process approaches, appraisals are conceived of operating through slow, deliberative, and detailed cognitive processes, or on the basis of fast, restricted, unconscious cognitive processing (Power & Dalgleish, 2008; Smith & Kirby, 2001). These two processes of appraisal reflect the distinction between controlled and automatic processing (Schneider & Chein, 2003; Schneider & Shiffrin, 1977). Controlled processing is constrained by attentional limits, but is flexible and under greater intentional control than automatic processing, and often includes conscious thought. Controlled processing can be influenced by information held in long-term memory, but is influenced more by information attended to in the environment than automatic processing. Automatic processing is unintentional, involuntary, occurs outside of awareness, is only minimally influenced by information in the environment, and there is a strong influence of information recalled from long-term memory. Automatic processing is less effortful, faster, and affords parallel processing of multiple cognitive tasks, but is less flexible or adaptable to changes in the environment than controlled processing. Distinguishing controlled from automatic processes does show promise for broadening our understanding of job design and affect in the workplace (Elfenbein, 2007; George, 2009; Parker & Ohly, 2008). One of the more sophisticated dual-process models of appraisals has been developed by Power and Dalgleish (2008). In their model, Power and Dalgleish proposed that mental models, or networks of beliefs concerning how events influence goals or affect, are the basis upon which work events are appraised and consequently alter the cognitive experience of affect. Daniels, Harris, and Briner (2004) extended Power and Dalgleish’s model in three ways. First, Daniels, Harris, and Briner developed the model to be applicable to work contexts. Secondly, they specified the processes by which work place events and episodes become the focus of affect-related cognitive processes. Thirdly, they explicitly made provision for coping in the model (see following sections). Following Power and Dalgleish, Daniels, Harris and Briner proposed that in controlled processing, mental models are fluid and developed to be specific to events as they happen, and change according to how events unfold, although there is an influence from long-term memory concerning similar events and
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
19
episodes encountered in the past. In automatic processing, Daniels, Harris, and Briner propose there is a very strong influence of long-term memory concerning events encountered in the past. In relation to how events and episodes at work initially influence affect, prior to engaging coping responses, three processes are relevant. First, the event or episode must be perceived, consciously or unconsciously, and then categorized as affectively toned. Categorization enables individuals to make inferences by using information on similar events or episodes encountered in the past to predict possible consequences of the current situation (Smith & Medin, 1981). In other words, what an individual believes might happen in current circumstances will be influenced by what that individual believes happened in previous circumstances that were somehow similar. Evidence indicates that people categorize events (Rifkin, 1985); people use affective information in categorization tasks (Halberstadt & Niedenthal, 1997; Niedenthal, Halberstadt, & Innes-Ker, 1999); and people can articulate events linked to particular categories of affects (Basch & Fisher, 2000). Daniels, Harris, and Briner (2004) proposed that controlled categorization of events: is more environmentally driven; involves categories that change according to circumstances; is related to pursuit of current goals; and involves inferences concerning how classes of events may impact upon pursuit of current goals. In contrast, Daniels, Harris, and Briner propose automatic categorization of events is based on a few fixed categories held in long-term memory and based on comparing the features of the current event with the features of events encountered in the past: there are no inferences concerning current goal progress in automatic categorization of work events. However, the precise means of how people categorize work events needs to be explored further empirically. Controlled categorization of work events enables inferences concerning goal progress. Where individuals make an inference that events are perceived to enhance progress towards goals, then Daniels, Harris, and Briner proposed pleasant affect will be experienced. Where individuals make an inference that progress towards goals will be hindered or halted, then unpleasant affects will be experienced. More specific propositions link specific inferences (e.g., concerning threat to goal progress or goal failure) to specific affects (e.g., anxiety, sadness). One important question concerns how individuals determine an acceptable rate of goal progress. Recently, Warr (2006, 2007) has proposed that thresholds on what is acceptable in relation to well-being are based on a number of judgments; for example, comparisons with other people, with the past, with an imagined future, and with expectations in a given situation. Notwithstanding the complexity of determining an acceptable rate of goal progress, there is evidence that the impact of adverse job characteristics is related to beliefs concerning those job characteristics’ influence on personal goal progress. In two studies, Harris and Daniels (2005, 2007) found affective well-being and job satisfaction at the end of working days was associated
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
20
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
with daily beliefs concerned with work demands’ influence on goal progress, with more unpleasant affect and low job satisfaction associated with beliefs that work demands impacted on goal progress adversely. In these studies, daily beliefs were taken to reflect daily appraisals of work demands. Daily levels of demands were controlled statistically. In two other samples, beliefs concerning a specific adverse job characteristic’s influence on goal progress assessed prior to an event sampling study were found to moderate the impact of the same job characteristic on negative affect during the period of the event sampling study (Daniels, Hartley, & Travers, 2006). Beliefs assessed prior to the event sampling study were taken to reflect information held in long-term memory upon which people draw to form appraisals in a given situation. The form of the interactions indicated that the relationship between the adverse job characteristic and negative affect was stronger for people who believed the adverse job characteristic had a detrimental impact on goal progress. Importantly, and reflecting the idea that controlled processing in any specific situation is only weakly informed by long-term memory, this moderating effect was evident only when hourly data gathered across a working week were aggregated. In relation to automatic processing, following Power and Dalgleish (2008), Daniels, Harris, and Briner (2004) proposed that inferences concerning current goal progress are bypassed, as the strong influence of long-term memory overrides current considerations. Instead, in a process like classic conditioning based on associative learning, the features of current events cue the affective experience associated with events with similar features encountered in the past that have had implications for pursuit of personal goals. In other words, repeated exposure to a class of events in the past that was appraised to impede goal progress leads to associations between the features of that class of events and the phenomenological experience of the affects originally experienced. When events are encountered with the same features, the affect is again experienced. Operationalizing these associative links between classes of events and experienced affects as beliefs concerning how events influence affect, two studies have found daily beliefs concerning demands’ influence on unpleasant affect were associated with more unpleasant affect and lower job satisfaction at the end of working days, after controlling for daily levels of job demands (Harris & Daniels, 2005, 2007). In two samples, beliefs that a specific job characteristic influenced negative affect, assessed a few days prior to an event sampling study, were found to accentuate hourly levels of that specific stressor’s association with momentary negative affect assessed at the end of the hour during the period of the event sampling study (Daniels, Hartley, & Travers, 2006). Reflecting the idea that automatic processes are persistent across situations and are not influenced by the specific characteristics of situations, the moderating effect of beliefs concerning negative affect was found for relationships between hourly levels of adverse job characteristics and end of hour assessments of momentary negative affect.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
21
How Do Appraisals Come To Be? Daniels, Harris, and Briner proposed that information from long-term memory has weaker effects on goal-related inferences inherent in controlled appraisal, but stronger influences on affect-related inferences inherent in automatic appraisal. Supporting the influence of long-term memory, Harris and Daniels (2005) reported that beliefs concerning job demands’ influence on both goal progress and affect assessed a few days previously to a diary study predicted end of day beliefs concerning job demands’ influence on goal progress and affect, respectively. Similarly, Daniels, Hartley, and Travers (2006) invoked the notion of the influence of beliefs held in long-term memory on appraisals to explain why relationships between negative affect and adverse job characteristics were moderated by prior assessments of beliefs concerning job characteristics’ influence on goal progress and affect. Whilst the above results indicate that long-term memory may influence appraisals as they happen, they do not explain how beliefs about work events come to be in long-term memory in the first place. This is important in order to develop effective interventions. If appraisals are formed from beliefs that are completely idiosyncratic, then job redesign to make certain work events more likely may not be an effective intervention, as each person will appraise events in different ways. However, if there is commonality between people, then interventions directed at changing collectively shared beliefs or shared work experiences may be effective. At present, there is little evidence on these issues. However, the evidence that does exist (Harkness, Long, Bermbach, et al., 2005; Harris, 2002) indicates that: 1. Person-level variables, such as trait affectivity and attitudes concerning risk, are associated with beliefs concerning work events; 2. The nature of the work situation is also associated with beliefs, so that there is a tendency across people to converge on beliefs concerning specific work events in consistent ways; 3. That within work groups, there may be a convergence of beliefs reflecting social influences. The evidence that does exist indicates some level of idiosyncrasy in beliefs, but also that elements may be shared within specific work groups, organizations, or social groups. What Happens After Appraisal? After the initial appraisal processes, many approaches consider subsequent appraisals concerned with regulating affective experience through coping (Lazarus, 1999). This is discussed in a following section. It is also likely that affective consequences of appraisals influence subsequent appraisals and work events (Daniels, Harris & Briner, 2004). Affect-congruence processes indicate
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
22
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
that people may interpret situations in a way consistent with their affective state (Clore, Wyer, Dienes, et al., 2001; Schwarz, 2000). For example, the experience of negative affect leads to recall of and attention to more negative information. Affect-congruence processes may influence appraisals either positively or negatively. Affect-congruence effects may lead to a level of longer term continuity in appraisals and detecting adverse events in the work environment, leading to spirals of accumulating ill-being or well-being, depending upon the affective starting point (e.g., Christie & Barling, 2009; Daniels & Guppy, 1997; Fuller, Stanton, Fisher, et al., 2003). However, there may be several moderators of this basic process, including ˆ e, 2005a; Lassiter, Koenig, & Apple, 1996), motivational factors (see e.g., Cot´ the extent to which information processing tasks are restricted (Forgas, 2001), and the extent to which relevant beliefs held in long-term memory are elaborate and easily accessed (Fiedler & Bless, 2000; Ruder & Bless, 2003; Williams, Watts, MacLeod, et al., 1996). Summary Cognitive appraisal approaches explain the detailed processes by which work events may influence affect. Yet, there are still issues to be resolved. For instance, although research has found links between presumed controlled and automatic appraisals and general classes of affect, relations between specific affects and specific goal-related appraisals or specific automatic processes have yet to be demonstrated. It is also possible that appraisals related to certain kinds of affect inhibit appraisals related to other kinds of affect (Daniels, Hartley, & Travers, 2006). Moreover, it is not necessarily the case that aspects of well-being, other than the affective aspects, are linked in a straightforward way to appraisals (Harris & Daniels, 2007). The complex mutual relationships between appraisals and their affective consequences need more thorough mapping in order to understand how various states of well-being and ill-being develop. The development of the networks of beliefs that underpin appraisals needs further investigation, which could include considering personal, environmental, or social influences.
REVISIONING THE WORKER AS AN ACTIVE AGENT Given the view that workers are active shapers of their jobs, identifying ways in which workers can protect and enhance their own well-being carries an important practical imperative (Daniels & de Jonge, 2010). However, coping and affect regulation are complex constructs that can be difficult to assess (Dewe & Cooper, 2007). Coping can be deliberate and controlled, or automatic and unconscious (Cramer, 2000). Coping consists of at least three elements: coping function, resources, and coping behaviour (Latack & Havlovic, 1992; Latack, Kinicki, & Prussia, 1995; Lazarus, 1999; Skinner, Edge, Altman, et al., 2003).
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
23
Coping function is the target of coping, and includes targets such as altering situations, regulating appraisals, or affective expression (Gross, 2008), and can also reflect approaching or avoiding/suppressing situations, thoughts, affects, and affective displays (cf. Latack, 1986; Latack & Havlovic, 1992; Roth & Cohen, 1986). Coping behaviours are the behaviours used to fulfil functions, such as changing work schedules to spend more time on a problem, talking to colleagues to distract oneself from thoughts of an upcoming performance review and so on. Resources are those things that enable individuals to perform coping behaviours. In the work environment, resources are typically considered to be job characteristics such as job control and social support (Karasek & Theorell, 1990; cf., Demerouti, Bakker, Nachreiner, et al., 2001). Coping resources embedded in job design may be viewed as latent job characteristics, which may enable or constrain action. As resources enable coping behaviours, they are more distal in the coping process than behaviours and functions, and so may not have a direct relationship with the outcomes of coping. The idea that resources allow specific behaviours to be enacted for a purpose echoes the idea of enacted job characteristics that can be enacted by people for specific purposes (see p. 15). When specific behaviours are enacted that involving shaping the work environment, enacted job characteristics and coping behaviour–function combinations become synonymous. Importantly, coping behaviour and function need to be considered together (Daniels, Beesley, Cheyne, et al., 2008; Daniels & Harris, 2005), as to separate artificially function from behaviour gives an incomplete picture (Skinner, Edge, Altman, et al., 2003). For example, simply assessing what coping functions people attempted does not given any information on whether these functions are more or less effective if enacted with specific behaviours, and may return a null relationship if some behaviours are effective and others not. Similarly, simply assessing which behaviours are enacted does not give any information on whether the behaviours are more or less effective for different functions, and may return a null relationship if behaviours (e.g., talking to colleagues as an example of enacting support, changing work schedules as an example of enacting job control) can be used for multiple functions (e.g., problem-solving or avoiding situations). A more precise strategy has been developed, in which assessments of coping tap what behaviours (job characteristics) were enacted and for what purpose simultaneously to link coping behaviours and functions (Daniels, Beesley, Cheyne, et al., 2008; Daniels, Boocock, Glover, et al., 2009; Daniels & Harris, 2005). Although latent job characteristics such as job control and support are thought to potentiate coping behaviours, finding a direct link might not be straightforward (Daniels, Harris, & Briner, 2004; Daniels, Boocock, Glover, et al., 2009). Typically, measures of job characteristics assess constructs like job control, support, and demands as generalized constructs. This level of analysis may be too imprecise: resources matched to specific demands (e.g., emotional, physical, cognitive) may be needed to enable coping behaviours that fulfil
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
24
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
specific functions to protect specific aspects of well-being (de Jonge & Dormann, 2006; de Jonge, Dormann, & van den Tooren, 2008; van den Tooren & de Jonge, 2010). Moreover, as coping can reflect both controlled and automatic processes, there might be different cognitive pathways that eventually lead to coping choice in response to any given event (Daniels, Harris, & Briner, 2004). Controlled processes involved in coping choice may involve detailed processing of environmental cues, the costs and benefits of specific coping attempts in a given situation, and inferences based on information held in long-term memory on the likely success of different coping options in relation to events of a similar kind encountered in the past. Automatic processes involved in coping choice may lead to certain work events cueing coping attempts that are associated in long-term memory with events of a similar nature to that encountered, irrespective of other features of the situation encountered. Despite these complexities, three forms of coping function may warrant further investigation because some research indicates these forms of coping may be effective, at least when combined with coping behaviours that relate to enactment of job characteristics such as job control and social support. These relate to problem-solving, recovery, and affective expression.
Problem-solving Problem-focused coping is one of the major classes of coping (Lazarus & Folkman, 1984). Supporting the idea that the latent job characteristics of job control and support might enhance problem-focused coping, a number of survey studies have demonstrated measures of problem-focused coping function is associated with better well-being in the presence of high job control or social support (de Rijk, Le Blanc, Schaufeli, et al., 1998; Ippolito, Adler, Thomas, et al., 2005; Shimazu, de Jonge, & Irimajiri, 2008; Shimazu, Shimazu, & Odara, 2005). Similarly, Elfering, Grebner, Semmer, et al. (2005) found that problem-focused coping function was associated with more adaptive outcomes when stressful work events were perceived to be controllable. While these studies suggest enacting control or support to solve problems is beneficial, they are not direct tests and do not rule out alternative interpretations. In a series of studies, coping function–behaviour combinations have been examined directly in relation to work demands using daily and hourly reports (Daniels, Beesley, Cheyne, et al., 2008; Daniels, Boocock, Glover, et al., 2009; Daniels & Harris, 2005). In these studies, the enactment of job control to solve problems was operationalized as changing aspects of work activities to solve problems and the enactment of social support to solve problems was operationalized as the extent to which to which participants discussed problems to solve problems. Findings indicate that problem-solving activity might be related to cognitive performance (Daniels, Beesley, Cheyne, et al., 2008) and relationships between problem-solving activity and well-being might be
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
25
mediated by cognitive variables such as goal progress (Daniels & Harris, 2005) and learning (Daniels, Boocock, Glover, et al., 2009). Recovery Recovery is process in which impaired well-being, psychological and physical, is restored following stressful events or episodes, and can encompass psychological detachment from work, relaxation, mastery oriented activities aimed at building internal resources (e.g., self-efficacy), and the exercise of control over activities (Sonnentag & Fritz, 2007). Most work on recovery has examined the impact of non-work experiences in successful recovery from work. Fritz and Sonnentag (2006) concluded that reflecting on the negative aspects of work during vacations may impair recovery, but reflecting on the positive aspects of work, relaxation, and engaging in challenging or developmental activities during vacations is associated with better well-being after vacations. Similarly, Fritz and Sonnentag (2005) found that weekend social activities and reflection on positive aspects of work over the weekend were associated with better well-being and aspects of performance after the weekend. Sonnentag and Jelden (2009) found that constraints experienced at work were associated with less time on sporting activities after work, that long hours worked were associated with low-effort recovery activities (e.g., watching television), and sporting activities were associated with better recovery experiences. In studies examining psychological detachment from work outside of working hours, studies have found: detachment was related to increased work engagement and proactive work behaviour (Sonnentag, 2003); detachment was related to less negative affect, more positive affect for people who are highly engaged with work and less fatigue, especially after days with high time pressure (Sonnentag & Bayer, 2005; Sonnentag, Mijza, Binnewies, et al., 2008); and detachment buffered some impacts of conflicts between work and non-work life on well-being (Moreno-Jim´enez, Mayo, Sanz-Vergel, et al., 2009). This research on recovery indicates the importance of out-of-work activities and experiences in regulating the impact of work on well-being and health. Yet, even within work, there might be opportunities for recovery activities. In one study, engaging in pleasant activities (e.g., socializing) during work breaks was associated with more pleasant affect and less unpleasant affect, but engaging in activities that required continued work (e.g., preparing future work) was associated with unpleasant affect (Trougakos, Beal, Green, et al., 2008). Trougakos, Beal, Green, et al., reasoned engaging in pleasant activities involves a greater level of voluntariness, which might suggest enacting job control to detach oneself psychologically from work is associated with better recovery experiences. For example, enacting job control may allow workers to choose when to schedule their activities and breaks to optimize recovery from demanding or unpleasant work situations (Karasek & Theorell, 1990; Trougakos & Hideg, 2009). In support of this assertion, job control was
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
26
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
associated with lower need for recovery in two separate samples (Sonnentag & Zijlstra, 2006). In another study, workers who enacted job control to achieve psychological detachment from work demands during one work day reported better affective well-being the following morning (Daniels & Harris, 2005). These results support the viability of enacting job control to allow recovery as influential in the maintenance of well-being at work, although contingencies may need to be explored (Trougakos & Hideg, 2009). Affective Expression One way of regulating unpleasant affect is through suppression. However, research indicates that the process of suppressing the experience and expression of unpleasant affect has a number of costs. First, there is the cognitive effort needed to suppress unpleasant affect, which can lead to information processing deficits and potentially poor performance in cognitively complex work (Beal, Weiss, Barros, et al., 2005; Gross, 2008). There are also social costs to suppressing or modification of experienced affect, in terms of inducing unˆ e, 2005b; Gross, favourable reactions in others (Butler & Gross, 2009; Cot´ 2008). Suppression can also lead to rebound effects, in which suppression of unpleasant affect is followed by re-emergence of unpleasant affect (Elfenbein, 2007). Unsurprisingly, suppression may have no or negative relationships with well-being (Nezlek & Kuppens, 2008). Affective expression is a sub-set of affect regulation strategies in which the function is to regulate the affective response to an event or episode by expressing that affect (Baker & Berenbaum, 2007; Gross, 2008; Gross & Thompson, 2007; Stanton & Franz, 1999). Several researchers have argued that affective expression allows adverse cognitive and physiological reactions to a stressor to diminish through catharsis and may also help individuals to understand their situation and their goals (Austenfeld & Stanton, 2004; John & Gross, 2007; Lepore, Ragan, & Jones, 2000). However, the links between affective expression and protected well-being are not clear and may be subject to several moderators (Austenfeld & Stanton, 2004). Research on public facing work that requires individuals to suppress expressions of felt affect, but express organizationally prescribed affects, indicates that the detrimental effects of suppressing felt affect are lower in jobs with high autonomy, both in terms of well-being and performance (Goldberg & Grandey, 2007; Grandey, Fisk, & Steiner, 2005; Johnson & Spector, 2007). These results can be interpreted as having latitude over when to express affect and over which affects to express can be adaptive. Similarly, there is evidence that job autonomy promotes the well-being and performance of those individuals who accept the affects they experience rather than suppress their experience of unpleasant affect (Bond & Bunce, 2003). This finding again might suggest the latitude to express affects is adaptive. However, research that has directly assessed the
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
27
enactment of job control in order to express affect has produced mixed results (Daniels, Beesley, Cheyne, et al., 2008; Daniels, Boocock, Glover, et al., 2009; Harris & Daniels, 2005). Some researchers have concluded that expressing affect in a receptive social context might be beneficial (Austenfeld & Stanton, 2004), bringing benefits such as empathic understanding of emotional difficulties (Clark & Finkel, 2004). However, the evidence from work settings is that expressing affect to others at work might lead to ill-being rather than protect well-being (Elfenbein, 2007). Some workplace studies have shown that social support can accentuate the harmful impact of stressors (Buunk & Hoorens, 1992; Kaufman & Beehr, 1986). Other workplace studies have found inverse associations between wellbeing and emotional support (Lowe & Bennett, 2003; Zellars & Perrew´e, 2001). Studies that have examined expressing affect to others at work directly also seem to indicate harmful effects in general, including ill-being and impaired cognitive performance (Daniels, Beesley, Cheyne, et al., 2008; Daniels, Boocock, Glover, et al., 2009). The reasons expressing affect to others at work might be harmful are related to norms of rationality in modern work organizations (Ashforth & Humphrey, 1995) and that it is not appropriate to express negative feelings about work (Moreno-Jim´enez, Mayo, Sanz-Vergel, et al., 2009). First, expressing affect to others might be viewed as inappropriate and indicative of incompetence, which might threaten an individual’s self-esteem (Daniels, Harris, & Briner, 2004). Secondly, expressions of affect may not be validated, as they violate norms of rationality, which may either dilute the effect of expression or render it harmful (Lepore, Ragan, & Jones, 2000). Thirdly, individuals may modulate their expression of affect to others in order to comply with norms of rationality. This modulation may deplete energy (Grandey, 2003; John & Gross, 2007), impede cognitive performance (Beal, Weiss, Barros, et al., 2005; Butler, Egloff, Wilhelm, et al., 2003), and subsequently reduce well-being. Rim´e (2007, 2009) has argued that the primary benefit of expressing affect to others is not to regulate affect or to recover from stressful events or episodes. Rather, the primary benefit is to strengthen social bonds through sharing affect and empathy. In Rim´e’s scheme, expressing affect to others may also have a cognitive impact, for instance, through stimulating individuals to reconsider and reorganize their goals. Indeed, Daniels and Harris (2005) found expressing affect to others at work led to less goal progress on the same day as expressing affect, but enhanced goal progress the day after. Others have indicated expressing affect to others, both individually and collectively, may serve other purposes such as: emphasizing the importance of certain goals and social identities within an organization or profession; communicating problems to others; and displaying symbolic resistance against more powerful groups in an organization (see e.g., Coupland, Brown, Humphreys, et al., 2008; Gibson & Callister, 2010).
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
28
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Summary The idea that workers attempt to regulate the impact of work on well-being and health is well established, and indicates workers’ agency in protecting and enhancing their well-being: Certain behaviours are used to fulfil certain functions that protect and promote well-being, and coping behaviours and functions need to be considered together to get a complete picture of how workers’ agency protects well-being in response to any given event. Whilst these behaviours can take place outside of work (e.g., Sonnentag & Jelden, 2009), they can also take place within work. Changing work schedules to solve problems or take breaks from demanding work are examples of enacting job control to fulfil specific coping functions. It is also important, then, to consider how latent job characteristics might potentiate the enactment of job characteristics, as well as the purposes for which the job characteristics are enacted. The enactment of job control and support for problem-solving might be effective for protecting and promoting well-being, but the effects may be mediated by cognitive processes, such as goal progress or learning. The enactment of job control to take breaks within the working day when workers feel they need them might also be effective to aid recovery. These are tentative conclusions, and more research is needed to build a more nuanced body of evidence. The expression of affect, either through enacting job control or talking to others at work, also requires more attention and closer specification of processes, moderators, and benefits, as links to well-being seem difficult to establish. However, given that affective suppression seems to be an ineffective or even harmful strategy, examining when and how affective expression can protect well-being seems important, especially given the affective demands of some jobs (cf. de Jonge & Dormann, 2006).
IMPLICATIONS FOR INTERVENTIONS AND THEIR IMPLEMENTATION Revisioning workers as active interpreters and shapers of their work environment does not entail victim-blaming and leaving individuals with no support in regulating the risks to their health and well-being. However, this revisioning does lead to a reconsideration of how best to protect worker well-being in ways that are consonant with the psychological processes underpinning the generation and regulation of affect at work. Worker interpretation and agency also means that workers will be active in shaping those interventions, and therefore their success. This leads onto considering how best to involve workers in the design and implementation of interventions. Interpretation and Agency as a Basis for Intervention Risk assessment is a necessary step in risk management. Many national surveillance schemes, including the UK’s Management Standards for Work-Related
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
29
Stress, use standardized instruments for assessing (perceived) adverse job characteristics using researcher or policy-defined categories of job characteristics (Cousins, MacKay, Clarke, et al., 2004; Dollard, Skinner, Tucker, et al., 2007). As indicated earlier, categorization of events as relevant to goal progress or affective experience precedes appraisal of events. Because there is no guarantee that categories generated through policy-related research correspond to the categories people use to classify work events, there is no guarantee that using standardized instruments provides an assessment of work that is meaningful for any group of workers. Instead, it might be better to base assessments on the categories and language used to describe those categories specific to any given worker or group of workers (Daniels, Harris & Briner, 2004). Qualitative techniques exist for representing idiosyncratic categories of work events and these can easily be adapted to include workers’ beliefs concerning effective interventions (Harris, Daniels, & Briner, 2002). Such qualitative techniques can be labour intensive. However, there might be sufficient convergence of categorization schemes in work groups and organizations to allow some degree of tailoring of instruments, so that categories and the language used to label those categories are specific to a work group, organization, or occupation (Cox, Griffiths, Barlow, et al., 2000; Daniels, Harris, & Briner, 2004). Such assessments might begin with qualitative methods and then develop tailored questionnaires specific to a work group, organization, or occupation (Cox, Griffiths, Barlow, et al., 2000). Moreover, rather than assessing the frequency with which categories of events occur, it might also be relevant to assess workers’ beliefs about those categories of events, and whether specific categories are perceived to have implications for goals and affective experience. The extent to which classes of events are perceived to adversely influence goal progress and affective experience might be given weight alongside frequency of occurrence in determining priorities for intervention (Harris & Daniels, 2005). Job redesign can also be directed toward providing workers with the infrastructure to regulate their own affective experience (Daniels, Harris, & Briner, 2004; Frese & Zapf, 1994). However, it is important to remember job characteristics can be enacted for several purposes, and providing workers with greater job autonomy or support does not necessarily mean they will use that autonomy or support to engage in problem-solving or recovery activities, for example (Daniels, Beesley, Cheyne, et al., 2008). Instead, job redesign aimed at allowing workers to regulate their own affective experience might need to be supplemented with additional training. For example, job redesign to encourage problem-solving might be usefully supplemented with problem-solving skills training, networking skills training to solicit good advice in problemsolving, establishing information sharing networks, training line managers in supporting worker problem-solving, and changing line managers’ performance criteria to emphasize a more supportive and mentoring role (Daniels, Beesley, Cheyne, et al., 2008; Daniels, Boocock, Glover, et al., 2009; Daniels & Harris, 2005; Harris & Daniels, 2005). Similarly, redesigning jobs to allow workers to
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
30
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
schedule breaks when best for them might be usefully supplemented with training in psychological detachment techniques to aid recovery or affect regulation techniques (cf. Bond & Bunce, 2000). Because goal progress is critical to well-being, interventions need not be limited to job redesign but can be targeted at enhancing goal progress or aligning individual and organizational goals so that both can be achieved (Daniels, Harris, & Briner, 2004). Well-designed performance-management systems that include developmental appraisals might be useful in this respect (Daniels & Harris, 2005; Daniels, Harris, & Briner, 2004). Such systems can allow individuals to identify their own goals over a coming period and help line managers identify ways in which the organization can support workers in achieving those goals (see e.g., Armstrong, 2001). Similarly, absence management and return to work programmes can incorporate procedures for identifying and supporting individuals’ personal goal attainment for those returning to work after illness or who are remaining in work despite illness. Workers Shaping Interventions The success of any intervention designed to protect and enhance well-being at work is dependent on the extent to which the intervention is implemented as planned, and many factors might serve to hinder the implementation of interventions (Carroll, Patterson, Wood, et al., 2007; Egan, Bambra, Petticrew, et al., 2009; Nielsen, Fredslund, Christensen, et al., 2006). The experience and interpretation of interventions can vary even within the same work group (Randall, Cox, & Griffiths, 2007; Randall, Griffiths, & Cox, 2005). Workers’ appraisals of interventions can also influence the extent to which they participate in interventions and the success of those interventions (Nielsen, Randall, & Albertsen, 2007; Randall, Nielsen, & Tvedt, 2009). Not surprisingly, like job characteristics, the picture emerges of workers actively interpreting and shaping interventions designed to protect and enhance their well-being. One way of ensuring effective implementation is to involve workers in decisions concerning interventions (ETUC, 2004; Leka & Cox, 2008). Indeed, worker involvement is encapsulated in the UK Health and Safety Executive’s Management Standards for Work-Related Stress (MacKay, Cousins, Kelly, et al., 2004). The Management Standards prescribe worker involvement in three ways. First, through indirect participation in decision making: the Management Standards recommend various stakeholders, including worker representatives, be involved in the steering committee to oversee the process or risk assessment and management. Secondly, employees become involved in a limited way through completing the indicator tool. Thirdly, through more direct participation in focus groups; this strategy is enacted after assessment of job characteristics has been completed, with worker involvement limited to deciding how to change those job characteristics identified as necessary for change by the indicator tool. However, as noted above in the discussion of
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
31
risk assessment, it might be more appropriate to involve workers in deciding much more directly on which job characteristics need to change as well as how to change them. That is, workers may be asked which job characteristics they wish to change through traditional job redesign interventions, and which job characteristics they feel able to shape and regulate directly through crafting their own jobs. Involvement of workers throughout the process of risk assessment and intervention requires a participatory and two-way process of communication on how best to regulate risks (National Research Council, 1989). Such communication itself can be an intervention. Knowledge of how people interpret their work environment and predispositions to use certain forms of coping can be used to develop tailored messages, indicating what it is workers can do to regulate their own well-being, or inform workers of the likely impact of work events on goal progress in order to mitigate against any concerns. Involvement of workers’ representatives in developing these messages may add to their credibility (Daniels, 1996).
IMPLICATIONS FOR METHODS Quantitative data collection methods and statistical analysis dominate research in the work stress area, most probably because the quantitative tradition typifies most psychological and epidemiological research. Within this tradition, differentiating latent, perceived, and enacted job characteristics entails using different quantitative methods for assessing these different facets (Daniels, 2006). Perceived job characteristics might be assessed using questionnaire methods, which require people to give generalized assessments of their own work (for self-perceptions) or others’ work (for perceptions of others’ work), in much the same way as many self-report measures of job characteristics already do. Latent job characteristics might be assessed through ratings and codes derived from examining organizational documents and processes (e.g., job descriptions, organizational charts). Perceived and enacted job characteristics might only be distal influences on affective experience, whilst more proximal influences are enacted job characteristics, appraisals of enacted job characteristics, and the coping functions that are the target of enacted job characteristics (Daniels, 2006; Daniels, Harris, & Briner, 2004). Here, experience sampling methods seem the most appropriate means of gathering quantitative data. This is because these methods can capture the dynamism and intra-individual change that characterizes the experience of work, affect, and the short-term impact of affect regulation strategies (e.g., Daniels, 2006; Ilies, Schwind, & Heller, 2007; Sonnentag, 2003; Weiss, Nicholas, & Daus, 1999). Because such methods can easily incorporate a range of variables, including lagged measures of dependent variables, experience sampling methods have high internal validity (Bolger, Davis, & Rafaeli, 2003).
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
32
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Experience sampling methods often require self-reports of events and behaviours, and can also assess the dynamic intra-psychic appraisals of those events and functions of coping behaviours. However, self-reports of events and behaviours may be accurate if self-reports concentrate on reports of specific events over a short, recent, and specified time (Frese & Zapf, 1988), such as over the previous few hours (e.g., Daniels, Boocock, Glover et al., 2009). Such a strategy may provide the most accurate means of assessing experience (Parkinson, Briner, Reynolds, et al., 1995; Stone, Schwartz, Neale, et al., 1998), especially if accompanied by electronic methods that are able to provide a reliable time stamp of when assessments were completed and to prevent ‘back filling’ of assessments some hours or days after the events were experienced (Tennen, Affleck, Coyne, et al., 2006). Despite their advantages, experience sampling methods have several limitations. Although they are capable of detecting short-run changes in affective well-being and other leading indicators of behaviours and experiences underpinning psychological and psychosomatic health, experience sampling methods are not easily adaptable to studying health outcomes characterized by longterm processes (Daniels, 2006). In these circumstances, experience sampling methods may be able to examine the micro-processes thought to underpin the development of certain occupational health outcomes (e.g., cross-referencing reports and appraisals of work events with data from portable heart rate monitors), so as to explain the results of longer term studies that examine links between latent and perceived job characteristics and disease outcomes (e.g., cardiovascular complaints). Experience sampling methods also have severe restrictions on the number of variables that can be assessed, given the demands placed on research participants to provide data over several days. These restrictions lend themselves to theory testing, where relationships are assessed between a small number of key dependent, independent, and important control variables. For more exploratory work, other real-time methods might be suitable: qualitative diaries and ethnographic methods are able to collect data as events happen, and also obtain participants’ interpretations, affective experiences, and coping responses to those events. Such methods can also provide richer interpretations of important phenomena related to stress and well-being, such as the development and maintenance of institutional patterns of interpretation (cf. Meyerson, 1994), the enactment and shaping of jobs (Nyberg, 2009), and patterns of coping (Korczynski, 2003). Perhaps one key limitation of experience sampling methods is that they often require specialist methodological expertise (e.g., concerning multilevel modelling), hardware (e.g., personal digital assistants), and software (e.g., statistical) which are not easily accessible to work organizations. The resource intensiveness of experience sampling methods also means hiring in such knowledge and equipment from established research organizations might incur prohibitive costs. Therefore, in many cases, the organizational beneficiaries of experience
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
33
sampling methods might only be those that are able and willing to participate in less applied research, usually undertaken by university based research teams funded through institutional, government, or charity grants. Accordingly, one key area is to develop practical methods and tools that can be used in work organizations to build upon the knowledge generated through experience sampling research on enacted job characteristics, appraisals, coping responses, and the affective experience of work. Methods that may show particular promise for further investigation and development as practical assessment tools include qualitative diaries (Clarkson & Hodgkinson, 2007) and event reconstruction interviews (Grube, Schroer, Hentzschel, et al., 2008).
CONCLUSIONS The dominant approach to policy and practice in the area of work stress and well-being has focused on jobs and job redesign, but has evolved to ignore how workers interpret their work and how they act to shape their work. By recognizing the active role workers have in protecting and enhancing their well-being, research focused on the dynamic nature of enacted job characteristics, workers’ appraisals, and coping responses addresses key limitations of the dominant approach. Rather than downplaying the importance of job redesign as an intervention strategy, recognizing worker interpretation and agency suggests a variety of ways in which workers can become involved in designing their own work in ways which best suit them and enables them better to cope with adverse aspects of work in ways that they think are best for them. Recognizing the importance of the worker in this process is likely to lead to better designed and better received job redesign interventions, which are more likely to be implemented and supported by other aspects of human resources management, such as training and performance appraisal systems. This chapter has been critical of the dominant approach, which has perhaps reached its practical zenith in the UK Health and Safety Executive’s Management Standards for Work-Related Stress. The Management Standards are evidence-based and reflect ‘the state of the science’ as it was in the late 1990s or early 2000s. At this time, experience sampling studies of work stress were rare. Experience sampling studies still are rare, although their numbers are increasing. The dominant approach to policy cannot be criticized for limitations in the evidence base, but can be praised for being evidence based. As the evidence base progresses and changes, then it is to be anticipated that policy, guidance, and practice should change accordingly. Moreover, and to conclude, another word of praise is in order for the Management Standards: the UK Health and Safety Executive has shown that it is possible to develop real and tangible public policy, guidance, and supporting tools in the area of work-related stress.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
34
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
ACKNOWLEDGEMENTS Parts of this work were supported by Engineering and Physical Sciences Research Council grants D04863X and EP/F02942X/1.
REFERENCES Andrews, F.M., & McKennell, A.C. (1980). Measures of self-reported well-being. Social Indicators Research, 25, 1–30. Armstrong, M. (2001). A Handbook of Human Resource Management Practice ( 8th edn). London: Kogan Page. Ashforth, B.E., & Humphrey, R.H. (1995). Emotion in the workplace: A reappraisal. Human Relations, 48, 97–125. Austenfeld, J.L., & Stanton, A.L. (2004). Coping through emotional approach: A new look at emotion, coping, and health-related outcomes. Journal of Personality, 72, 1335–63. Baker, J.P., & Berenbaum. H. (2007). Emotional approach and problem-focused coping: A comparison of potentially adaptive coping strategies. Cognition and Emotion, 21, 95–118. Bakker, A., Schaufeli, W.B., Leiter, M.P., & Taris, T. (2008). Work engagement: An emerging concept in occupational health psychology. Work and Stress, 22, 187– 200. Bambra, C., Egan, M., Thomas, S., Petticrew, M., & Whitehead, M. (2007). The psychosocial and health effects of workplace reorganisation 2. A systematic review of task structuring interventions. Journal of Epidemiology and Community Health, 61, 1028–37. Basch, J., & Fisher, C.D. (2000). Affective events–emotions matrix: A classification of work events and associated emotions. In N.M. Ashkanasy, C.E.J. H¨artel, & W.J. Zerbe (Eds), Emotions in the Workplace: Research, Theory and Practice. Westport: Quorum Books. Beal, D.J., Weiss, H.M., Barros, E., & MacDermid, S.M. (2005). An episodic process model of affective influences on performance. Journal of Applied Psychology, 90, 1054–68. Berg, J.M., Wrzesniewski, A., & Dutton. J.E. (2010). Perceiving and responding to challenges in job crafting: When proactivity requires adaptivity. Journal of Organizational Behavior, 31, 158–186. Bolger, N., Davis, A., & Rafaeli, E. (2003). Diary methods: Capturing life as it is lived. Annual Review of Psychology, 54, 579–616. Bond, F.W., & Bunce, D. (2003). The role of acceptance and job control in mental health, job satisfaction, and work performance. Journal of Applied Psychology, 88, 1057–67. Bond, F.W., & Bunce, D. (2000). Mediators of change in emotion-focused and problem-focused worksite stress management interventions. Journal of Occupational Health Psychology, 5, 156–63. Bosma, H., Marmot, M.G., Hemingway, H., Nicholson, A.C., Brunner, E., & Stansfeld, S.A. (1997). Low job control and risk of coronary heart disease in Whitehall II (prospective cohort) study. British Medical Journal, 314, 558–65. Brief, A., & George, J.M. (1991). Psychological stress and the workplace: A brief comment on Lazarus’ outlook. Journal of Social Behavior and Personality, 6, 15–20.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
35
Brief, A.P., Burke, M.J., George, J.M., Robinson, B.S. & Webster, J. (1988). Should negative affectivity remain an unmeasured variable in the study of job stress. Journal of Applied Psychology, 73, 193–8. Briner, R.B., Harris, C., & Daniels, K. (2004). How do stress and coping work? Toward a fundamental theoretical reappraisal. British Journal of Guidance and Counselling, 32, 223–34. Butler, E.A., & Gross, J.J. (2009). Emotion and emotion regulation: Integrating individual and social levels of analysis. Emotion Review, 1, 86–7. Butler, E.A., Egloff, B., Wilhelm, F.H., Smith. N.C., Erickson, E.A., & Gross, J.J. (2003). The social consequences of expressive suppression. Emotion, 3, 48–67. Buunk, B.P., & Hoorens, V. (1992). Social support and stress: The role of social comparison and social exchange processes. British Journal of Clinical Psychology, 31, 445–57. Carroll, C., Patterson, M., Wood, S., Booth, A., Rick, J., & Balain, S. (2007). A conceptual framework for implementation fidelity. Implementation Science, 2, 40. Carver, C.S., & Scheier, M.F. (1990). Origins and functions of positive and negative affect: A control-process view. Psychological Review, 97, 19–35. Cass, M., Faragher, B., & Cooper, C.L. (2002). Health and Employment: A Review and Meta-analysis. Sudbury: HSE Books. Christie, A.M., & Barling, J. (2009). Disentangling the indirect links between socioeconomic status and health: The dynamic roles of work stressors and personal control. Journal of Applied Psychology, 94, 1466–78. Clark, M.S., & Finkel, E.J. (2004). Does expressing emotion promote well-being? It depends on relationship context. In L.Z. Tiedens & C.W. Leach (Eds), The Social Life of Emotions. Cambridge, UK: Cambridge University Press. Clarkson, G.P., & Hodgkinson, G.P. (2007). What can occupational stress diaries achieve that questionnaires can’t. Personnel Review, 36, 684–700. Clegg, C.W., & Spencer. C. (2007). A circular and dynamic model of the process of job design. Journal of Occupational and Organizational Psychology, 80, 321–39. Clore, G.L., Wyer, R.S., Dienes, B., Gasper, K., Gohm, C., & Isbell, L. (2001). Affective feelings as feedback: Some cognitive consequences. In L.L. Martin & G.L. Clore (Eds), Theories of Mood and Cognition: A User’s Handbook. Mahwah, NJ: Erlbaum. Cooper, C.L., & Marshall, J. (1976). Occupational sources of stress: A review of the literature relating to coronary heart disease and mental ill health. Journal of Occupational Psychology, 49, 11–28. Cooper, C.L., Dewe, P.J., & O’Driscoll, M.P. (2001). Organizational Stress: A Review and Critique of Theory, Research, and Applications. Thousand Oaks, CA: Sage. ˆ e, S. (2005a). Reconciling the feelings-as-information and hedonic contingency Cot´ models of how mood influences systematic information processing. Journal of Applied Social Psychology, 35, 1656–79. ˆ e, S. (2005b). A social interaction model of the effects of emotion regulation on Cot´ work strain. Academy of Management Review, 30, 509–30. Coupland, C., Brown, A.D, Daniels, K., & Humphreys, M. (2008). Saying it with feeling: analysing speakable emotions. Human Relations, 61, 327–53. Cousins, R., MacKay, C.J., Clarke, S.D., Kelly, C., Kelly, P.J., & McCaig, R. (2004). ‘Management Standards’ work-related stress in the UK: Practical development. Work and Stress, 18, 113–36. Cox, T. (1978). Stress. London: MacMillan. Cox, T., Griffiths, A., Barlow, C.A., Randall, R., Thomson, L, & Rial-Gonzalez, E. (2000). Organisational Interventions for Work Stress: A Risk Management Approach. Sudbury, UK: HSE Books.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
36
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Cox, T., Karanika-Murray, M., Griffiths, A., Wong, Y.Y.V., & Hardy, C. (2009). Developing the Management Standards Approach Within the Context of Common Health Problems in the Workplace: A Delphi Study. Norwich, UK: HSE Books. Cramer, P. (2000). Defense mechanisms in psychology today: Further process for adaptation. American Psychologist, 55, 637–46. Cropanzano, R., Weiss, H.M., Hale, J.M.S., & Reb, J. (2003). The structure of affect: Reconsidering the relationship between negative and positive affectivity. Journal of Management, 29, 831–57. Daniels, K. (1996). Why aren’t managers concerned about occupational stress ? Work and Stress, 10, 352–66. Daniels, K. (2000). Measures of five aspects of affective well-being at work. Human Relations, 53, 275–94. Daniels, K. (2006). Rethinking job characteristics in work stress research. Human Relations, 59, 267–90. Daniels, K., Beesley, N.J., Cheyne, A.J.T., & Wilmarisiri, V.P. (2008). Coping processes linking the Demands-Control-Support model, affect, and risky decisions at work. Human Relations, 61, 845–74. Daniels, K., Boocock, G., Glover, J., Hartley, R., & Holland, J. (2009). An experience sampling study of learning, affect, and the Demands Control Support model. Journal of Applied Psychology, 94, 1003–17. Daniels, K., & Guppy A. (1997). Stressors, locus of control and social support as consequences of psychological well-being. Journal of Occupational Health Psychology, 2, 156–74. Daniels, K., & Harris, C. (2005). A daily diary study of coping in the context of the job demands-control-support model. Journal of Vocational Behavior, 66, 219–37. Daniels, K., Harris, C., & Briner, R.B. (2004). Linking work conditions to unpleasant affect: Cognition, categorisation and goals. Journal of Occupational and Organizational Psychology, 77, 343–64. Daniels, K., Hartley, R., & Travers, C. (2006). Beliefs about stressors alter stressors’ impact: Evidence from two experience-sampling studies. Human Relations, 59, 1261–85. Daniels, K., Jones, D., Perryman, S., Rick, J., & Fergusson, E. (2004). Cognitive and Perceptual Factors’ Influence on the Expression and Reporting of Work-Related Stress. Sudbury, UK: HSE Books. Daniels, K., & de Jonge, J. (2010). Matching-making and match-breaking: The nature of match within and around job design. Journal of Occupational and Organizational Psychology, 83, 1–16. Danna, K., & Griffin, R.W. (1999). Health and well-being in the workplace: A review of the literature and synthesis of the literature. Journal of Management, 25, 357–84. de Jonge, J., & Dormann, C. (2006). Stressors, resources, and strain at work: A longitudinal test of the triple-match principle. Journal of Applied Psychology, 91, 1359–74. de Jonge, J., Dormann, C., & Tooren van den, M. (2008). The Demand-Induced Strain Compensation Model: Renewed theoretical considerations and empirical evidence. In K. N¨aswall, J. Hellgren, & M. Sverke (Eds), The Individual in the Changing Working Life. Oxford, UK: Cambridge University Press. de Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.R.D., & Bongers, P.M. (2003). ‘The very best of the millennium’: Longitudinal research and the demandcontrol-(support) model. Journal of Occupational Health Psychology, 8, 282–305. Demerouti, E., Bakker, A.B., Nachreiner, F., & Schaufeli, W.B. (2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86, 499–512. de Rijk, A.E., Le Blanc, P., Schaufeli, W., & de Jonge, J. (1998). Active coping and need for control as moderators of the job-demand-control model: Effects on burnout. Journal of Occupational and Organizational Psychology, 71, 1–18.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
37
Dewe, P., & Cooper, C.L. (2007). Coping research and measurement in the context of work related stress. In G.P. Hodgkinson & J.K. Ford (Eds), International Review of Industrial and Organizational Psychology (volume 22). Chichester, UK: Wiley. Dick, P. (2000). The social construction of the meaning of acute stressors: A qualitative study of the personal accounts of police officers using a stress counselling service. Work and Stress, 14, 226–44. Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542–75. Diener, E., & Larsen, R.J. (1993). The experience of emotional well-being. In M. Lewis & J.M. Havilland (Eds), Handbook of Emotions. New York: Guildford Press. Diener, E., Suh, E.M., Lucas, R.E., & Smith, H.L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125, 276–302. DiMaggio, P.J., & Powell, W.W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48, 147–61. Dollard, M., Skinner, N., Tuckey, M.R., & Bailey, T. (2007). National surveillance of psychosocial risk factors in the workplace: An international overview. Work and Stress, 21, 1–29. Egan, M., Bambra, C., Petticrew, M., & Whitehead, M. (2009). Reviewing evidence on complex social interventions: Appraising implementation in systematic reviews of the health effects of organizational-level workplace interventions. Journal of Epidemiology and Community Health, 63, 4–11. Egan, M., Bambra, C., Petticrew, M., Whitehead, M., Thomas, S., & Thompson, H. (2007). The psychosocial and health effects of workplace reorganisation 1: A systematic review of interventions that aim to increase employee participation or control. Journal of Epidemiology and Community Health, 61, 945–54. Elfenbein, H.A. (2007). Emotion in organizations: A review and theoretical integration. Academy of Management Annals, 1, 315–86 Elfering, A., Grebner, S., Semmer, N.K., Kaiser-Freiburghaus, D., Lauper-Del Ponte, S., & Witschi, I. (2005). Chronic job stressors and job control: Effects of eventrelated coping success and well-being. Journal of Occupational and Organizational Psychology, 78, 237–52. ETUC (2004). Framework Agreement on Work-Related Stress. Brussels: European Trade Union Confederation. Ferguson, E., Daniels, K., & Jones, D. (2006). The relative contributions of work conditions and psychological differences to health measures: A meta-analysis with structural equations modelling. Journal of Psychosomatic Research, 60, 45–52. Fiedler, K., & Bless, H. (2000). The formation of beliefs at the interface of affective and cognitive processes. In N.H. Frijda, A.S.R. Manstead, & S. Bem (Eds), Emotions and Beliefs: How Feelings Influence Thoughts. Paris: Cambridge University Press. Fisher, C.D. (2000). Mood and emotions while working: Missing pieces of job satisfaction. Journal of Organizational Behavior, 21, 185–202. Fisher, C.D. (2002). Antecedents and consequences of real-time affective reactions to work. Motivation and Emotion, 26, 3–30. Forgas, J.P. (2001). The affect infusion model (AIM); an integrative theory of mood effects on cognition and judgments. In L.L. Martin & G.L. Clore (Eds), Theories of Mood and Cognition: A User’s Handbook. Mahwah, NJ: Erlbaum. Frese, M. (1985). Stress at work and psychosomatic complaints: A causal interpretation. Journal of Applied Psychology, 70, 314–28. Frese, M., & Zapf, D. (1988). Methodological issues in the study of work stress: Objective vs subjective measurement work stress and the question of longitudinal studies. In C.L. Cooper & R. Payne (Eds), Causes, Coping, and Consequences of Stress at Work. Chichester: Wiley.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
38
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Frese, M., & Zapf, D. (1994). Action as the core of work psychology: A German approach. In H.C. Triandis, M.D. Dunnette & L.M. Hough (Eds), Handbook of Industrial and Organizational Psychology ( 2nd edn). Palo Alto, CA: Consulting Psychologists Press. Frijda, N.H. (1986). The Emotions. Cambridge: Cambridge University Press. Fritz, C., & Sonnentag, S. (2005). Recovery, health, and job performance. Effects of weekend experiences. Journal of Occupational Health Psychology, 10, 187–99. Fritz, C., & Sonnentag, S. (2006). Recovery, well-being, and performance-related outcomes: The role of workload and vacation experiences. Journal of Applied Psychology, 91, 936–45. ¨ Fuller, J.A., Stanton, J.M., Fisher, G.G., Spitzmuller, C., Russell, S.S., & Smith, P.C. (2003). A lengthy look at the daily grind: Time series analysis of events, mood, stress, and, satisfaction. Journal of Applied Psychology, 88, 1019–33. George, J.M. (2009). The illusion of will in organizational behavior research: Nonconscious processes and job design. Journal of Management, 35, 1318–39. Gibson, D.E., & Callister, R.R. (2010). Anger in organizations: Review and integration. Journal of Management, 36, 66–93. Goldberg, L.S., & Grandey, A.A. (2007). Display rules versus display autonomy: Emotion regulation, emotional exhaustion, and task performance in a call center simulation. Journal of Occupational Health Psychology, 12, 301–18. Grandey, A.A. (2003). When ‘the show must go on’: Surface acting and deep acting as determinants of emotional exhaustion and peer-rated service delivery. Academy of Management Journal, 46, 86–96. Grandey, A.A., Fisk, G.M., & Steiner, D.D. (2005). Must ‘service with a smile’ be stressful? The moderating role of personal control for American and French employees. Journal of Applied Psychology, 90, 893–904. Grant, A.M., Fried, Y., Parker, S.K., & Frese, M., (2010). Putting job design in context: Introduction to the special issue. Journal of Organizational Behavior, 31, 145–57. Grant, A.M., & Parker, S.K. (2009). Redesigning work design theories: The rise of relational and proactive perspectives. Academy of Management Annals, 3, 317–75. Gross, J.J. (2008). Emotion regulation. In M. Lewis, J.M. Haviland-Jones & L. Feldman-Barrett (Eds), Handbook of Emotions ( 3rd edn). New York: Guildford. Gross, J.J., & Thompson, R.A. (2007). Emotional regulation: Conceptual foundations. In J.J. Gross (Ed.), Handbook of Emotional Regulation. New York: Guilford. Grube, A., Schroer, J., Hentzchel, C., & Hertel, G. (2008). The event reconstruction method: An efficient measure of experience based job satisfaction. Journal of Occupational and Organizational Psychology, 81, 669–89. Halberstadt, J.B., & Niedenthal, P.M. (1997). Emotional state and the use of stimulus dimensions in judgment. Journal of Personality and Social Psychology, 72, 1017–33. Harkness, A.M.B., Long, B.C., Bermbach, N., Patterson, K., Jordan, S. & Kahn, H. (2005). Talking about stress: Discourse analysis and implications for stress interventions. Work and Stress, 19, 121–36. Harris, C. (2002). Mental models of psychosocial risk perception and well-being. PhD thesis. University of Nottingham. Harris, C., & Daniels, K. (2005). Daily affect and daily beliefs about work demands. Journal of Occupational Health Psychology, 10, 415–28. Harris, C., & Daniels, K. (2007). The role of stressor appraisals in psychological wellbeing and physical symptom reporting. European Journal of Work and Organizational Psychology, 16, 407–31. Harris, C., Daniels, K., & Briner, R. (2002). Using cognitive mapping for psychosocial risk assessment. Risk Management: An International Journal, 4, 7–22.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
39
Harris, C., Daniels, K., & Briner, R. (2003). A daily diary study of goals and affective well-being at work. Journal of Occupational and Organizational Psychology, 76, 401–10. Hart, P.M., & Cooper, C.L. (2001). Occupational stress: Toward a more integrated framework. In N.A. Anderson, D.S. Ones, H.K. Sinangil, & C. Viswesveran (Eds), Handbook of Industrial, Work, and Organizational Psychology (volume 2). London: Sage. Haslam, N. (1995). The discreteness of emotion concepts: Categorical structure in the affective circumplex. Personality and Social Psychology Bulletin, 21, 1012–9. Hepburn, A., & Brown, S. (2001). Teacher stress and the management of accountability. Human Relations, 54, 691–716. Holland J.L. (1985). Making Vocational Choices. Englewood Cliffs, NJ: Prentice Hall. Hodgkinson, G.P., & Healey, M.P. (2008). Cognition in organizations. Annual Review of Psychology, 59, 387–417. Hurrell, J.J., Nelson, D.L., & Simmons, B.L. (1998). Measuring job stressors and strains: Where have we been, where are we: And where do we need to go. Journal of Occupational Health Psychology, 3, 368–89. Ilies, R., Schwind, K.M., & Heller, D. (2007). Employee well-being: A multilevel model linking work and nonwork domains. European Journal of Work and Organizational Psychology, 16, 326–41. ILO (2001). Guidelines on Occupational Safety and Health Management Systems. Geneva: International Labor Office. ¨ Ippolito, J., Adler, A.B., Thomas, J.L., Litz, B.T., & Holzl, R. (2005). Extending and applying the demand control model: The role of soldier’s coping on a peacekeeping deployment. Journal of Occupational Health Psychology, 10, 452–64. Jackson, S.E. (1983). Participation in decision making as a strategy for reducing jobrelated strain. Journal of Applied Psychology, 68, 3–19. Jackson, S.E., & Schuler, R.S. (1985). A meta-analysis and conceptual critique of research on role ambiguity and role conflict in work settings. Organizational Behavior and Human Decision Processes, 36, 16–78. Jex, S.M. (1998). Stress and Job Performance: Theory, Research, and Implications for Management Practice. Thousand Oaks, CA: Sage. John, O.P, & Gross, J.J. (2007). Individual differences in emotional regulation. In J.J. Gross (Ed.). Handbook of Emotional Regulation. New York: Guilford. Johnson, H.M., & Spector, P.E. (2007). Service with a smile: Do emotional intelligence, gender, and autonomy moderate the emotional labor process. Journal of Occupational Health Psychology, 12, 319–34. Karanika-Murray, M., Antoniou, A.S., Michaelides, G., & Cox, T. (2009). Expanding the risk assessment methodology for work-related health: A technique for incorporating multivariate curvilinear effects. Work and Stress, 23, 99–119. Karasek, R.A., & Theorell, T. (1990). Healthy Work. New York: Basic Books. Kasl, S.V. (1983). Pursuing the link between stressful life experiences and disease: A time for re-appraisal. In C.L. Cooper (Ed.), Stress Research: Issues for the Eighties. Chichester, UK: Wiley. Kaufman, G.M., & Beehr, T.A. (1986). Interactions between job stressors and social support: Some counterintuitive results. Journal of Applied Psychology, 71, 522–6. Kehr, H.M. (2003). Goal conflicts, attainment of new goals, and well-being among managers. Journal of Occupational Health Psychology, 8, 195–208. Kompier, M. (2004). Does the ‘Management Standards’ approach meet the standard? Work and Stress, 18, 137–9. Korczynski, M. (2003). Communities of coping: Collective emotional labour in service work. Organization, 10, 55–79.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
40
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
LaMontagne, A.D., Keegel, T., Louie, A.M., Ostry, A., & Landsbergis, P.A. (2007). A systematic review of the job-stress intervention evaluation literature 1990–2005. International Journal of Occupational and Environmental Medicine, 13, 268–80. Lassiter, G.D., Koenig, L.J., & Apple, K.J. (1996). Mood and behavior perception: Dysphoria can increase and decrease effortful processing of information. Personality and Social Psychology Bulletin, 22, 794–810. Latack, J.C. (1986). Coping with job stress: Measures and future directions for scale development. Journal of Applied Psychology, 71, 377–85. Latack, J.C., & Havlovic, S.J. (1992). Coping with job stress: A conceptual evaluation framework for coping measures. Journal of Organizational Behavior, 13, 479–508. Latack, J.C., Kinicki, A.J., & Prussia, G.E. (1995). An integrative process model of coping with job loss. Academy of Management Review, 20, 311–42. Lazarus, R.S. (1999). Stress and Emotion: A New Synthesis. New York: Springer. Lazarus, R.S., & Folkman, S. (1984). Stress, Appraisal and Coping. New York: Springer. Leka, S., & Cox, T. (2008). PRIMA-EF: Guidance on the European Framework for Psychosocial Risk Management: A Resource for Employers and Worker Representatives. Geneva: World Health Organization. Leka, S., Griffiths, A., & Cox, T. (2003). Work Organization and Stress: Systematic Problem Approaches for Employers, Managers and Trade Union Representatives. Geneva: World Health Organization. Lepore, S.J., Ragan, J.D., & Jones, S. (2000). Talking facilitates cognitive-emotional processes of adaptation to an acute stressor. Journal of Personality and Social Psychology, 78, 499–508. Lerner, D., Amick, B. C., Malspeis S., & Rogers, W. H. (2000). A national survey of health-related work limitations among employed persons in the United States. Disability and Rehabilitation, 22, 225–32. Lindsay, J., & Powell, G.E. (Eds). (1994). Handbook of Clinical Adult Psychology (2nd edn). London: Routledge. Lowe, R., & Bennett, P. (2003). Exploring coping reactions to work-stress: Application of an appraisal theory. Journal of Occupational and Organizational Psychology, 76, 393–400. Lunt, J., Fox, D., Bowen, J., Higgins, G., Crozier, S., & Carter, L. (2007). Applying the Biopsychosocial Approach to Managing Risks of Contemporary Occupational Health Conditions: Scoping Review. Health and Safety Laboratory: Buxton. Macik-Frey, M., Quick, J.C., & Nelson, D.L. (2007). Advances in occupational health: From a stressful beginning to a positive future. Journal of Management, 33, 809–40. MacKay, C.J., Cousins, R., Kelly, P.J., Lee, S., & McCaig, R.H. (2004). ‘Management Standards’ and work-related stress in the UK: Policy background and science. Work and Stress, 18, 91–112. Maier, G.W., & Brunstein, J.C. (2001). The role of personal goals in newcomers job satisfaction and organizational commitment: A longitudinal analysis. Journal of Applied Psychology, 86, 1034–42. Manning, M.R, Jackson, C.N., & Fusilier, M.R. (1996). Occupational stress, social support, and the costs of health care. Academy of Management Journal, 39, 738–50. Mason, J.W. (1975). A historical view of the stress field (parts 1 and 2). Journal of Human Stress, 1, 6–12, 22–36. Meyerson, D.E. (1994). Interpretations of stress in institutions: The cultural production of ambiguity and burnout. Administrative Science Quarterly, 39, 628–53. Morgeson, F.P., & Humphrey, S.E. (2006). The work design questionnaire (WDQ): Developing and validating a comprehensive measure for assessing job design and the nature of work. Journal of Applied Psychology, 91, 1321–39.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
41
˜ Moreno-Jim´enez, B., Mayo, M., Sanz-Vergel, A.I., Geurts, S., Rodr´ıguez-Munoz, A., & Garrosa, E. (2009). Effects of work-family conflict on employees’ well-being: The moderating role of recovery strategies. Journal of Occupational Health Psychology, 14, 427–10. Munir, F., Jones, D., Leka, S., & Griffiths, A. (2005). Work limitations and employer adjustments for employees with chronic illness. International Journal of Rehabilitation Research, 28, 111–7. Munir, F., Leka, S., & Griffiths, A. (2005). Dealing with self-management of chronic illness at work: Predictors for self-disclosure. Social Science and Medicine, 60, 1397–407. Munir, F., Yarker J., & Haslam C. (2008). Sickness absence management: Encouraging attendance or risk-taking presenteeism in employees with chronic illness? Disability and Rehabilitation, 30, 1461–72. Munir, F., Yarker, J., Haslam C., Long, H., Leka, S., Griffiths, A., & Cox, S. (2007). Work factors related to psychological and health-related distress among employees with chronic illnesses. Journal of Occupational Rehabilitation, 17, 259–77. Murphy, L.R. (1988). Workplace interventions for stress reduction and prevention. In C.L. Cooper & R. Payne (Eds), Causes, Coping and Consequences of Stress at Work. Chichester, UK: Wiley. National Research Council (1989). Improving Risk Communication. Washington DC: National Academy Press. Nezlek, J.B., & Kuppens, P. (2008). Regulating positive and negative emotions in daily life. Journal of Personality, 76, 561–79. ˚ .H. (1999). Emotional response Niedenthal, P.M., Halberstadt, J.B., & Innes-Ker, A categorization. Psychological Review, 106, 337–61. Nielsen, K., Fredslund, H., Christensen, K.B., & Albertsen, K. (2006). Success or failure? Interpreting and understanding the impact of interventions. Work and Stress, 20, 272–87. Nielsen, K., Randall, R., & Albertsen, K. (2007). Participants’ appraisals of process issues and the effects of stress management interventions. Journal of Organizational Behavior, 28, 793–810. Noon, M., & Blyton, P. (1997). The Realities of Work. Basingstoke, UK: MacMillan. Nyberg, D. (2009). Computers, customer service operatives and cyborgs: Intra-actions in call centers. Organization Studies, 30, 1181–99. Oatley, K., & Johnson-Laird, P. (1987). Towards a cognitive theory of emotions. Cognition and Emotion, 1, 29–50. Parent-Thirion, A., Fern´andez Mac´ıas, E., Hurley, J., & Vermeylen, G. (2007). Fourth European Working Conditions Survey. Luxembourg: Office for Official Publications of the European Communities. Parker, S.K. & Ohly, S. (2008). Designing motivating jobs: An expanded framework for linking work characteristics and motivation. In R. Kanfer, G. Chen & R. Pritchard (Eds), Work Motivation: Past, Present, and Future (pp. 233–84). New York: Routledge. Parkinson, B., Briner, R.B., Reynolds, S., & Totterdell, P. (1995). Time frames for mood: Relations between momentary and generalized ratings of affect. Personality and Social Psychology Bulletin, 21, 331–9. Peterson, M.F. (1998). Embedded organizational events: The units of process in organization science. Organization Science, 9, 16–33. Pierce, J.L., Jussila, I., & Cummings, A. (2009). Psychological ownership within the job design context: Revision of the job characteristics model. Journal of Organizational Behavior, 30, 477–96. Pillow, D.R., Zautra, A.J., & Sandler, I. (1996). Major life events and minor stressors: Identifying mediational links in the stress process. Journal of Personality and Social Psychology, 70, 381–94.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
42
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Pomaki, G., Karoly, P., & Maes, S. (2009). Linking goal progress to subjective wellbeing at work: The moderating role of goal-related self-efficacy and attainability. Journal of Occupational Health Psychology, 14, 206–18. Power, M., & Dalgleish, T. (2008). Cognition and Emotion: From Order to Disorder (2nd edn). Hove, UK: Psychology Press. Quick, J.C., Cooper, C.L., Gibbs. P.C., Little, L.M., & Nelson, D.L. (2010). Positive organizational behaviour. In G.P. Hodgkinson & J.K. Fords (Eds), International Review of Industrial and Organizational Psychology (volume 25). Chichester, UK: Wiley. Randall, R., Cox, T., & Griffiths, A. (2007). Participants’ accounts of a stress management intervention. Human Relations, 60, 1181–209. Randall, R., Griffiths, A., & Cox, T. (2005). Evaluating organizational stressmanagement interventions using adapted study designs. European Journal of Work and Organizational Psychology, 14, 23–41. Randall, R., Nielsen, K., & Tvedt, S.D. (2009). The development of five scales to measure employees’ appraisals of organizational-level stress management interventions. Work and Stress, 23, 1–23. Richardson, K.M., & Rothstein, H.R. (2008). Effects of occupational stress management intervention programs: A meta-analysis. Journal of Occupational Health Psychology, 13, 69–93. Rick, J., & Briner, R.B. (2000). Psychosocial risk assessment: Problems and prospects. Occupational Medicine, 50, 310–4. Rick, J., Thomson, L., Briner, R., O’Regan, S., & Daniels, K. (2002). Review of Existing Scientific Knowledge to Underpin Standards of Good Practice for Key Work-related Stressors – Phase 1. Sudbury, UK: HSE Books. Rifkin, A. (1985). Evidence for a basic level in event taxonomies. Memory and Cognition, 13, 538–56. Rim´e, B. (2007). Interpersonal emotion regulation. In J.J. Gross (Ed.), Handbook of Emotional Regulation. New York: Guilford. Rim´e, B. (2009). Emotions elicits the social sharing of emotion: Theory and empirical review. Emotion Review, 1, 60–85. Roth, S., & Cohen, L.J. (1986). Approach, avoidance, and coping with stress. American Psychologist, 41, 813–9. Ruder, M., & Bless, H. (2003). Mood and the reliance on the ease of retrieval heuristic. Journal of Personality and Social Psychology, 85, 20–32. Ryff, C.D., & Keyes, C.L.M. (1995). The structure of psychological well-being revisited. Journal of Personality and Social Psychology, 69, 719–27. Salancik, G.R., & Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 23, 224–53. Schaubroeck, J. (1999). Should the subjective be the objective? On studying mental processes, coping behavior, and actual exposures in organizational stress research. Journal of Organizational Behavior, 20, 753–60. Schneider, W., & Chein, J.M. (2003). Controlled and automatic processing: Behavior, theory, and biological mechanisms. Cognitive Science, 27, 525–59. Schneider, W., & Shiffrin, R.M. (1977). Controlled and automatic human information processing: I. Detection, search, and attention. Psychological Review, 84, 1–66. Schwartz, J.E., Pickering, T.G., & Landsergis, P.A. (1996). Work-related stress and blood pressure: Current theoretical models and considerations from a behavioral medicine perspective. Journal of Occupational Health Psychology, 1, 287– 310. Schwarz, N. (2000). Emotion, cognition, and decision making. Cognition and Emotion, 14, 433–40. Scott, W.R. (1995). Institutions and Organizations. Thousand Oaks, CA: Sage.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
43
Shimazu, A., de Jonge, J., & Irimajiri, H. (2008). Lagged effects of active coping within the demand-control model: A three-wave panel study among Japanese employees. International Journal of Behavioral Medicine, 15, 44–53. Shimazu, A., Shimazu, M., & Odara, T. (2005). Divergent effects of active coping on psychological distress in the context of the Job Demands-Control-Support Model: The roles of job control and social support. International Journal of Behavioral Medicine, 12, 192–8. Skinner, E.A., Edge, K., Altman, J., & Sherwood, H. (2003). Searching for the structure of coping: A review and critique of category systems for classifying ways of coping. Psychological Bulletin, 129, 216–69. Smith, C.A., & Kirby, L.D. (2001). Affect and cognitive appraisal processes. In J.P. Forgas (Ed.), Handbook of Affect and Social Cognition. New Jersey: Erlbaum. Smith E.E., & Medin, D.L. (1981). Categories and Concepts. Cambridge, MA: Harvard University Press. Sonnentag, S. (2003). Recovery, work engagement, and proactive behavior: A new look at the interface between work and nonwork. Journal of Applied Psychology, 88, 518–28. Sonnentag, S., & Bayer, U. (2005). Switching off mentally: Predictors and consequences of psychological detachment from work during off-job time. Journal of Occupational Health Psychology, 10, 393–44. Sonnentag, S., & Fritz, C. (2007). The recovery experience questionnaire: Development and validation of a measure for assessing recuperation and unwinding from work. Journal of Occupational Health Psychology, 12, 204–21. Sonnentag, S., & Jelden, S. (2009). Job stressors and the pursuit of sport activities: A day-level perspective. Journal of Occupational Health Psychology, 14, 165–81. Sonnentag, S., Mijza E,J., Binnewies, C., & Scholl, A. (2008). Being engaged at work and detached at home: A week-level study of work engagement, psychological detachment, and affect. Work and Stress, 22, 257–76. Sonnentag, S., & Zijlstra, F.R.H. (2006). Job characteristics and off-job activities as predictors of need for recovery, well-being, and fatigue. Journal of Applied Psychology, 91, 330–50. Spector, P.E. (1992). A consideration of the validity and meaning of self-report measures of job conditions. In C.L. Cooper & I.T. Robertson (Eds), International Review and Industrial and Organizational Psychology (volume 7). Chichester, UK: Wiley. Spector, P.E., & Goh, A. (2001). The role of emotions in the occupational stress process. Exploring Theoretical Mechanisms and Perspectives, 1, 195– 232. Spector, P.E., Zapf, D., Chen, P.Y., & Frese M. (2000). Why negative affectivity should not be controlled in job stress research: Don’t throw the baby out with the bath water. Journal of Organizational Behavior, 21, 79–95. Stanton, A.L., & Franz, R. (1999). Focusing on emotion: An adaptive coping strategy? In C.R. Snyder (Ed.), Coping: The Psychology of What Works. New York: Oxford University Press. Stone, A.A., Schwartz, J.E., Neale, J.M., Shiffman, S., Marco, C.A., Hickcox, M., Paty, J., Porter, L.S., & Cruise, L.J. (1998). A comparison of coping assessed by ecological momentary assessment and retrospective recall. Journal of Personality and Social Psychology, 74, 1670–80. Suh, E., Diener, E., & Fujita, F. (1996). Events and subjective well-being: Only recent events matter. Journal of Personality and Social Psychology, 70, 1091–102. Taris, T.W., & Schreurs, P.J.G. (2009). Well-being and organizational performance: An organizational level test of the happy-productive worker hypothesis. Work and Stress, 23, 120–36.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
44
January 21, 2011
12:53
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Tennen, H., Affleck, G., Coyne, J.C., Larsen, R.J., & Delongis, A. (2006). Paper and plastic in daily diary research: Comment on Green, Rafaeli, Bolger, Shrout, & Reis (2006). Psychological Methods, 11, 112–8. ter Doest, L., Mase, S., Gebhardt, W.A., & Koelwijn, H. (2006). Personal goal facilitation through work: Implications for employee satisfaction and well-being. Applied Psychology: An International Review, 55, 192–219. Totterdell, P., Wood, S., & Wall, T. (2006). An intra-individual test of the demandscontrol model: A weekly diary study of psychological strain in portfolio workers. Journal of Occupational and Organizational Psychology, 79, 63–84. Trougakos, J.P., Beal, D.J., Green, S.G., & Weiss, H.M. (2008). Making the break count: An episodic examination of recovery activities, emotional experiences, and positive affective displays. Academy of Management Journal, 51, 131–46. Trougakos, J.P., & Hideg, I. (2009). Momentary work recovery: The role of within-day work breaks. In S. Sonnentag & S.A.E. Guerts (Eds), Research in Occupational Stress and Well-being (volume 7): Current Perspectives on Job-Stress Recovery. Bingley, UK: Emerald. Vahtera, J., Kivim¨aki, M., Pentti, J., & Theorell, T. (2000). Effect of change in the psychosocial work environment on sickness absence: A seven year follow up of initially healthy employees. Journal of Epidemiology and Community Health, 54, 484–93. van den Tooren, M., & de Jonge, J. (2010). The role of matching job resources in different demanding situations at work: A vignette study. Journal of Occupational and Organizational Psychology, 83, 39–54. van Horn, J.E., Taris, T.W., Schaufeli, W.B., & Schreurs, P.J.G. (2004). The structure of occupational well-being: A study among Dutch teachers. Journal of Occupational and Organizational Psychology, 77, 365–76. Wall, T.D., Kemp, N.J., Jackson, P.R., & Clegg, C.W. (1986). Outcomes of autonomous workgroups: A long term field experiment. Academy of Management Journal, 29, 280–304. Warr, P. (1994). A conceptual framework for the study of work and mental health. Work and Stress, 8, 84–97. Warr, P. (2006). Differential activation of judgments in employee well-being. Journal of Occupational and Organizational Psychology, 79, 225–44. Warr, P. (2007). Work, Happiness and Unhappiness. Mahwah, NJ: Erlbaum. Warr, P.B. (1987). Work, Unemployment and Mental Health. Oxford: Oxford University Press. Warr, P.B. (1990). The measurement of well-being and other aspects of mental health. Journal of Occupational Psychology, 63, 193–210. Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. Psychological Bulletin, 98, 219–35. Weiss, H.M., & Cropanzano, R. (1996). Affective events theory: A theoretical discussion of the structure, causes, and consequences of affective experiences at work. In B.M. Staw & L.L. Cummings (Eds), Research in Organizational Behavior (volume 18). Greenwich, CT: JAI Press. Weiss, H.M., Nicholas, J.P., & Daus, C.S. (1999). An examination of the joint effects of affective experiences and job beliefs on job satisfaction and variations in affective experiences over time. Organizational Behavior and Human Decision Processes, 78, 1–24. Wiese, B.S., & Freund, A.M. (2005). Goal progress makes one happy, or does it? Longitudinal findings from the work domain. Journal of Occupational and Organizational Psychology, 78, 287–304. Williams, J.M.G., Watts, F.N., MacLeod, C., & Mathews, A. (1996). Cognitive Psychology and Emotional Disorders ( 2nd edn). Chichester, UK: Wiley.
P1: OTA/XYZ P2: ABC c01 JWST046-Hodgkinson
January 21, 2011
12:53
Printer Name: Yet to Come
STRESS AND WELL-BEING
45
Wrzesniewski, A., & Dutton, J. (2001). Crafting a job: Revisioning employees as active crafters of their work. Academy of Management Review, 26, 179–201. Xanthopoulou, D., Bakker, A.B., Demerouti, E., Schaufeli, W.B. (2009). Work engagement and financial returns: A diary study on the role of job and personal resources. Journal of Occupational and Organizational Psychology, 82, 183–200. Zapf, D., Dormann, C., & Frese, M. (1996). Longitudinal studies in organizational stress research: A review of the literature with reference to methodological issues. Journal of Occupational Health Psychology, 1, 145–69. Zellars, K.L., & Perrew´e, P.L. (2001). Affective personality and the content of emotional support: Coping in organizations. Journal of Applied Psychology, 86, 459–67.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
Chapter 2 BRAIN, EMOTION, AND CONTINGENCY IN THE EXPLANATION OF CONSUMER BEHAVIOUR Gordon R. Foxall Consumer Behaviour Analysis Research Group, Cardiff Business School, Cardiff University, Cardiff, CF10 3EU, UK The Behavioural Perspective Model (BPM) of purchase and consumption (Foxall, 1990) aims to improve understanding of behaviourism, intentionality, and cognition in the explanation of consumer choice. The initial phase of the BPM research programme has been to examine an extensional explanation of choice (based on radical behaviourism). This perspective has been adopted in order, first, to establish the boundaries of so parsimonious an approach to explanation and, secondly, to identify the scope for intentional and cognitive explanations of consumer choice. Empirical work demonstrates the value of the extensional construal in explaining the nature of consumer brand and product choices and the interpretation of consumer behaviour in relation to the situations in which it occurs. A limitation of this approach arises from the attempt to account in purely extensional terms for the continuity of consumer behaviour over situations. This makes the use of intentional language an inevitable part of the explanation of consumer behaviour but there remains the problem of employing intentionality in a logical and scientifically consistent manner rather than opportunistically. This chapter is concerned with the role of one aspect of intentionality, emotion, in the explanation of consumer choice in order to comprehend behavioural continuity. It draws upon the philosophy of psychology known as intentional behaviourism (Foxall, 2004) in order to ascribe intentionality in accordance with scientific canons of procedure that are consistent with evolutionary reasoning. This methodology extends Dennett’s (1969) emphasis on the role of evolutionarily consistent neuronal activity in the adoption of intentional explanations of behaviour by using the BPM to International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
48
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
relate emotionality to the contingencies of reinforcement that are responsible for the shaping and maintenance of consumer choice. All too often, consumer behaviour texts portray emotions as causes of behaviour that can be manipulated for managerial purposes rather than as elements of experience that participate in the social–scientific explanation of choice. There is a particular gap in our knowledge of the way in which emotions are related to publicly available influences on behaviour, the ‘contingencies of reward (or reinforcement) and punishment’, that are the most apparent determinants of choice in the contemporary marketplace. This chapter relates emotion to the environmental contingencies of which consumer choice is demonstrably a function and suggests how both the contingencies and the emotions related to them contribute to the explanation of consumer behaviour. An earlier chapter in the International Review of Industrial and Organizational Psychology (Foxall, 1997a) presented a critique of the social cognitive approach to the attitude–intention–behaviour sequence that remains a prevalent feature of models of consumer behaviour in the marketing literature. Finding a lack of empirical support for such a relationship, except where measures of emotion, cognition, and response reflected high levels of situational correspondence, the chapter argued in favour of an alternative, behaviour-analytical model of purchase and consumption, the Behavioural Perspective Model (BPM). The BPM was proposed as a means of systematizing the likely effects of those situational influences, conceptualized in terms of contingencies of reinforcement and punishment. This obviated the need to demonstrate attitude–intention–behaviour consistency on the understanding that all three were the result of underlying cognitive events; instead it enabled the verbal behaviours involved in the expression of attitudes and intentions and the overt behaviours involved in purchase and consumption to be differentially explained in terms of the contingencies of reinforcement and punishment to which each was uniquely subject. The overall aims of the model and the research programme in which it features has been to ascertain the feasibility of constructing a radical behaviourist model of consumer choice and, if this proves possible, to examine the epistemological status of such a model in order to build and test more complex theories of consumer behaviour. The underlying philosophical basis of the model has evolved (e.g., Foxall, 2004, 2007b) and empirical work has been generated that supports its underlying approach to the explanation of such aspects of consumer choice as produce and brand selection (Foxall, Oliveira-Castro, James, et al., 2007). One strand of empirical research indicates how reported emotional responses to consumer situations are related to specific patterns of environmental contingency (Foxall & Greenley, 1998, 1999, 2000; Foxall & Yani-de-Soriano, 2005) and the results of this work are applied in this chapter to illustrate and extend the theoretical developments that have taken place over the last dozen or so years. Recent developments in the theoretical portrayal of emotion and contingency, and its relationship to brain processes involved in the evaluation of alternative
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
49
courses of action, have increased the relevance of neuroscience to consumer choice; this emphasis is reflected in both the progression of the BPM approach generally (Foxall, 2008a) and the discussion in this chapter of the explanation of consumer behaviour. The chapter first reviews the BPM as a means of predicting such aspects of consumer behaviour as brand and product choice and examines the (extensional) nature of its explanation, going on to argue that, although this is sufficient for the prediction and control of its subject matter, it is not able to account for some aspects of behaviour such as its continuity. That emotion provides a means by which this gap can be filled permits an interpretation of consumer behaviour derived from extensional behavioural science and neuroscience. ‘Emotion’, understood as consisting in subjective experience and as intentional, can then be employed in the explanation of consumer behaviour in the rigorous terms proposed by ‘intentional behaviourism’ (Foxall, 2007b). Intentional behaviourism involves the responsible ascription of emotions on the basis of molar patterns of operant behaviour and related afferent–efferent linkages at the neuronal level. The BPM contingency matrix offers an alternative, empirically based, means of categorizing patterns of contingency in relation to emotional responses to consumer environments to that, say, of Rolls (1999). Mehrabian and Russell’s (1974) depiction of emotionality in terms of pleasure, arousal, and dominance is evaluated as a framework for the empirical investigation of contingency and emotion. The chapter continues with a discussion of the extent to which the intentionally construed model accounts for behavioural continuity and the personal level, whilst indicating how a behavioural interpretation can be bounded. It concludes by considering briefly the emerging theoretical synthesis among these variables and cognition.
THE BEHAVIOURAL PERSPECTIVE MODEL The Generic Model Consumer behaviour is influenced by both the economic and technical properties of goods on one hand and the social meaning of acquiring, owning, and using them on the other. People drive cars in order to get around and in order to be seen getting around, wear clothes for protection from the elements and to signal to everyone how well they are doing at work, adorn themselves with jewellery not only to impress their fellows or fit in with their expectations but to raise or confirm their own self-esteem. To the extent that consumption is influenced by these consequences, it is operant; to the extent that it reflects both the functional and the symbolic, it is under the influence of a complex of utilitarian and informational reinforcers. Businesses meet these consumer wants by offering marketing mixes that stress product attributes of both kinds, advertising and distribution channels that complement and enhance them, and price
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
50
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Consumer behaviour setting (CBS)
Utilitarian reinforcement Consumer situation
Utilitarian punishment
Behaviour Consumer situation Learning history
Informational reinforcement Informational punishment
Figure 2.1
Generic Behavioural Perspective Model (BPM).
levels that are consonant with both the technical–economic purposes and the social–psychological meanings that the resulting brands address. Both sources of reinforcement must be included in a behaviour–analytic model of consumer choice. So must the punishing consequences associated with each, for every economic transaction meets with aversive outcomes as well as those that reward. These consequential causes of behaviour are depicted on the right-hand side of the BPM (Figure 2.1). Shown on the left are the stimuli that set the occasion for these causal consequences should particular acts of purchase and consumption be enacted. The consumer behaviour setting (CBS) is composed of stimuli that signal the outcomes of behaviour – the availability of particular brands, for instance, within a supermarket – and stimuli that motivate the behaviour – say a pointof-sale advertisement that emphasizes the unique taste or value-for-money that buying the item will generate. Open settings permit a wider range of behaviours to be enacted (‘offer more choice’) than closed settings in which just one or a few behaviours are possible. CBSs can be described on a continuum from relatively open to relatively closed. This conceptualization is especially relevant to the study of consumer behaviour and, particularly, retail research. Generally, although not inevitably, in the relatively closed setting, persons other than the consumer arrange the discriminative stimuli that compose the setting in a way that compels conformity to the desired behaviour. Such conformity is achieved by making reinforcement contingent on such conformity. The open setting, however, is marked by a relative absence of physical, social, and verbal pressures to conform to a pattern of activity that is determined by others (what ecological psychologists call a behaviour programme: Schoggen, 1989); it is comparatively free of constraints on the consumer, who, thus, has an increased range of choices. He or she has some ability to determine personal rules for choosing among the products and brands on offer, which stores to visit, and so on. A typical open setting is represented by a departmental store in which the consumer can move from section to section, browsing here, considering
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
51
there, making a purchase, or leaving altogether to find another store, or even giving up on shopping and going home. In contrast, extremely closed CBSs are exemplified by the dental surgery or the gymnasium where only one course of action is reinforced and removing oneself from the situation, while not impossible, is fraught with social and, ultimately, health-related costs. Less extreme but still distinctly closed for the consumer behaviour context, a bank is usually a physically closed setting, arranged to encourage orderly queuing by customers and to discourage behaviour that detracts from the efficient execution of transactions. Social and verbal elements also enter into the closed nature of the setting: the single-file line that leads to the teller window does not encourage conversation, at least not to the point where the business of the bank is likely to be delayed. Social and regulatory aspects of the CBS are also apparent in less formal contexts such as having to purchase a birthday gift for a friend, which is closer to the centre of the open–closed continuum. The setting is closed in so far as the consumer conforms to social rules that describe moral or material rewards for reciprocity or punishments for ignoring generosity in others, although it has facets of openness stemming from the capacity of friends to depart from social norms or even break the rules on occasion, not only without censure but with a strengthening of the relationship. Also on the left of the BPM shown in Figure 2.1 is the consumer’s learning history for this and similar products, what he or she has done in the past and the reinforcing and punishing outcomes this has had. The learning history primes the discriminative stimuli (SD ) and motivating operations (MO) that make up the CBS and evokes the behaviour that will generate or avoid the consequences on offer. SD are stimuli in the presence of which the individual discriminates behaviourally by performing a response that has previously been reinforced in these or similar circumstances; MO are stimuli that enhance the ability of a reinforcer to strengthen a response. For instance, while the wording of an advertisement, ‘Persil washes whiter!’ may be a SD for buying this product, the accompanying picture of child wearing pristine, clean clothes might enhance the efficacy of the reinforcer if this symbol has previously been associated with sound parenting (Fagerstrom, Foxall, & Arntzen, 2010). It is the consumer situation that results from the interaction learning history and CBS that is the immediate precursor of consumer behaviour. The consumer situation induces or inhibits particular consumer behaviours depending on whether the consumer behaviour analysis is relatively open or relatively closed. In this non-intentional construal of the BPM, the consumer situation thus amounts to the scope of the setting, that is, its degree of openness or closeness weighted by the individual’s consumption history directly impacts upon the probability that particular consumer behaviours will occur. The stimuli that comprise the CBS and that enter into the consumer situation induce the consumer to discriminate his or her behaviour by purchasing or consuming certain products and services, marques and brands rather
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
52
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
than others. The behaviours performed are those that have been reinforced in the past and the discriminative stimuli, motivating operations, and learning history that interact to form the consumer situation are associated with utilitarian or functional and informational or symbolic reinforcements that will result from current behaviours. These consequences of behaviour, shown on the right-hand side of the model in Figure 2.1, may be positive or aversive, reinforcing or punishing in their effects on future consumer choice. Utilitarian reinforcers, which are mediated by the products themselves, are associated with the technical and operational qualities of the item bought and consumed. Informational reinforcers are socially mediated, however, and consist in performance feedback on the consumer behaviour in question or other behaviours instrumental in making it possible. Almost any car will provide the utilitarian benefits of transporting its owner or driver, ‘getting from A to B’, that is. But a Porsche usually delivers the performance feedback that comes from recognition of the owner’s occupational status, social position, and other sources of honour and prestige. Like other socially constructed, symbolic outcomes of behaviour, informational reinforcers are relative to the values the community: in a social system conscious of CO2 emission or fossil fuel consumption, a prestige car might not confer the positive social feedback just assumed. Consumers acquire combinations of utilitarian and informational benefits in the course of buying and using products, represented as a pattern of low/high utilitarian reinforcement and low/high informational reinforcement. The idea of a pattern of reinforcement replaces that of schedule of reinforcement, something applicable more to the precision of the laboratory than interpreting complex choices in the marketplace. Defined in terms of pattern of reinforcement, consumer behaviour falls into one of four operant classes: maintenance, accumulation, hedonism, and accomplishment (Figure 2.2). The BPM Contingency Matrix (Figure 2.3) comprises eight distinct categories of contingencies, the outcome of combining CBS scope and reinforcement patterns (Foxall, 2010). The following sections reveal that the generic BPM shown in Figure 2.1 can be construed in both extensional and intentional forms and that these offer different levels of explanation of consumer behaviour. Thus far, empirical work has emphasized the extensional construal of the model; this chapter
Low utilitarian Reinforcement
High utilitarian reinforcement
Low informational reinforcement
MAINTENANCE
HEDONISM
High informational reinforcement
ACCUMULATION
ACCOMPLISHMENT
Figure 2.2
Patterns of reinforcement and operant classes of consumer behaviour.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
53
BEHAVIOR SETTING SCOPE
Open
Closed CC2
ACCOMPLISHMENT
Fulfillment
CC1
high
Status consumption IR CC4
HEDONISM
CC3
Popular entertainment
Inescapable entertainment CC6
ACCUMULATION
Token-based consumption
low
CC5
high
Saving and collecting IR CC8
MAINTENANCE
Mandatory consumption Figure 2.3
High UR
Low UR
CC7
Routine purchasing
low
The BPM Contingency Matrix.
argues for an intentional construal based on research into the role of emotionality in consumer choice.
Empirical Research What kind of model is the BPM and what confidence may we have that its explanatory categories actually relate consistently to consumer behaviour? The model was devised to overcome the drawbacks of a social cognitive portrayal of choice by examining the possibility that a model of consumer choice devoid of unobservables could be constructed and employed in the interpretation of purchase and consumption. The aim was then to test this model to destruction. Two outcomes of this procedure were possible. Either an entirely non-intentional approach would be shown adequate for the explanation of consumer behaviour or, more probably, the necessity of employing intentional terminology and therefore intentional explanation would become evident. These founding objectives guided the BPM research programme since its inception (Foxall, 1988). When the earlier chapter in this series was published (Foxall, 1997a), the BPM had been used solely as an interpretive device, a means of recasting familiar themes in consumer research in behavioural terms in order to demonstrate that this portrayal accounted for the empirical data at least as well as did cognitive theories and to suggest alternative avenues for research
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
54
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
not included in the social cognitive paradigm. Hence, attitude–behaviour relationships, consumer innovativeness, ‘green’ purchasing and consumption, as well as general purchasing, consumption, and saving behaviours were reexamined from this antithetical perspective. At that point little more was expected of the BPM approach than its contribution to the growth of knowledge in a Feyerabendian manner (Feyerabend, 1975) through a clash of competing interpretations. The interim has been marked by a large volume of empirical research intended to test the predictive capacity of the BPM. This work has increasingly incorporated techniques pioneered in behaviour analysis and, especially, behavioural economics (Hursh, 1984) in order to investigate behaviour in the non-intentional terms that are the hallmark of operant psychology and experimental economics. Defining economic choice as the allocation of behaviour within a framework of costs and benefits (Staddon, 1980), this perspective adopted matching and maximization techniques to the study of consumers’ brand and product choices (Foxall, 1999; Foxall & James, 2002, 2003; Foxall & Schrezenmaier, 2003; cf. Curry, Foxall, & Sigurdsson, 2010). This translational research programme has demonstrated that the matching phenomena explored by Herrnstein (1961, 1970, 1997) provide psychological measures of standard microeconomic variables such as product and brand substitutability, complementarity, and independence (Foxall, James, Oliveira-Castro, et al., 2010; Romero, Foxall, Schrezenmaier, et al., 2006) which also serve to define product categories, sub-categories, and brands (Foxall, James, Chang, et al., 2010). Further behavioural economics research has involved the operational measurement of utilitarian and informational reinforcement and the estimation of price elasticity of demand coefficients for brands feature varying combinations of these elements of reward (Foxall, Oliveira-Castro, Schrezenmaier, et al., 2004; Foxall, Oliveira-Castro, James, et al., 2007; Foxall, James, Chang, et al., 2010; Oliveira-Castro Foxall, & Schrezenmaier, 2006; Oliveira-Castro, Foxall, & James, 2008; Oliveira-Castro, Foxall, James, et al., 2008). This work underpins the BPM approach by showing that consumer demand is a function not only of price but of the pattern of reinforcement delivered by goods.
An Extensional Construal Extensional language Theories of behaviour are a matter of how language is deployed to make a subject matter intelligible: they are methodological devices rather than ontological descriptions. We understand behaviour in daily life by reference to a person’s intentions: he or she did that because they expected to, desired the outcome, or believed it was morally incumbent. Such intentional explanation is also at the heart of social and behavioural science: If A desires x and believes
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
55
that doing y will lead to x, he or she will do y (e.g., Rosenberg, 1988) and economics (Rosenberg, 1992). The essence of the ‘attitudes’ represented by the italicized words is that they form sentences that are about something other than themselves: we do not simply ‘expect’ or ‘desire’ or ‘like’ per se; we ‘expect that p’, ‘desire that p’ or ‘believe that p’. Rediscovering the scholastic philosophers’ analysis of aboutness or intentionality, Brentano (1874) took it to be the mark of the mental, that which distinguished the substances that composed the world. The modern interpretation of intentionality inheres, however, in the recognition that the attitudes and the proposition they take as their predicates (hence, ‘propositional attitudes’) serve only to distinguish one kind of sentence, one kind of explanation, from another (Chisholm, 1957; Dennett, 1969; Russell, 1912). The radical behaviourist explanation that the BPM research programme seeks to evaluate is wholly different from an intentional explanation. First, it consists of the identification of the environmental stimuli that control behaviour; when these have been identified and described (in non-intentional terms), the behaviour has been explained. Secondly, it resolutely adheres to a form of explanation that strenuously avoids intentional terms such as ‘believes’ and ‘desires’ – it is extensional. The essence of behaviourism is the avoidance of intentionality in its scientific discourse (Foxall, 2004). As the founder of radical behaviourism, Skinner (1945) strove to avoid intentional terms in scientific discourse. His meticulous standards of linguistic expression inhere in his later writing that, ‘We say that spiders spin webs in order to catch flies and that men set nets in order to catch fish. The “order” is temporal’ (Skinner, 1969: 193). That is, we are saying simply that first the spider spins and then it catches flies, that men first set their nets and then catch fish. Neither the spider nor the men pursue a purpose or seek to fulfil an intention when spinning or setting. Skinner (1971: 18) is also scrupulously careful to avoid intentional language in defining operant behaviour as ‘behavior that operates on the environment to produce consequences’. There is now no suggestion that the operation is performed ‘in order’ to produce consequences, emphasizing that the order implied is just that of temporal sequence. Extensional linguistic convention is the heart of radical behaviourism, the locutionary style that defines it as a philosophy of psychology (Foxall, 2004). More technically, extensional language consists in sentences whose truth value lies in the substitutability of coextensive terms: our rephrasing, salve veritate, ‘This consumer has purchased Brand X’ as ‘This consumer has purchased Product Y’ when X is the sole member of the Y product category is an example of extensionality. This is the language that is generally taken to be that of hypothetico-deductive science: it leads directly to statements that can be tested by comparison with empirical evidence (Chisholm, 1957; Dennett, 1969; Quine, 1960). By extensional language is meant, therefore, sentences that are referentially transparent, that conform to the normal usages of science, contain no intentional terms, sentences that allow substitution of identicals.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
56
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Extensional language is a means to express statements that met the truth value of science devised by Quine who argues: ‘All failures of extensionality are failures of substitutability of identity’ (1960: 151). Terms that have the same extension can be substituted for one another in extensional sentences salve veritate but such substitutability of identity is not possible salve veritate in intensional sentences. In the latter, the terms have different intensions or meanings in the individual’s mental or subjective or private life (or at least that which is being attributed to him or her). These two expressions nevertheless have different intensions or meanings and are not therefore substitutable if the truth value of the sentence is to be preserved. The extensional construal of the BPM The structure of the extensionally conceived model is such that consumer behaviour is portrayed as the outcome of functional relationships between a consumer situation and a response, where the consumer situation is the intersection of a CBS and a learning history of reinforcement and punishment by utilitarian and informational consequences. CBS scope, in so far as it contains the consequences of behaviour that have formed the individual’s learning history, can thus be said to be the ‘cause’ of consumer behaviour, in the sense that the behaviour is a function of the stimuli that compose CBS scope. The consumer situation is thus understood in the extensional model solely in terms of the scope of the CBS. An extensional model incorporates causal influences but does not employ intentional idioms or reasoning to explain its dependent variable(s). As such, the BPM locates consumer choice at the intersection of the consumer’s learning history and the current CBS, that is, where the experience of consumption meets an opportunity to consume anew. This intersection of time and space forms the consumer situation, the immediate shaper of approach–avoidance responses involved in purchase and consumption. We have already seen that in line with a theory that takes the prediction and control of behaviour as its raison d’´etre, the extensionally defined consumer situation is coterminous with the scope of the CBS. The consumer situation is defined in terms of the range of options available to the consumer as determined by the stimulus antecedents of feasible behaviours, some of which will have been present on earlier consumption occasions; in the presence of the individual’s learning history, these initially neutral stimuli are transformed into the discriminative stimuli and motivating operations that set the occasion for current choice. The consumer’s consumption history invests the initially neutral stimuli with a kind of meaning, which consists in no more than the capacity to generate specific kinds of approach and/or avoidance behaviours that produce consequences to regulate the rate of recurrence of the behaviours that produced them. Thus, both CBS and consumer situation simply set the occasion for three types of behavioural consequence: utilitarian reinforcement, which consists in the functional outcomes
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
57
of behaviour; informational reinforcement, which stems from the symbolic outcomes, principally performance feedback; and aversive/punishing consequences, the costs of purchase and consumption. Such aversive outcomes can themselves be subdivided into those that are utilitarian in nature and those that are symbolic. The components of the model are operationally defined, specified in terms of the functional relationships that stem from their observable impacts upon behaviour. Hence, an operant response, one that operates on the environment to produce the consequences that govern its subsequent rate of emission, ‘is not simply a response that the organism thinks will have a certain effect, it does have that effect’ (Smith, 1994: 127–8). Further, a reinforcer ‘is not simply a stimulus that the organism desires to occur. It is a stimulus that will alter the rate of behaviour upon which its occurrence is contingent’ (Smith, 1994: 127–8). A discriminative stimulus ‘is not simply a stimulus that has been correlated with a certain contingency in the organism’s experience, it is one that successfully alters the organism’s operant behaviour with respect to that contingency’ (Smith, 1994: 129). It does not signal, refer to, or represent the utilitarian and informational reinforcers or punishers likely to be contingent on the performance of particular responses: it simply ‘sets the occasion’ for these consequences. The rationale for building a model of consumer behaviour in these terms derives not from the conventional wisdom of hypothetico-deductive scientific methodology but from the need to examine whether a theory of choice can avoid the intentional language of beliefs and desires, that is, statements that do not permit the substitutability of co-extensives. The key motivation for this is the finding that both cognitive and behaviourist accounts of consumer choice are equally supported by the empirical evidence on attitudinal–behavioural correspondence (Foxall, 1997c, 2005): to favour the former whilst ignoring the latter represents not only a slavish adherence of an applied field to the prevailing paradigm of the disciplines from which it derives, in this case cognitive psychology, but an intellectually closed perspective that will not conceive of explanation in terms not belonging to this framework of conceptualization and analysis. But why radical behaviourism? The central fact in the delineation of radical behaviourism is its conceptual avoidance of propositional content. This eschewal of the intentional stance sets it apart not only from cognitivism but from other neo-behaviourisms. Indeed, the defining characteristic of radical behaviourism is not that it avoids mediating processes per se but that it sets out to account for behaviour without recourse to propositional attitudes. Based rather on the contextual stance, it provides definitions of contingencyshaped, rule-governed, verbal and private behaviours that are non-intentional. For reasons of disinterested curiosity, therefore, as well as the more pragmatic search for a general explanation of choice, a research programme based on the development and evaluation of an extensional model of consumer choice becomes inevitable. This is what the extensional construal of the BPM attempts.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
58
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
The extensional model has made considerable advances in the interpretation and prediction of complex human choice, some of which are reviewed above. But there are limits to extensionality as a result of which it is necessary to incorporate intentional language into the model (Foxall, 2004, 2007c). Indeed, radical behaviourists already use such language when they are faced with the limitations of extensionality. What is essential to the BPM research programme is to understand why radical behaviourists do this and, if it is necessary, to do it rigorously and consistently. The evolution of the BPM as a generic portrayal of consumer choice which includes extensional, intentional, and cognitive theories elucidates the kind of theoretical development required by applied behaviour analysis in order to make comprehensive sense of its subject matter. We shall argue that emotion is an example of intentionality that provides a missing link in the attempt to account for the continuity of behaviour. An interesting parallel development is that of cognitive psychologists who have also shifted their focus to emotion in recent years, although for rather different reasons (see e.g., Hodgkinson & Healey, in press). What is being emphasized here is that the philosophically legitimate ascription of intentionality requires great care if it is to avoid being a mere convenience. There needs, first, to be a solid rationale for the emergence of intentional language in the BPM account of consumer behaviour, followed by a reasoned understanding of why emotion has assumed an essential role in the explanation of behaviour. These are pursued below in terms of the inability of the extensional model to explain behavioural continuity and the exposition of why a particular conception of emotionality can fill this gap. The logic by which the ascription of emotion can be justified is discussed in terms of neurobiology and environment–behaviour relationships and the resulting philosophy of psychology, intentional behaviourism, is outlined. Accounting for Behavioural Continuity Although an extensional account facilitates prediction and control by reference to the stimuli that determine the rate of recurrence of behaviour, it cannot explain the continuity of behaviour over settings and situations. As the consumer moves from setting to setting, they may be faced with stimuli that differ from those previously encountered; yet they act in a manner consistent with the behaviours displayed in those rather different settings. In contrast, there are other occasions when the pattern of behaviour displayed by the consumer in familiar settings deviates markedly from what might be expected as, for example, when a lazy glutton starts to eat only less-fattening foods and to take up exercise. What accounts for such deviations from patterns of behaviour that have hitherto remained constant over time? It is beyond the capacity of a purely extensional model to explain the continuity of behaviour across settings or the discontinuity of behaviour that occurs when the consumer switches to a new pattern of choice that has not previously been reinforced. The recurrence of
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
59
the same or similar stimuli in a succession of settings is not generally sufficient for such explanation: only in the most closed experimental setting could it be taken as such. In complex situations of purchase and consumption, it is usually impossible to isolate the stimuli that are responsible for consumer responses with the precision available in the laboratory and without interpretation based on the ascription of intentionality. Moreover, most stimuli differ somewhat from setting to setting. Physiological changes resulting from behaving once in one setting cannot be shown to explain the continuity of behaviour even across settings that exhibit stimulus similarities let alone among divergent stimulus contexts. Rules cannot account for behavioural continuity or pattern shifts unless some mechanism of perception, encoding, interpreting can be identified. (Rules may be MO in an extensional account, but while they may predict or control behaviour they cannot explain behavioural continuity or deviations from established behaviour patterns.) Only by employing intentional language can we provide an explanatory account. Foxall (2007b) presented the argument for incorporating intentionality into theories of choice on the grounds that an extensional theory could not of itself account for the continuity/discontinuity of behaviour (see also Foxall, 2004, 2007c, 2008b). Some examples may make this argument more concrete. First, take a person whom we have observed drink alcohol heavily on a daily basis but who, we note, now drinks only on Friday evenings and confines himself to two drinks. As Rachlin (1995) says, we might explain this behaviour by saying that the individual concerned has ‘decided’ on this change. The use of intentional language appears inevitable if we are to account for this behavioural discontinuity. Second, consider a heavy user of the four brands, A, B, C, and D, that comprise this individual’s consideration set for a particular consumer nondurable, who now includes a new brand, E, in their repertoire. As is the case for many consumers in affluent societies, we cannot assume anything about the individual’s learning history except that they are a heavy user of the product category. It seems impossible to account for their inclusion of the new brand without referring to the individual’s beliefs and desires. Finally, let us consider the case of a participant in an operant experiment who maintains their behaviour pattern even though the contingencies governing reinforcement of that behaviour have changed. Again, there is little we can say about the individual’s learning history. It seems reasonable to assume some control of their overt behaviour by their private verbal behaviour especially since we have no evidence of prior control of overt behaviour via instructions. It does not seem that behaviour such as this can be explained other than ascribing certain beliefs and desires to the experimental participant. These examples of behavioural continuity/discontinuity define a continuum of behavioural change which relates the sequence of observed behaviour to changes in the attendant sequence of reinforcement. The behaviour of the heavy user of alcohol which reflects some early signs of addiction (such as bingeing followed by remorse which is not sufficient to allay further bouts of
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
60
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
heavy drinking) but whose behaviour changes to a more restrained pattern of moderated drinking cannot be explained in terms of the contingencies alone. The initial phase leads to aversive consequences which do not reduce the level of alcohol consumption; the subsequent behaviour pattern is adopted before the novel consequences of restrained consumption have had time to exert an effect on choice. The abrupt change in the first molar pattern of behaviour can be explained only in terms of the individual’s having made a decision to try a different style of behaviour. Such change is described as major or discontinuous. The consumer who adopts a new brand also exhibits a change in their sequence of behaviour, not by abandoning the existing pattern of choice but by supplementing and extending it. There is a change in behaviour but it amounts to no more than trying a new brand in a product category of which the consumer has much experience (i.e., a novel version of a familiar pattern of reinforcement). Most consumers of a product category purchase within a small consideration set of tried and tested brands; many, especially the heavier users of the product, try new brands that appear to contain the characteristics of the product class; some of those who try it incorporate the new brand into their future consideration set (Ehrenberg, 1972). Most consumers who select a new brand in this way or change to another in their existing consideration set choose one that contains a similar combination of functional and symbolic benefits (the pattern of reinforcement) as existing members of the set (Foxall, Oliveira-Castro, & Schrezenmaier, 2004). The prediction of such behaviour follows easily enough from consideration of the contingencies alone (at least for aggregates, not necessarily for individuals) but an explanation of the change itself requires consideration of the processes of comparison and recognition that must precede the change. How are the verbal stimuli (e.g., advertisements) translated into the new pattern of consumer behaviour via comparison with the characteristics of the brands already in the consumer’s repertoire? Selective perception, beliefs, and desires must be used as part of the explanation of such behaviour. It is not sufficient, therefore, to say that more continuous change of this sort, even though it may be readily related to the contingencies, is ‘explained’ by its embodiment of stimulus or reinforcer discrimination and generalization. Use of such terminology merely redescribes the observed choices. Finally, the behaviour of the experimental participant who exhibits rigidity in the face of changing contingencies is an example of behavioural continuity that cannot be explained in terms of the contingencies themselves (Lowe, 1983). The situation is exemplified by the consumer who continues to purchase and use a particular brand of razor blades even though the quality of the shave obtained from them has markedly diminished. Why is human behaviour so insensitive to changes in contingencies when this is not true of nonhumans? The person presumably has not perceived the change in contingencies and is operating according to a self-generated rule reached in decision making prior to the contingency change. The behaviour of persons in this situation often
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
61
comes to conform to the contingencies after time. How does this change in perception occur? Is there further decision making?
INTENTIONAL BEHAVIOURISM ‘Neural Intentionality’ It is one thing to accept the need to incorporate intentional terms into an explanation of consumer choice, quite another to propose a legitimate means by which this might be accomplished without succumbing to the temptation to base intentional inferences naively on the behaviour to be explained. Intentional behaviourism proposes reasoned criteria for attributing intentionality in a manner consistent with evolutionary criteria. The starting point is Dennett’s (1969) suggestion that intentionality be ascribed on the basis of evolutionarily consistent afferent–efferent neural links that account for how organisms select a particular behaviour (reaching out) as the appropriate response to a given internal state of deprivation (hunger) and a given environmental context (availability of alternative food sources, each with its own costs of procurement). ‘Afferent’ and ‘efferent’ denote functions of neurons which are cells in the nervous system that transmit impulses to other neurons. Dendrites, of which there are a number to each cell, receive signals from other neurons and are accordingly known as afferent. Axons, of which each cell has only one, transmit signals to other neurons and are, therefore, known as efferent. These terms are used to denote the functions of neurons by reference to the direction in which they transmit impulses: towards the central nervous system (CNS) in the case of afferent or sensory neurons, away from the CNS in the case of efferent or motor neurons. The thrust of Dennett’s (1969) system appears on first reading to be that intentionality can be ascribed at the personal level (that of people and minds) on the basis of considerations that arise at the sub-personal level (that of brains and neuronal functioning). As his work developed, however, Dennett (1978a, 1987) increasingly sought method for ascribing intentionality at sub-personal levels, a complicating factor that a keener reading of his earliest work reveals to be an underlying current. At the heart of his method of using extensional sentences as the basis of ascribing intentionality, is Dennett’s linking of basic biological facts with their evolutionary history: ‘Intentional description presupposes the environmental appropriateness of antecedent-consequent connections; natural selection guarantees, over the long run, the environmental appropriateness of what it produces’ (Dennett 1969: 41). Dennett (1969) portrays the peripheralist approach of the behaviourists as incapable of producing a coherent science of behaviour because of its inability to incorporate the use of intentional language in everyday discourse. He assumes as evidence for the existence of entities described intentionally the fact that people speak as though these things existed. Above all, Dennett berates
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
62
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
the behaviourists for being unable to produce an account of human behaviour including verbal behaviour that proceeds in the absence of intentional terms. The behaviourist paradigm, parsimoniously reduced to observable stimuli and responses, is sufficient to predict and control learning but is incapable of explicating in nonintentional language what is learned. Dennett’s solution is to offer first intentional characterizations of behavioural events and their causes and then to seek extensional justifications for doing so. He seeks ground rules for this translation from the intentional to the extensional in the logic of evolution by natural selection. His argument is that the brain must have evolved in a way that enables it to discriminate the significance of afferent stimuli in terms of the effectiveness of efferent responses on the environment on which the organism depends for its survival and biological fitness. However, this argument anticipates a resolution of the very problem Dennett is trying to solve by arbitrarily ascribing intentionality to neural systems in the form of their intuiting the ‘significance’ of afferent stimuli. Dennett’s project fails by dint of its circularity and by not recognizing that evolution is a process of elimination rather than of insight. Three methodological points emerge from this consideration, all of which entail breaks with Dennett’s centralist approach. First, intentionality can be responsibly ascribed only at the personal level. Secondly, intentionality can be ascribed only to entities whose behaviour can reasonably be ascribed to intentional functioning – computers, incapable of feeling pain, are eliminated (Dennett, 1978b). Thirdly, intentionality can be ascribed only on the basis of demonstrated causal relationships between behaviour and the products of evolution by natural selection (i.e., appropriate neural functioning) and, as is argued below, environment–behaviour connections established by operant learning. In view of this, intentional behaviourism applies to patterns of molar operant behaviour rather than single instances of choice. Intentional behaviourism (Foxall, 2007b) rests on the premise that intentionality can be legitimately ascribed to the individual in order to explain his or her behaviour on the following bases. First, in line with Dennett’s thinking, it must be possible to show afferent–efferent linkages at the neuronal level that are logically consistent with both the behaviour to be explained and the intentionality to be ascribed in explaining it. However, the means of making such ascriptions is more complicated than in Dennett’s original treatise (Dennett, 1969). These linkages must, as was mentioned above, be shown to have evolved through natural selection; but neuronal plasticity occurring during the lifetime of the individual must also be related to the contingencies responsible for it. Secondly, it must be possible to demonstrate that the pattern of behaviour to be explained is a molar sequence of operant choice rather than a single instance. Dennett’s concentration on the molecular level, which refers to the explanation of a single instance of behaviour, makes no contribution to our primary concern to explain the continuity of behaviour. After all, the task we have set ourselves
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
63
is the elucidation of the means by which such continuity can be explained in behavioural science (including occasions of discontinuity). The argument that the ascription of intentionality relies on demonstrating appropriate afferent–efferent links (Dennett, 1969) can be conveniently called the ‘neuro-intentional approach,’ although this is not Dennett’s term. It is a necessary although insufficient element of the explanatory method of intentional behaviourism. Dennett’s (1969) argument that content may be reliably ascribed on the basis of evolutionarily consistent afferent–efferent linkages is consistent with the view that such content belongs at the personal level of explanation while the extensional neuroscience from which it derives is part of the sub-personal level. As long as these levels are kept separate, the use of both intentional and extensional routes to knowledge can be legitimized within the same framework of exposition. Intentionality and Contingency Intentionality is to be ascribed based on evolutionarily consistent afferent–efferent linkages identified by neuroscience (Dennett, 1969). Neuroscience is an extensional science that is concerned with the sub-personal level of explanation. Intentionality is ascribed at the personal level. The consideration of afferent–efferent neuronal links justifies the role of neuroscience in the explanation of behaviour through the ascription of intentionality. Another extensional approach to knowledge, behavioural science, is also involved and, within that, the demonstration that molar patterns of the behaviour that is to be explained can be functionally related to their consequences. This requirement has two implications. First, we are dealing with operant relationships, that is, with the assumption that the behaviour under investigation is explicable in terms of the patterns of reinforcement and punishment associated with it. The insistence on operant behaviour is particularly justified in that reinforcement and punishment are clearly implemented in the brain (Rolls, 2005). This is presumably an evolutionary endowment and makes the identification of relevant afferent–efferent links of the kind Dennett insisted upon. Secondly, we are concerned not with the interpretation of single instances of behaviour but with sequences of choice that correlate with sequences of consequences. The necessity of demonstrating a molar pattern of operant behaviour is, therefore, twofold: (i) knowing what we want to explain, and (ii) including ontogenetic as well as phylogenetically shaped afferent–efferent links in the process of ascribing intentionality (i.e., incorporating development as well as evolution). In terms of the first requirement, Dennett is sketchy on what behaviour is, why it matters, and how it enters into the explanation. The intentional stance approach does not show how behaviour is to be included in the ascription of intentionality. Dennett says that the system to be predicted must be invested with the beliefs and desires it ought to have given its history and position. Only if we know its behavioural learning history and its current
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
64
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
behaviour setting can this be done. This requires something closely akin to the BPM: the elements entering in to the consumer situation and the consumer situation itself, including the various consequences of behaviour that are (i) signalled by the discriminative stimuli and motivating operations that compose the setting and which have (ii) previously enabled a learning history to be established. One way of getting at this information is through consideration of the molar pattern of choice. A more refined approach is yet to be established in which we can appreciate more fully the contingent consequences of behaviour (informational reinforcement and utilitarian reinforcement and the pattern of reinforcement to which they give rise) so that the attitudinal and propositional components of intentionality can be properly adjudged and ascribed. In order to meet the second requirement, we must reach an understanding of how afferent–efferent links are formed and strengthened by ontogenetic development as opposed to the phylogenetic processes to which Dennett alludes. Natural selection is not the only source of afferent–efferent linkages that influence behaviour. In the course of individual development, synaptic strength is affected by the effects of reinforcement on the firing rate of neurons making particular responses more or less probable in the future (Hebb, 1949; Knorski, 1948). ‘Just like the muscles of your body, connections in the brain will strengthen and grow as they are exercised. . . The brain cells that are involved in the activities that occur most frequently will have extensive connections, whereas those that are used less frequently will be pushed out of the way, and their targets will be taken over by their more hardworking neighbours’ (Greenfield, 2000: 62; for further discussion see Frey, 1997). These processes are themselves the result of natural selection of course but they represent the effects of voluntary or operant behaviour on synaptic strength, an additional influence to that produced phylogenetically and which is largely responsible for involuntary or reflexive behaviour. In defining voluntary and involuntary responses, Skinner (1953: 110–13) is careful to emphasize that they are equally determined by environmental events and that neither implies free will. Such synaptic strength is brought about in the course of the organism’s lifetime as a result of cumulative behavioural repetitions. The resulting efficacy of the afferent–efferent links involved should be taken into consideration in the ascription of intentionality just as those that evolved in the lifetime of the species deserve notice. In fact, given the importance of voluntary behaviour to human behaviour, it is likely that such influences may be the greater not only in determining choice but in allocating intentionality to it.
THE ASCRIPTION OF EMOTION Salient Facets of Emotion Emotions do not constitute a uniquely definable class of entities and an absolute definition seems impossible. Philosophers and psychologists differ considerably in defining emotion (Frijda, 2008; Solomon, 2008). Even separating
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
65
emotion from other sensational states demands intense precision (Bennett & Hacker, 2003). Price (2005: 11) distinguishes emotions from bodily feelings (nausea), reflex responses (startle), moods (persistent anger), character traits (such as cowardice and possibly also depression and anxiety), emotional attitudes (love which manifests as other emotions such as pride or sadness depending on the occasion and what happens to the object of love). What is especially germane to the discussion of emotion in the current context, however, are the understandings that emotions manifest as subjective feelings based on physiological events, that they are intentional, may have a cognitive component (depending on what we believe, e.g. about the likelihood that a tiger will attack us), and may seem to have a causative role in directing behavioural responses (Roberts, 1988). In this section, we pay particular attention to the first two of these, briefly returning later to questions of cognitive content and causation. Emotion as subjective experience Whatever view is taken of the causes and components of emotion, subjective feelings are especially relevant to the mode of explanation pursued here. The theoretical basis of the present argument emphasizes the evolutionary logic by which felt emotion arises from behaviour that is performed in a particular context of contingent reinforcement and punishment. We may add that felt emotion influences learning history to guide further behavioural choices in similar contexts. Unless this minimal assumption is made, it is difficult to appreciate why strong and insistent emotional responses evolved. Emotional feelings undoubtedly are caused by neurophysiological processes but it is important (as Barrett, Mesquita, Ochsner, et al., 2007, who adopt Searle’s, 2000, ‘biological naturalism’ approach, argue) to keep separate the phenomenological experience of an event and its causation (cf. Izard, 2009). The fact that emotional feelings are instantiated at the sub-personal level of neurophysiological processing is not sufficient to define emotion; equally important is the content of emotion, the subjective experience which is the outcome of both neural firings and contextualized behaviour. It is a reasonable assumption that emotion, however it is defined, will not influence further behaviour unless something is felt, even though the feeling itself, as an entity that cannot be subjected to direct laboratory investigation, may not be a demonstrable cause of that behaviour. Having argued that an account of emotion must treat both its causes and its content, Barrett, Mesquita, Ochsner, et al. (2007) abstract two further principles of biological naturalism that relate to the role of emotional feelings in the explanation of behaviour. First, they note that a wholly causal account of emotional content may not be feasible. In wider philosophical terms we may note that there are two levels of explanation involved here, the personal and sub-personal which may not be conflated (see the discussion of Dennett’s development of the intentional stance in Foxall, 2007a). McGinn (1991, 2004)
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
66
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
points out that conscious experience is not and may never be reducible to physiological operations. It is, as Barrett, Mesquita, Ochsner, et al. (2007: 376) point out, the scientist’s work to supply the means of bridging these levels, which is precisely the task adopted by intentional behaviourism (Foxall, 2004), which, as noted earlier, is the primary explanatory device adopted in this chapter. Secondly, they point out that the ontological subjectivity of emotional feelings refers to their unique, personal, and private nature. The ontological status of emotion feelings cannot be abrogated by measures of behaviour, physiology, or neural functioning even if these are reliable indices of felt emotion. Barrett, Mesquita, Ochsner, et al.’s solution inheres in the dictum that ‘To know what emotion feels like, it is necessary to ask people what they experience’ (2007: 376; see also Barrett, Niedenthal, & Winkielman, 2005). However, while the truth of this is unassailable, in itself it brings us no closer to understanding the nature of explanatory significance of felt emotion than the attempt to reduce it to its physiological causes. Biological naturalism lacks an account of how emotional verbalizations are linked to the appetitive and aversive contingencies that control them and, therefore, of the role of emotional feelings in learning history and subsequent behaviour. Verbal reports do not stem directly from feelings but are the products of patterns of reinforcement and punishment (Foxall & Greenley, 1999; Foxall & Yani-de-Soriano, 2005). Intentional behaviourism supplies this understanding based on some of the methodological imperatives set by the private nature of emotionality and the methodological limitations posed by the inability to subject emotionality to a direct experimental analysis. The subjective nature of emotional feeling carries further methodological implications. No matter how real an individual’s emotional feelings are to that person, the subjective experience of emotion in others is, for that person, no more than an inference. Emotions themselves, whatever ontological status we assign them, cannot enter directly into experimental analyses and it is therefore impossible to identify exactly their influences on behaviour. A scientist making ascriptions of emotion in order to interpret consumer behaviour must therefore justify these ascriptions by means of empirical findings of systematic relationships between that behaviour and its neurophysiological causes on the one hand and its contextual causes on the other. This requires a means of adding meaning to verbal reports of privately experienced emotion. Biological naturalism also lacks an understanding of why and how explanation in terms of emotional feelings becomes necessary. Although psychologists and philosophers are quick to dismiss behaviourism as a relic of the past which has been superseded by the cognitive revolution, they do not point out what was specifically lacking in its explanations that makes a role for experienced emotion necessary. Finally, there may be something to be gained from the pursuit of an emotion-based explanation of behaviour in the confines of a specific range of behaviour. Rather than plot a way through the myriad of available definitions of emotion, the present context emphasizes two implications of emotion for the
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
67
explanation of behaviour. First, emotion embodies characteristic feelings that are mediated by neurobiological and/or experiential events, but emotion is not to be identified with either neuronal activity or overt behaviour. ‘Insofar as emotion is at bottom sensation, then generating a feeling ipso facto generates a state of consciousness. Thus, an emotion feeling always registers in phenomenal consciousness’ (Izard, 2009: 12). Understanding emotion requires recognition that, except for what each of us takes to be his or her own subjectively felt emotions, emotion is not empirically available for a third-person scientific analysis. What we take, with varying degrees of validation, to be reliable correlates of emotion (e.g., verbal and nonverbal behaviours, neuronal activities) are not the emotions themselves, not even emotion feelings. Although an individual may appear to have emotion feelings of their own and be willing to attribute these also to others on the basis of their similar behaviours and physiologies, that individual cannot adduce these feelings as material for an experimental analysis. In that respect, an attempt to extend the extensional explanation of behaviour by reference to emotion feelings depends on: (i) reconstructing a plausible set of felt emotions for another person – what Dennett calls heterophenomenology – as a means of making the data intelligible, and (ii) recognizing that this is a matter of supplementing rather than replicating scientific analysis. (However, compare Lowenstein, Rick, & Cohen, 2008, for an account of the expanding treatment of emotion in neuroeconomics which posits hot and cold cognition – a standpoint that contrasts markedly with the heterophenomenological approach.) Emotion and intentionality Secondly, in ascribing emotions to others in order to make their behaviour intelligible, we take emotions to be intentional (about or directed towards something other than themselves) and express statements about them in terms of propositional attitudes (Solomon, 1973). They are not to be identified with their physiological or behavioural correlates even though these causal factors are the basis of their formal attribution to others. What part do emotions play in explanation? Specific emotions often accompany specific behaviours: even the most irascible person only throws shoes at the TV when feeling angry with the programmes. But it is not possible to use emotional feelings experimentally in order to demonstrate their necessary or sufficient status in bringing about this behaviour. Since emotional feelings cannot be shown to be causal, how can they enter into explanations of overt behaviour? It is not a matter of showing that they are ontologically causal; they are methodologically necessary because we have to adopt the intentional language of emotion to account for the continuity of behaviour. Having established that some operant behaviour is shaped and maintained by its environmental consequences, we are unable to account for its continuity or discontinuity in operant terms. The only way to proceed (and behaviour analysts do this all the time) is to speak of the organism
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
68
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
in terms of its beliefs and desires (i.e., to adopt the language of intentionality: propositional attitudes). The ontological implications of this are Rylean: it would amount to a category mistake to assume that speaking in terms of beliefs and desires meant that entities of a different order existed over and above the contingencies of reinforcement that are the stock in trade of extensional behavioural science and the afferent–efferent neural links that are part and parcel of extensional neuroscience (Ryle, 1949). But the methodological implications accord more with Dennett (1969): the intentional attributions are a necessary linguistic means of coping with the lacunae found in the extensional accounts. ‘Emotion’ as it is used here is, then, an inference of a potentially felt state that accounts for the continuity of behaviour. No ontological claims are made for this construct: it is ascribed on the basis of neurophysiological and behavioural evidence to provide for gaps in the attribution of behavioural causation based on contingencies of reinforcement and neural functioning. While there is no means of discovering the nature of a subjective emotion or its actual effect on behaviour, out inferences are based on personal experience with respect to the naming and functional attribution of emotions. There remains, nevertheless, a need to ascribe intentionality to actors in order to complete the explanation of their behaviour. Behaviour analysis would do this via the concept of the learning history but this is usually an empirically unavailable entity that itself requires explication at neurophysiological and intentional levels. It is here that emotional ascription can provide an augmented explanation that is consistent with the behaviour analytical framework without being confined to its restrictively extensional language. Since emotions are intentional, they are inferences and the product of linguistic usage; they are not something that, in themselves, can enter into a scientific experiment. The physical and behavioural correlates of emotions, from which emotions are inferred, are not the emotions themselves. The ascription of an emotion to another requires a more sophisticated rationale than an everyday, folk-psychological inference from his or her behaviour. What we are seeking to do is to create an interpretation of a consumer’s observed behaviour: first, by identifying the environmental consequences of that behaviour, reinforcers and punishers that can be shown empirically to be causal elements at the operant level; secondly, by identifying the neural structures and functions that are causal as a result of natural selection; and, thirdly, by using these causal elements to justify an intentional interpretation that supplements them by providing a language in which to describe the aspects of behaviour such as its continuity, discontinuity, and cross-situational consistency that are not amenable to a causal analysis. Emotion and explanation An intentional behaviourist account of emotion makes three demands. The first is a definition and measure of emotionally expressive behaviour. Such
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
69
behaviour might in principle be verbal or nonverbal. However, it is important to separate such behaviour from physiological indices of emotion which might be conflated with other neurological data employed in intentional behaviourism. The requirement is, therefore, for an empirical measure of verbal reactions to felt emotion based on a well-reasoned and research-based typology of human emotionality. A suitable measure would be a psychometrically well-founded instrument for the assessment of fundamental emotional feelings based on verbal responses to settings. The second and third are empirical evidence that neuronal activity and environment–behaviour links are respectively and consistently related to the self-reported behaviour revealed by the psychometric measure. These correlates of self-reported emotion with molar sequences of consumer behaviour provide the rationale for the ascription of emotion feeling. Let us look at these three requirements in turn. Behavioural Expression of Emotion While several typologies of emotion have been employed in consumer research (Havlena & Holbrook, 1986; Havlena, Holbrook, & Lehmann, 1989; Holbrook & Batra, 1987; Huang, 2001; Sherman, Mathur, & Smith, 1997; Sweeney & Wyber, 2002), many investigators have employed the pleasure, arousal, and dominance (PAD) scales devised by Mehrabian and Russell (1974). This approach is especially relevant to the present project in that these authors focus on the physical and social stimuli that influence an individual’s emotional state and behaviours within a specific environment. This environment corresponds to the consumer behaviour setting defined in the BPM as comprising physical and social SD and MO. Mehrabian and Russell (1974) argue on the basis of a thorough literature review (see also Mehrabian, 1980) that pleasure, arousal, and dominance capture the emotion-eliciting qualities of environments and mediate approach–avoidance behaviours such as preference, exploration, affiliation, and work performance. These dimensions of emotional response are measured questionnaire-based self-reports of respondents’ verbal reactions to descriptions of situations elicited by semantic differential scales (Mehrabian & Russell, 1974). Mehrabian (1980) defends the selection of pleasure, arousal, and dominance on the basis of their multi-modal (synesthetic) effects, reports of physiological reactions to such intermodal stimulation, and the findings of work using the semantic differential method of verbal scaling (Osgood, May, & Miron, 1975; Osgood, Suci, & Tannenbaum, 1957) which established evaluation, activity, and potency as the basic dimensions in terms of which the meanings of concepts are delineable. Pleasure–displeasure is a feeling state measured as a continuum ranging from extreme pain or unhappiness to extreme happiness that can be assessed readily with self-report, such as semantic differential measures or with behavioural indicators, such as smiles, laughter, and, in general, positive versus negative facial expressions. Arousal–nonarousal is a feeling state varying along a single
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
70
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
dimension ranging from sleep to frantic excitement which is most directly assessed by verbal report or with behavioural indicators such as vocal activity (positive and negative), facial activity (positive and negative expressions), speech rate, and speech volume. Dominance–submissiveness ranges from extreme feelings of being influenced and controlled by one’s environment to feelings of mastery and control over it; it is also a feeling state that can be assessed from verbal reports using the semantic differential method. It is assumed that there is an inverse relationship between dominance and the judged potency of the environment. Behaviourally, dominance is measured in terms of postural relaxation (i.e., body lean and asymmetrical positioning of the limbs). An individual’s feeling of dominion in a situation is based on the extent to which he or she feels unrestricted or free to act in a variety of ways (Mehrabian & Russell, 1974). Neurobiological Basis of Pleasure, Arousal and Dominance Mehrabian (1980) argues that pleasure, arousal, and dominance are primary adaptations, a contention that entails the identification of these emotions’ neural substrates and relatedness to adaptive behaviours. Accordingly, Barrett, Mesquita, Ochsner, et al. (2007) confirm Mehrabian and Russell’s judgment that these three emotions are fundamental to the mental representation of emotion and relate them to reinforcement and punishment (see also Barrett, 2005; Russell & Barrett, 1999). Foxall (2008a) suggests that Panksepp’s (1998, 2005, 2007) seven core emotional systems – SEEKING, RAGE, FEAR, LUST, CARE, PANIC, PLAY – correspond at a general level to the three emotions adopted by Mehrabian and Russell (1974). (Panskepp employs uppercase letters for these core emotions which represent complex propositional systems in terms of ‘convenient vernacular heuristics’.) The approximation is strengthened if, following Toronchuk and Ellis (2010), PLEASURE and POWER/DOMINANCE are also included as core emotions (Figure 2.4). PLEASURE PLAY RAGE FEAR LUST SEEKING CARE PANIC POWER/DOMINANCE
Pleasure
Arousal
Dominance
Figure 2.4 Panskepp’s (1998) seven core emotional systems, augmented by PLEASURE and POWER/DOMINANCE (after Toronchuk & Ellis, 2010) and related to Mehrabian and Russell’s (1974) tripartite classification of emotions.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
71
Pleasure Barrett, Mesquita, Ochsner, et al. (2007) specifically link ‘core affect’, pleasure–displeasure, with the helpfulness or harmfulness of stimulus events, the likelihood that these outcomes will lead to rewarding or aversive consequences, and their consequent acceptance or rejection (Barrett, Mesquita, Ochsner, et al., 2007: 377). They thereby corroborate the conclusion that these hedonic considerations indicate utilitarian reinforcement, a view expressed also by Panksepp (1998: 112) who argues that such positive feelings indicate to the organism that biologically useful consequences are contingent upon its responses to the stimuli in question: ‘ “Useful” stimuli . . . inform the brain of their potential to restore the body toward homeostatic equilibrium when it has deviated from its biologically dictated “set point” level.’ Affective processes are central to the homeostatic process: the experience of pleasure denotes that equilibrium has been restored (e.g., because hunger or thirst has been assuaged). As an index of material, biological equilibrium, mediated by tangible physical events, and based on the biogenic influences on behaviour that give rise to primary reinforcement, the occurrence of pleasure is consistent with what consumer behaviour analysis understands as utilitarian reinforcement. Pleasure is intrinsically involved in the setting of goals, in evaluating the means available for their achievement, monitoring progress towards them, and judging accomplishment (Politser, 2008). And what can be assumed to follow in the case of primary reinforcers can also be expected to hold for their associated secondary reinforcers. It is encouraging also to note that research on the ‘evolutionary substrates of socio-emotional processes’ provides a bridge between the neurochemical level and the human social and cultural context. The connection between the opioids, in particular, pleasure, and appetitive seeking behaviour strongly supports the current interpretation of the nature of consumer choice and its explanation. While the experience of emotion cannot be reduced to neural substrates (McGinn, 1996; Strawson, 1994), it is possible to identify the neurophysiological correlates of reported mental representation of emotion. In the case of pleasure–displeasure, this involves the amygdala, orbito-frontal cortex (OFC), and the ventromedial prefrontal cortex (vmPFC) (see, inter alia, Rolls, Grabenhorst, & Leonardo, 2009; see also Cardinal, Parkinson, Hall, et al., 2002). In line with this, Wager, Feldman Barrett, Bliss-Moreau, et al. (2008) conclude from a meta-analysis that increased activation in the brainstem areas ventral tegmental area (VTA), and the subcortical telencephalon areas nucleus accumbens (NAC) and portions of the ventral striatum (vStr), all of which are rich in dopamine (DA), is associated with pleasant experiences which also correlate with activity in the hypothalamus (Hy), vmPFC, and right OFC. In contrast, unpleasant experience is associated with activation in amygdala, anteria insula (aINS), periaqueductal gray (PAG), left OFC and more posterior portions of the vStr and ventral globus pallidus (vGP). ‘The results provide a
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
72
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
promising indication that different gross anatomical areas may be differentially sensitive to pleasant and unpleasant stimuli, although they do not imply that activation in any of these regions is uniquely associated with either category’ (Wager, Feldman Barrett, Bliss-Moreau, et al., 2008: 259). These areas, associated closely with pleasure–displeasure, form a brain region ‘that is involved in establishing the threat or reward value of a stimulus’ (Barrett, Mesquita, Ochsner, et al., 2007: 382). This is related to representations of core affect that has been the consequence of prior behaviour with respect to the stimulus in question (i.e., it is a representation of learning history). Arousal The elements of emotion, other than pleasure, that Mehrabian and Russell identify as central – namely, arousal and dominance – suggest the content of core emotion (the propositional component of its intentionality), of which Barrett, Mesquita, Ochsner, et al. (2007) speak in terms of arousal-based content, relational content, and situational content. Arousal signifies activeness and is manifest in reports of being active, attentive, wound-up; while unarousal manifests in stillness, reported as being still, quiet, sleepy. This active-versus-still dimension corresponds with Mehrabian and Russell’s arousal and incorporates the affective effects of informational reinforcement. Relational content is social and reflects the degree of domination or submissiveness that is felt in the presence of others. Situational content reflects the extent to which a setting is novel or unexpected, conducive to or obstructive of a goal, compatible or not with norms and values, and confers responsibility or agency. These considerations call forth maintenance of or change in the individual’s behavioural stance, his or her readiness for action. These content-conferring dimensions of emotion are environmentally determined and Barrett, Mesquita, Ochsner, et al. (2007) point out that it may not be so easy to point to specific neural substrates in their cases as for core affect. Nevertheless, we can make some suggestions. The capacity of humans to consistently and communicatively make the mental state attributions on which depictions of emotional content are based is indicative of both biological and cultural mechanisms for this function. The ascription of content to one’s core affect is ostensibly a matter of interpreting subjective experiences of emotion which process includes the appraisal of the social and physical situations in which the individual find’s him or herself. Barrett and her colleagues proceed from these considerations to the observation that experienced emotion correlates with activation of medial prefrontal cortex (MPFC) and the anterior cingulate cortex (ACC). These and other results confirm the case for the attribution of content to experienced emotion based on neurophysiological reasoning (see also Coull, 1998; Lewis, Critchley, Rothstein, et al., 2007; Paul, Lautzenhiser, Brown, et al., 2006; Rauch, Shin, Dougherty, et al., 1999). The concern here is with an interaction of emotional and cognitive functions to
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
73
produce content attributions appropriate to the situation in which the person and his or her behaviour are located. Barrett and her colleagues draw attention also to the operation of the ventrolateral prefrontal cortex (VLPFC) in generating the selection and inhibition of responses, and working memory; these are in turn implicated in the retrieval and integration of cognitive elements of memory processing such as the interpretation and evaluation of conceptual knowledge. Specifically with respect to arousal and dominance, we can make some general observations on the likely brain areas and physiological functions associated with the ascription of appropriate content to core affect. Neuroscientists’ frequent definition of arousal in terms of anxiety, fear, and anger chimes with Mehrabian’s (1980) understanding of the term. Environments that signal or encourage competitiveness and uncertainty as to the outcomes of behaviour are most likely to engender arousal. Sex differences in arousal-seeking tendency have evolutionary support and, expressed in terms of aggression vs. inhibition and flight vs. fight, manifest well-established sex differences in behaviour (Campbell, 2007; Taylor, Klein, Lewis, et al., 2000). Hormones such as oxytocin and testosterone have a role in the regulation of fear and aggression and nurturance and affiliation. Among the neurotransmitters, serotonin is associated not (as in popular imagination) with pleasure but with the reduction of anxiety. Reduced functioning of CNS serotonin underlies impaired impulse control and is further implicated in violence, impatience, and the assumption of risks of punishment or injury (Higley, Mehlman, Poland, et al., 1996). The administration of serotonin, in contrast, as in selective serotonin reuptake inhibitor (SSRI) medication, modulates antisocial tendencies (Knutson, Wolkowitz, Cole, et al., 1998). These are behaviours closely connected with feelings of arousal and, although they are too extreme to find a place in most consumption activities and research, they are indicative of a role for arousal at all points along the consumer continuum. Arousal and impulsivity are clearly seen in everyday consumer behaviours such as innovativeness, novelty-seeking, and unplanned purchasing; compulsiveness is at the root of unregulated consumption and addiction (Foxall, in press). Finally, while dopamine has a general role in the anticipation of rewarded behaviour, it has a particular affinity with behaviour that eventuates in (reported) arousal. Dopamine is associated with feelings of excitement, engagement, and the involved pursuit of primary reinforcers; it is responsible for the energization of higher areas of the motor cortex that is essential to SEEKING (Panksepp, 1998: 144). The biochemical bases of behavioural responses link also to the arousalseeking tendencies evoked by environments that provide varying information rates (Mehrabian & Russell, 1974), that is, the rate at which stimuli impinge upon the individual to create excitatory or inhibitory reactions. This is the basis of feelings of arousal, some of which reflect environmental dynamism which in turn can and often does provide the individual with performance feedback on the behaviour that he or she is enacting (Foxall, 2005).
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
74
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
The role of dopamine in pleasure and arousal is subtle. Berridge and Robinson (1998, 2003) do not confine their analysis of reward to learning processes in which the consequences of action are associated with the stimulus context in which behaviour has taken place. Rather, they call attention to both a hedonic or affective element (that denotes ‘liking’ or pleasure), and a motivational element (reflecting ‘wanting’ or incentive salience). ‘Liking’ is associated with opioid transmission on to GABAergic neurons in the nucleus accumbens (Winkielman, Berridge, & Wilbarger, 2005). ‘Wanting’ or incentive salience, the motivational element in reward, is a separate process, more likely associated with dopamine release and retention. Hence, in contradistinction to some early views, dopamine is not an instigator of pleasure: it is in fact neither necessary nor sufficient for ‘liking.’ Manipulation of the dopamine system does, however, change motivated behaviour by increasing instrumental responses and the consumption of rewards; incentive salience is a motivational rather than an affective component of reward that transforms neutral stimuli into compelling incentives (see also Berridge, 2004; Robinson & Berridge, 2003). In line with Berridge’s (2000) argument that liking and wanting should be separated, Toronchuk and Ellis (2010) contrast PLEASURE which is relevant to consummatory behaviours and associated with opioid and GABA release, and Panskepp’s (1998) SEEKING which is associated with dopamine release and which marks appetitive responses. This dichotomy is well-accommodated to the distinction drawn here since the wanting which is inherent in SEEKING is indicated by arousal rather than pleasure. Toronchuk and Ellis also note that social dominance is related to hedonic experience (pleasure–displeasure), which leads to the final dimension of emotion considered central by Mehrabian and Russell (1974). Dominance The term ‘dominance’ is frequently employed in social psychology to refer to interpersonal control, a behaviour known to be context-specific (Sulloway, 2007). The dominance to which Mehrabian and Russell (1974) drew attention is an emotional response to both physical and social environments and differs with the extent to which the consumer setting permits autonomy or induces conformity. This latter definition includes but is broader than the former. Barrett, Mesquita, Ochsner, et al. (2007) include dominance in their model of emotion, associating it with the autonomy that is an American value and contrasting it with the submissiveness and harmoniousness that are often thought to be more highly valued in eastern societies such as Japan (2007: 380). At the social psychological level, dominance undoubtedly is connected with some of the traits we have considered to constitute arousal. Dopamine and opioids are associated with sociability, prosocial behaviour, and affiliation; the neuropeptide oxytocin increases feelings of trust. Moreover, the reward-cantered
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
75
dopamine promotes social preferences; but the attractiveness of psychostimulants like cocaine increases with dopamine deficit (Panksepp, 2007: 156). Hence, submissiveness rather than dominance is associated with cocaine consumption. Both the neurotransmitter serotonin and the hormone testosterone are associated with high social status and feelings of dominance; challenges to status and increased stress are associated with higher cortisol levels (Buss, 2004, 2005; Cummins, 2005; Foxall, 2007b). The relationship between dominance and the BPM resides in a tendency of consumers to report high levels of this emotional response in more open settings, those that provide a wider range of behavioural outcomes, and which are usually under the control of the consumer him or herself rather than another agent such as a marketer or government department. In a study that addressed the neurophysiology basis of pleasure, arousal, and dominance in the context of consumer behaviour, Morris, Klahr, Shen, et al. (2008) sought to relate brain regions to PAD in the context of exposure to a TV commercial (see also Bernat, Patrick, Benning, et al., 2006). Their study revealed bilateral activations in the inferior frontal gyri and middle temporal gyri for pleasure, activations on the right superior temporal gyrus for arousal. No significant differences were found for dominance which the authors argue is unsurprising since dominance accounts for much less variance in emotional responses than pleasure or arousal ‘and is often not a factor in vicarious experiences such as watching a television commercial’ (Bernat, Patrick, Benning, et al., 2006: 795). However, after positively reviewing the evidence for a model of emotionality that includes dominance as well as pleasure and arousal, Demaree, Everhart, Youngstrom, et al. (2005) propose that ‘the lateralization of emotional “dominance” be explored with the hypothesis that relative left- and right-frontal activation would be associated with feelings of dominance and submissiveness, respectively’ (2005: 3). Figure 2.5 summarizes the relationships among core emotions, their putative neurophysiological bases, and the patterns of contingency with which they are associated as defined by the BPM.
Linking Emotion and Contingency The importance of emotion–contingency links That emotion is consistently related to contingencies of reinforcement and punishment is a recurring theme in modern theories including Damasio’s (1994) somatic marker hypothesis, Ledoux’s (1998, 2000) emotional brain perspective, and Rolls’s (1999) neurophysiological approach. But it is the last of these that makes the relationship most explicit and systematic. Rolls (1999) integrates classical (or respondent) and operant conditioning, and links the acquisition of behaviour and the generation of emotion to genetic factors, by proposing that emotions are states elicited by the reinforcing stimuli that condition instrumental behaviour. Genes specify not individual behaviours but the
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
76
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY BEHAVIOR SETTING SCOPE
Closed
Open
CC2
ACCOMPLISHMENT
CC1
PLEASURE AROUSAL DOMINANCE
PLEASURE AROUSAL dominance CC4
HEDONISM
PLEASURE arousal dominance
ACCUMULATION
pleasure AROUSAL dominance
CC3
PLEASURE arousal DOMINANCE
CC6
MAINTENANCE
pleasure arousal dominance
CC5
pleasure AROUSAL DOMINANCE
CC8
pleasure arousal DOMINANCE
CC7
Figure 2.5 Contingencies and emotions: research hypotheses and summary of findings. Studies show that: (i) pleasure scores for contingency categories (CCs) 1, 2, 3, and 4 each exceed those of CCs 5, 6, 7, and 8; (ii) arousal scores for CCs 1, 2, 5, and 6 each exceed those of CCs 3, 4, 7, and 8; (iii) dominance scores for CCs 1, 3, 5, and 7 each exceed those for CCs 2, 4, 6, and 8. Moreover, (iv) approach–avoidance (aminusa) scores for CCs 1, 2, 3, and 4 each exceed those for CCs 5, 6, 7, and 8; and (v) approach–avoidance (aminusa) scores for CCs 1 and 3 each exceed those for CCs 2, 4, 5, 6, 7, and 8. (For explication, see text and Foxall, Yani-de-Soriano, Yousafzai, et al., 2010).
goals that will serve as reinforcers (i.e. influence the performance of operant responses), those whose rate of repetition is influenced by their consequences. The process is even more flexible than this, however, since an operant is not a single response but a class of behaviours that lead to and are selected and maintained by a common set of reinforcers or goals. This proposition is derived from a broader biological purview of emotionality and behaviour based on evolutionary considerations. The adaptive propensity of genes stems not from their leading to the elicitation of fixed action patterns but from the specification of goals (related to the enhancement of survival and fitness) towards which many behaviours will be directed and in the process selected (Rolls, 2005). Genes
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
77
do not determine particular behaviours but something rather more general: the goals or consequences that stem from a number of functionally (although not topographically) equivalent responses each of which is a flexible adaptation to an appropriate set of environmental contingencies of reinforcement and punishment. Gene-specified primary reinforcers can in this process become arbitrarily related to a wide range of secondary reinforcers, producing patterns of contingency that are the proximal determinants of which operant responses are selected in particular circumstances. The ultimate agent of selection is biological, however: the emotional experience evoked by the contingencies of reinforcement and punishment. The relatively small number of specifications required at the genetic level by Rolls’s (2005) theory is balanced in the lifetime of an organism by the learning of secondary reinforcers; instead of encoding an incalculable number of stimulus–response associations, the genes perform the restricted function of encoding relationships between goals and the variety of means by which they may be achieved. What is crucial is that the genes specify goals (behaviour–contingent consequences) which elicit particular emotions. It is the relationship between these emotions and the selection of behaviour that is of primary interest to us here. Hence, the conclusion that ‘. . .the heritability of behaviour is best understood as the heritability of reinforcers’ and, therefore, of the emotions they elicit; the outcome is an adaptive construal of emotion which has implications for the genetic specification of synaptic structure and function (Rolls, 2005: 61). It is necessary to note here for later reference, the sort of mechanism that Rolls proposes for this example of genetic specification. It requires that synapses be specified across the nervous system from the point of sensory input to whatever region(s) of the brain at which the reinforcing or punishing value of the reinforcing stimulus (goal) is explicitly represented (Rolls, 2005: 61). He proposes neural activity in the amygdala and OFC as defining the site(s) where such explicit representation may be located, and exemplifies the process in terms of sweet taste receptors on the tongue being connected to neurons that specify food reward, the actions of which are modulated by hunger signals (Rolls, 2005: 61; 2006). Rolls (1999, 2005) proposes that increases in intensity of positive reinforcement are associated with increases in pleasure leading on to elation and ecstasy while the increasing intensity of aversive stimuli (the reception of which punish behaviour) is associated with increasing apprehension leading on to fear and terror. When, as in the processes of behavioural extinction and time out, respectively, a reinforcer is omitted or terminated, the emotional outcome is sadness and frustration leading on to anger and grief and ultimately rage. However, if an aversive consequence is omitted or terminated, as in the processes of avoidance and escape, respectively, the emotion felt is increasing relief. For all its logical insight, however, Roll’s classification of emotions in relation to contingencies of reinforcement appears to reflect more the a priori reasoning of a generally informed account of emotion and behaviour rather than a framework based on empirical research specifically tailored to the study of
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
78
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
economic behaviour with its dual sources of reinforcement and its sensitivity to the scope of consumer behaviour settings. PAD and consumer behaviour Systematic relationships between the emotional variables identified by Mehrabian and Russell (1974) and consumer behaviour emerge from a number of studies. Pleasure, arousal, and dominance mediate more overt consumer behaviour such as desire to affiliate with others in the setting, desire to stay in or escape from the setting, and willingness to spend money and consume (Donovan & Rossiter, 1982; Donovan, Rossiter, Marcoolyn, et al., 1994; Foxall, 1997d; Foxall & Yani-de-Soriano, 2005; Gilboa & Rafaeli, 2003; Groeppel-Klein, 2005; Mehrabian, 1979; Mehrabian & de Wetter, 1987; Mehrabian & Riccioni, 1986; Mehrabian & Russell, 1975; Russell & Mehrabian, 1976, 1978; Tai & Fung, 1997; Van Kenhove & Desrumaux, 1997). Donovan and Rossiter (1982) used the Mehrabian–Russell (1974) model in retail settings, focusing on store atmosphere, that is, within-store variables that influence shopping behaviour (Kotler, 1973). They report pleasure to be a very powerful determinant of approach–avoidance behaviours within the retail environment, including the tendency of the consumer to spend beyond his or her original expectations. Arousal similarly increased time spent on in-store browsing and exploring and willingness to interact with sales personnel. Bright lighting and upbeat music were noted as stimuli that induce arousal. Dominance, however, exerted little influence on consumer behaviour, to the extent that a second study (Donovan, Rossiter, Marcoolyn, et al., 1994) incorporated only the pleasure and arousal of shoppers during the shopping experience. The results confirmed that pleasure predicted consumer behaviour such as extra time spent in the store and overspending but arousal did not predict overspending, failing to support the earlier research, but arousal predicted underspending in unpleasant store environments, something not revealed by the earlier study. These and other studies have focused mainly on understanding the role of pleasure and arousal in shaping consumer choice, while the role of dominance has been ignored or controlled (Biggers, 1981; Biggers & Pryor, 1982; Biggers & Rankins, 1983; Groeppel-Klein, 2005;Yani-de-Soriano & Foxall, 2006). Moreover, in research involving the BPM dominance has proved to discriminate very effectively between open and closed CBSs. The identification of differences in dominance among consumer settings requires a model capable of generating consumer situations that discriminate adequately between varying levels of situational inducement (Foxall, 1990). Several investigations employing the BPM reveal that the influence of dominance on approach–avoidance behaviour is more clearly apparent from an array of consumer situations that distinguish open from closed CBSs than by the use of haphazardly developed consumer situations (Foxall, 1997d, 2005). Discriminant analyses showed the
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
79
significantly high power of the dominance dimension in discriminating between open and closed settings (Foxall & Greenley, 1999, 2000). Similar results were obtained in subsequent studies in Venezuela (Foxall & Yani-de-Soriano, 2005). The BPM basis of emotion–contingency links The wording of the items that Mehrabian and Russell (1974) employ in their scales to measure pleasure, arousal, and dominance, plus the neurobiological considerations described above, suggest that these emotions can each be associated with particular structural variables of the BPM (Seva, Duh, & Helander, 2010). By emphasizing the satisfaction and pleasure consumers are likely to receive from certain purchase and consumption environments, plus the usefulness of this emotion in indicating the presence of positively reinforcing behavioural outcomes that will eventuate in enhanced survival potential and biological fitness, pleasure is redolent of utilitarian reinforcement. Pleasure, therefore, should be expected to be a disproportionately encountered response to consumer situations that result in utilitarian (functional) consequences. This does not mean that it will be exclusively the emotional response reported for such consumer situations: each consumer situation is defined in terms of relative contributions from utilitarian and informational reinforcement and CBS scope. But situations that are defined as relatively high in utilitarian reinforcement should produce relatively high pleasure reactions. In as much as it is defined primarily in terms of performance feedback, informational reinforcement can be expected to co-occur with reactions of arousal which is often associated with the identification of discrepancies between current and indicated performance. To illustrate, road markings that indicate one’s excess speed, for instance, are presumably intended to embody such feedback and to engender perceptual responses that bring drivers’ behaviour into line with socially accepted norms. Mehrabian and Russell (1974) define a measure of the information rate of environments, scores on which correlate significantly with arousal scores; this idea of information rate provides an indication of the variation from norms present in environmental stimuli such as noise, crowdedness, temperature, lighting, and social interaction rates and thus includes the criteria of informational reinforcement without being coterminous with it. The expectation, therefore, is that consumer situations that are relatively high in informational reinforcement will be those for which reports of arousal are also disproportionately high. Open CBSs are by definition those in which the consumer has relatively large numbers of behavioural options (sources of reinforcement) and hence those in which he or she can be predicted to experience a higher degree of dominance than in closed settings in which behavioural choices are severely limited perhaps to a single option. In the former, the consumer is more likely to be a determinant of the range of options available (through choice of store, time of shopping, lack of external compulsion) whereas in the latter agents other
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
80
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
than the consumer are likely to arrange these matters. It is probable, therefore, that consumer situations that are relatively open will engender higher feelings of dominance than those that are relatively closed. Eight studies have shown strong consistency of results in showing that pleasure scores are higher for consumer situations marked by higher utilitarian reinforcement; arousal scores, for consumer situations marked by higher informational reinforcement; and dominance scores, for situations marked by greater openness (Foxall, 1997a,b; Foxall & Yani-de-Soriano, 2005; Yani-deSoriano, Foxall, & Newman, in press). The hypothesized patterns of emotional reaction to consumer situations defined in terms of the structural variables of the BPM shown in Figure 2.5 are also those identified by these empirical studies (Foxall, Yani-de-Soriano, Yousafzai, et al., 2010).
BEHAVIOURAL CONTINUITY REVISITED The aim of intentional behaviourism is to ascribe intentionality (in this case emotionality) on the basis of evolutionarily consistent neuronal functioning and molar behaviour–environment relationships. The language of intentionality has become necessary in order to explain aspects of behaviour (the present account concentrates on its continuity or discontinuity) that cannot be explained solely in extensional terms. Although the extensional account identifies environmental stimuli that normally allow the behaviour in question to be predicted and controlled, it is not possible to locate stimuli that account for changes in behaviour or persistence when the contingencies of reinforcement have changed. It is therefore impossible to construct a consumer behaviour setting to aid the explanation of the behaviour because the SD and MO cannot be ascertained. It is clear from the examples presented above (the experimental participant who is insensitive to changing schedule parameters, the addict who changes behaviour pattern, and the consumer who expands the choice set of brands considered for purchase) that some discriminative, motivating, and/or reinforcing stimuli cannot be located. An extensional account requires, moreover, that these stimuli be incorporated into a learning history that accounts for the observed pattern of behaviour in the past. While learning histories may be easily constructed for laboratory animals, it is a different matter to find one for the average consumer wandering around a large departmental store. Not only is such a learning history not immediately available to the consumer researcher: it is never likely to be empirically available in nonintentional terms. That is, the attempt to reconstruct a learning history for a consumer entails verbal interactions between that consumer and an investigator which inevitably means that it is expressed intentionally. One answer to this problem is to assume that a learning history that explains current behaviour is somewhere available and that until it can
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
81
be specified scientifically it must be assumed to effect some kind of ‘actionat-a-distance’ on behaviour (Baum, 2007; Baum & Heath, 1982). However, to speak as though a learning history were available in the face of an inability to specify it with the precision required of a scientific account is misleading to say the least. The point at which the use of intentional language becomes inevitable is that at which an extensional account has become unfeasible. The use of the language of intentionality demarcates this impasse (Foxall, 2008b, 2009). Why does finding afferent–efferent and environment–behaviour links justify the use of intentionality? In the case of emotion, we cannot investigate the subjective experience of a consumer and relate this to his or her behaviour simply because the subjective experience is not publicly available. We are assuming, however, on the basis of Rolls’s kind of reasoning, that felt emotion is a necessary link in a sequence from reinforced behaviour to learning history. That is, the emotion is the result of reinforcing stimuli evoking emotion feelings by respondent conditioning and that these feelings act as rewards that are connected in some way with the continuity of behaviour since they make the re-enactment of the behaviour more probable. We cannot reconstruct the processes whereby respondent conditioning produces emotion feelings in the individual, but we can find general criteria for the ascription of emotion feelings based on the general operation of biological and environmental procedures. That is, we can isolate the afferent–efferent and environment–behaviour links that are associated with people’s verbal (and feasible nonverbal) expressions of specific emotions. From these we can attribute specific emotions, the content of which is judged from the verbal responses, to individuals in order to explain their behaviour in the absence of specific stimuli to which to attribute it. The use of emotion-related words and facial expressions are among the means by which emotion feelings are indexed in experiments relating emotion to behaviour (e.g., in pain research; see, inter alia, Flor, Knost, & Birbaumer, 2002). It is clear from the above review of the neural correlates of pleasure, arousal, and dominance that neuroanatomy and neurophysiology are closely related to the verbal and nonverbal expression of emotion feelings. And it is clear from the review of environment–behaviour relationships also reviewed that patterns of contingency are systematically related to verbal expressions of emotion feelings. It is the establishment of these relationships by the procedures of extensional sciences that gives confidence that the ascription of intentionality (in lieu of a learning history) is a justified part of the explanation of behaviour. This is a different kind of justification for the use of intentional language than that presented by Dennett (1969) who relies on the pre-emptive assumption of intentionality at the level of the neuron. In contrast, intentional behaviourism ascribes intentionality only at the personal level of explanation and only on the basis of the evidence of extensional neuroscience and behavioural science.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
82
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
CONCLUSIONS One strategy for coping with the limitations of an extensional account of such aspects of behaviour as its continuity is to manufacture a fictitious learning history, which, however plausible, cannot be subjected to a scientific analysis and is never likely to become available to the empirical investigator. An alternative is to recognize that a different way of talking about the causes of such behaviour is required and to substitute the beliefs and desires that can be ascribed to the consumer. If this ascription is to carry reasonable legitimacy, it must be founded on inferences that can be reasonably made from the extensional sciences which reflect evolutionary logic. The use of language relating to emotions, for instance, must be securely founded upon the operation of a nervous system that has evolved to enable the organism to solve the kinds of problem posed by the behaviour, a problem that is related to its individual survival and biological fitness. The means of solving problems of obtaining food and other primary reinforcers are assumed to be hard-wired into the genetic composition of the consumer which specifies the goals necessary to optimize fitness. A second source of afferent–efferent support for the ascription of emotionality derives from the neuronal plasticity that is the result of behaviour that secures secondary reinforcers which contribute to the individual’s survival and fitness. In both cases, the operant processes that eventuate in obtaining the reinforcer are the prelude to classical or respondent conditioning which engenders emotional feeling. Pleasure, arousal, and dominance are subjective emotional feelings that have been shown to have a role in the phylogenetic history of the species and in the ontogentic learning of the individual. They are, therefore, indicated as outcomes of the processes of selection by consequences that inhere in natural and environmental selection. This seems a far surer basis for the linguistic ascription of at least a quasicausal mechanism to account for behavioural continuity than the imagination of a learning history. By incorporating intentional language, by speaking of quasi-causation in the case of intentional inferences, this approach is also intellectually honest in its demarcation of a sphere of behaviour that cannot be accounted for by extensional science alone. It makes clear, for instance, that the linguistic entities to which appeal is made in order to explain behaviour are not of the kind that are amenable to an experimental analysis, that are not empirically available, not likely to become so. The faux-extensional approach that involves the fabrication of a learning history to account for continuity is subject to all of these limitations but its vulnerability is not obvious given the language in which it is expressed. The intentional terms are not inferred from the very behaviour they are said to explain (a practice that has marked social science at some stages of its development); rather they are founded upon the very extensional considerations that an entirely scientific account would embrace were the evidence available. By saying that the experimental subject believes that the contingencies
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
83
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
operate in a particular way and pitches his or her behaviour accordingly, that the addict has decided to change his or her behaviour, that the buyer of consumer nondurables desires an alternative source of product category benefits, we are not providing the sort of evidence that the establishment of causal relationships via experimentation delivers. But we are acknowledging a gap in our extensionally expressible knowledge and indicating that we can fill it only by the adoption of intentionality. By surmising that felt emotions produced via operant and respondent processes contribute to the consumer’s learning history by providing a priming mechanism for the evaluation of stimuli that constitute the CBS, we are linking the intentional explanation proposed with the radical behaviourist approach which provides the general framework for investigation. Figure 2.6 summarizes the development of the BPM framework in view of the relationships between reinforcement contingencies and emotion feeling that have been explored. Rolls’s (1999) argument that emotions are states elicited by reinforcers is made explicit, and the classification of consumer behaviours and patterns of emotion identified in empirical research are drawn Operant conditioning
CONSUMER SITUATION SD LH MO
R
Respondent conditioning
SR/A Pattern of reinforcement
Operant class
P, A, D Emotional pattern
Cl
Op
Accomp P+ A+ D- P+ A+ D+ low low
UR
Maint
high
IR high
Accum
Hed P+ A- D-
P+ A- D+
Accum P- A+ D-
P- A+ D+
Maint P- A- D-
P- A- D+
Hed
Accomp
Figure 2.6 Behavioural contingencies and patterns of emotion. A, arousal; Accomp, Accomplishment; Accum, Accumulation; Cl, closed consumer behaviour setting; D, dominance; Hed, Hedonism; LH, learning history; Maint, Maintenance; MO, motivating operations; Op, open consumer behaviour setting; P, pleasure; R, response; SD , discriminative stimuli; SR/A , reinforcing and aversive consequences.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
84
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
out for further study. The next stage in the research programme, suggested by the dotted arrow, aims to clarify the contribution of the felt emotions of pleasure, arousal, and dominance to learning history. This is likely to rest upon the identification of brain regions such as the amygdala which are related not only to emotional processing but to behavioural reinforcement and memory. Further research is also directed towards understanding better the relationship between emotionality and propositional knowledge. Such enquiry recognizes that emotional events are closely allied with cognitive processing. The fact that emotion, cognition, and approach–avoidance dispositions seem to occupy identical brain regions in terms of their physiological grounding (Barrett, Mesquita, Ochsner, et al., 2007; Gray & Braver, 2002; Phelps, 2005, 2006; Spielberg, Stewart, Levin, et al., 2008) suggests that the attribution of propositional attitudes to consumers is closely related to the emotional experiences that we ascribe to them (see also Whalen & Phelps, 2009). Emotion and cognition may be separate only phenomenologically, not causally; they may not exist as independent systems since ‘the separation does not seem to be respected by the brain’ (Barrett, Mesquita, Ochsner, et al., 2007: 390). Brain structures involved in the neural circuitry for emotion such as the amygdala are also implicated in such cognitive activities as the allocation of attention, perceptual processing, and memory. Moreover, cognitively embedded brain structures such as dorsomedial prefrontal cortex and VLPFC are involved in the expression of emotion; decision processes with respect to moral reasoning are founded in core emotion; and economic decisions are emotionally coloured (Hodgkinson & Healey, 2008, in press; Lowenstein, Rick, & Cohen, 2008). Finally, it is important to recognize that in our role as psychological theorists, we are not here doing the work of the neuroscientist or behavioural scientist. We are using their work to show that in terms of current canons of scientific practice within those extensional sciences, it is possible to identify a sufficiently convincing empirical relationship between the dependent variable, behaviour, and the independent environmental and biological variables of which it is a function. Science is never static; in time, new results will clarify and perhaps even make redundant the relationships we have posited; new paradigms will alter the theoretical bases of sciences and the extensional understanding of the findings they generate. The extensional explanations will change and the intentional interpretation of behaviour will alter accordingly. We can only work where we are. We accept that behaviour is a function directly and proximally of neuronal processing and directly, indirectly, and distally of patterns of environmental contingency. Even here, however, we are not attempting qua psychological theorists to be at the leading edge of either neuroscience or behavioural science. But we acknowledge the imperative of supplementing the extensional language of these sciences and seek a logical rationale for resorting to the intentional.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
85
ACKNOWLEDGEMENTS The author is grateful to Dr Helen Hancocks, Professor Gerard Hodgkinson and Professor Leonard Minkes for helpful comments on earlier drafts.
REFERENCES Barrett, L.F. (2005). Feeling is perceiving: Core affect and conceptualization in the experience of emotion. In L.F. Barrett, P.M. Niedenthal, & P. Winkielman (Eds), Emotion and Consciousness (pp. 255–84.) New York: Guilford. Barrett, L.F., Niedenthal, P.M., & Winkielman, P. (Eds) (2005). Emotion and Consciousness. New York: Guilford. Barrett, L.F., Mesquita, B., Ochsner, K.N., & Gross, J.J. (2007). The experience of emotion. Annual Review of Psychology, 38, 173–401. Baum, W.M. (2007). Commentary on Foxall ‘Intentional Behaviourism’. Behaviour and Philosophy, 35, 57–60. Baum, W.M., & Heath, J.L. (1982). Behavioural explanations and intentional explanations in psychology. American Psychologist, 47, 1312–7. Bennett, M.R., & Hacker, P.M.S. (2003). Philosophical Foundations of Neuroscience. Oxford: Blackwell. Bernat, E., Patrick, C.J., Benning, S.D., & Tellegen, A. (2006). Effects of picture content and intensity on affective physiological response. Psychophysiology, 43, 93–103. Berridge, K.C. (2000). Reward learning: Reinforcement, incentives, and expectations. In D.L. Medin (Ed.), The Psychology of Learning and Motivation, 49 (pp. 223–78). San Diego: Academic Press. Berridge, K.C. (2004). Motivation concepts in behavioural neuroscience. Physiology and Behaviour, 81, 179–209. Berridge, K.C., & Robinson, T.E. (1998). What is the role of dopamine in reward: Hedonic impact, reward learning, or incentive salience? Brain Research Reviews, 28, 309–69. Berridge, K.C., & Robinson, T.E. (2003). Parsing reward. Trends in Neuroscience, 26, 507–12. Biggers, T. (1981). The function of dominance-submissiveness within Mehrabian’s theory of emotion. PhD thesis. Florida State University. Biggers, T., & Pryor, B. (1982). Attitude change: A function of emotion-eliciting qualities of environment. Personality and Social Psychology Bulletin, 8, 94–9. Biggers, T., & Rankins, O.E. (1983). Dominance-submissiveness as an effective response to situations and as a predictor of approach-avoidance. Social Behavior and Personality, 11(2), 61–9. Brentano, F. (1874). Psychology from an Empirical Standpoint. Leipsig: Meiner. Buss, D.M. (2004). Evolutionary Psychology. Boston, MA: Pearson. Buss, D.M. (Ed.) (2005). The Handbook of Evolutionary Psychology. Hoboken, NJ: Wiley. Campbell, A. (2007). Sex differences in aggression. In R.I.M. Dunbar & L. Barrett (Eds), The Oxford Handbook of Evolutionary Psychology (pp. 365–82). Oxford: Oxford University Press. Cardinal, R.N., Parkinson, J.A., Hall, J., & Everitt, B.J. (2002). Emotion and motivation: the role of the amygdala, central striatum, and prefrontal cortex. Neuroscience and Biobehavioral Reviews, 26, 321–52.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
86
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Chisholm, R. (1957). Perceiving: A Philosophical Study. Ithaca, NY: Cornell University Press. Coull, J.T. (1998). Neural correlates of attention and arousal: Insights from electrophysiology, functional neuroimaging and psychopharmacology. Progress in Neurobiology, 55, 343–61. Cummins, D.D. (2005). Dominance, status, and social hierarchies. In D.M. Buss (Ed.), The Handbook of Evolutionary Psychology (pp. 676–97). Hoboken, NJ: Wiley. Curry, B., Foxall, G.R., & Sigurdsson, V. (2010). On the tautology of the matching law in consumer behavior analysis, Behavioural Processes, 84, 390–9. Damasio, A. (1994). Descarte’s Error. New York: Putnam. Demaree, H.A., Everhart, D.E., Youngstrom, E.A., & Harrison, D.W. (2005). Brain lateralization of emotional processing: Historical roots and a future incorporating ‘dominance’. Behavioral and Cognitive Neuroscience Reviews, 4, 3–20. Dennett, D.C. (1969). Content and Consciousness. London: Routledge and Kegan Paul. Dennett, D.C. (1978a). Brainstorms: Philosophical Essays on Mind and Psychology. Montgomery, VT: Bradford Books. Dennett, D.C. (1978b). Why you can’t make a computer that feels pain. In D.C. Dennett (Ed.), Brainstorms: Philosophical Essays on Mind and Psychology (pp. 190– 229). Montgomery, VT: Bradford Books. Dennett, D.C. (1987). The Intentional Stance. Cambridge, MA: MIT Press. Donovan, R.J. & Rossiter, J.R. (1982). Store atmosphere: An environmental psychology approach. Journal of Retailing, 58(1), 34–57. Donovan, R.J., Rossiter, J.R., Marcoolyn, G., & Nesale, A. (1994). Store atmosphere and purchasing behavior. Journal of Retailing, 70(3), 283–94. Ehrenberg, A.S.C. (1972). Repeat Buying. Amsterdam: North Holland. Fagerstrom, A., Foxall G.R. & Arntzen, E. (2010). Implications of motivating operations for the functional analysis of consumer behavior. Journal of Organizational Behavior Management, 30, 110–26. Feyerabend, P. (1975). Against Method. London: Verso. Flor, H., Knost, B., & Birbaumer, N. (2002). The role of operant conditioning in chronic pain: An experimental investigation. Pain, 95, 111–8. Foxall, G.R. (1988). The behavioural analysis of consumer choice: Issues for theory and research. Proceedings of the European Institute for Advanced Studies in Management Conference ‘New Challenges in Management Research’, Leuven, The Netherlands. (Reprinted in Foxall, 1996, Chapter 5.) Foxall, G.R. (1990). Consumer Psychology in Behavioral Perspective. London and New York: Routledge. (Reprinted 2004 by Beard Books, Frederick, MD). Foxall, G.R. (1996). Consumers in Context: The BPM Research Program. London and New York: Routledge. Foxall, G.R. (1997a). Explaining consumer behaviour: From social cognition to environmental control. In C.L. Cooper & I.T. Robertson (Eds), International Review of Industrial and Organizational Psychology (volume 12) (pp. 229–87). Chichester, UK: Wiley. Foxall, G.R. (1997b). Marketing Psychology. London and New York: Macmillan. Foxall, G.R. (1997c). Affective responses to consumer situations. International Review of Retail, Distribution and Consumer Research, 7, 191–225. Foxall, G.R. (1997d). The emotional texture of consumer environments: A systematic approach to atmospherics. Journal of Economic Psychology, 18, 505–23. Foxall, G.R. (1999). The substitutability of brands. Managerial and Decision Economics, 20, 241–57. Foxall, G.R. (2004). Context and Cognition: Interpreting Complex Behavior. Reno, NV: Context Press.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
87
Foxall, G.R. (2005). Understanding Consumer Choice. London and New York: Palgrave Macmillan. Foxall, G. R. (2007a). Explaining Consumer Choice. New York and London: Palgrave Macmillan. Foxall, G.R. (2007b). Intentional behaviorism. Behavior and Philosophy, 35, 1–56. Foxall, G.R. (2007c). Explaining consumer choice: Coming to terms with intentionality. Behavioural Processes, 75, 129–145. Foxall, G.R. (2008a). Reward, emotion and consumer choice: From neuroeconomics to neurophilosophy. Journal of Consumer Behaviour, 7, 368–96. Foxall, G.R. (2008b). Intentional behaviourism revisited. Behavior and Philosophy, 37, 113–56. Foxall, G.R. (2009). Ascribing intentionality. Behavior and Philosophy, 38, 1–6. Foxall, G.R. (2010). Interpreting Consumer Choice: The Behavioural Perspective Model. New York and London: Routledge. Foxall, G.R. (in press). Accounting for consumer choice: Inter-temporal decisionmaking in behavioural perspective. Marketing Theory. Foxall, G.R., & Greenley, G.E. (1998). The affective structure of consumer situations. Environment and Behavior, 30, 781–98. Foxall, G.R., & Greenley, G.E. (1999). Consumers’ emotional responses to service environments. Journal of Business Research, 46, 149–58. Foxall, G.R., & Greenley, G.E. (2000). Predicting and explaining responses to consumer environments: An empirical test and theoretical extension of the Behavioural Perspective Model. The Service Industries Journal, 20, 39–63. Foxall, G.R., & James, V.K. (2002). Behavior analysis of consumer brand choice: A preliminary analysis. European Journal of Behavior Analysis, 2, 209–20. Foxall, G.R., & James, V. (2003). The behavioral ecology of brand choice: How and what do consumers maximize? Psychology and Marketing, 20, 811–36. Foxall, G.R., & Schrezenmaier, T.C. (2003). The behavioural economics of consumer brand choice: Establishing a methodology. Journal of Economic Psychology, 24, 675– 95. Foxall, G.R., & Yani-de-Soriano, M.M. (2005). Situational influences on consumers’ attitudes and behavior. Journal of Business Research, 58, 518–25. Foxall, G.R., & Yani-de-Soriano, M. (in press). Influence of reinforcement contingencies and cognitive styles on affective responses: An examination of Rolls’s theory of emotion in the context of consumer choice. Journal of Applied Social Psychology. Foxall, G.R., Oliveira-Castro, J.M., & Schrezenmaier, T.C. (2004). The behavioral economics of consumer brand choice: Patterns of reinforcement and utility maximization. Behavioural Processes, 65, 235–60. Foxall, G.R., Oliveira-Castro, J.M., James, V.K., & Schrezenmaier, T.C.M. (2007). The Behavioral Economics of Brand Choice. London and New York: Palgrave Macmillan. Foxall, G.R., James, V.K., Chang, J., & Oliveira-Castro, J.M. (2010). Substitutability and complementarity: Matching analyses of brands and products. Journal of Organizational Behavior Management, 30, 145–60. Foxall, G.R., James, V.K., Oliveira-Castro, J.M., & Ribier, S. (2010). Product substitutability and the matching law. The Psychological Record, 60, 185–216. Foxall, G.R., Yani-de-Soriano, M., Yousafzai, S. & Javed, U. (2010). Emotion, contingency, and neuroeconomics in the explanation of consumer choice. Working Paper, Cardiff Business School, Cardiff University, UK. Frey, U. (1997). Cellular mechanisms of long-term potentiation: Late maintenance. In J.W. Donahoe, & V.P. Dorsel (Eds), Neural-Network Models of Cognition: Biobehavioral Foundations (pp. 105–28). Amsterdam: North-Holland.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
88
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Frijda, N.H. (2008). The psychologist’s point of view. In M. Lewis, J.M. HavilandJones & L.F. Barrett (Eds), Handbook of Emotions ( 3rd edn), (pp. 68–87). New York: The Guilford Press. Gilboa, S. & Rafaeli, A. (2003). Store environment, emotions and approach behavior: Applying environmental aesthetics to retailing. The International Review of Retail, Distribution and Consumer Research, 13, 195–211. Gray, J.R. & Braver, T.S. (2002). Integration of emotion and cognitive control: A neurocomputational hypothesis of dynamic goal representation. In S. Moore, & M. Oaksford (Eds), Emotional Cognition (pp. 289–316). Amsterdam: John Bejamins. Greenfield, S. (2000). The Secret Life of the Brain. London: Penguin. Groeppel-Klein, A. (2005). Arousal and consumer in-store behavior. Brain Research Bulletin, 67, 428–37. Havlena, W. J., & Holbrook, M. B. (1986). The varieties of consumption experience: Comparing two typologies of emotion in consumer behavior. Journal of Consumer Research, 13, 394–404. Havlena, W.J., Holbrook, M.B., Lehmann, D.R. (1989). Assessing the validity of emotional typologies. Psychology and Marketing, 6, 97–112. Hebb, D.O. (1949). The Organization of Behavior: A Neurophysiological Theory. New York: Wiley. Herrnstein, R.J. (1961). Relative and absolute strength of response as a function of frequency of reinforcement. Journal of the Experimental Analysis of Behavior, 4, 267– 72. Herrnstein, R.J. (1970). On the law of effect. Journal of the Experimental Analysis of Behavior, 13, 243–66. Herrnstein, R.J. (1997). The Matching Law: Papers in Psychology and Economics. H. Rachlin & D.I. Laibson (Eds). Cambridge, MA: Russell Sage Foundation, New York/Harvard University Press. Higley, J.D., Mehlman, P.T., Poland, R.E., Taub, D.M., Vickers, J., Suomi, S.J., Linnoila, M. (1996). CSF testosterone and 5-H1AA correlate with different types of aggressive behaviors. Biological Psychiatry, 40, 1067–82. Hodgkinson, G. P., & Healey, G.P. (2008). Cognition in organizations. Annual Review of Psychology, 59, 387–417. Hodgkinson, G.P., & Healey, M.P. (in press). Psychological foundations of dynamic capabilities: Reflexion and reflection in strategic management. Strategic Management Journal. Holbrook, M., & Batra, R. (1987). Assessing the role of emotions as mediators of consumer responses to advertising. Journal of Consumer Research, 14, 404–20. Huang, H.M. (2001). The theory of emotions in marketing. Journal of Business and Psychology, 16, 239–47. Hursh, S.R. (1984). Behavioral economics. Journal of the Experimental Analysis of Behavior, 42, 435–52. Izard, C.E. (2009). Emotion theory and research: Highlights, unanswered questions, and emerging issues. Annual Review of Psychology, 60, 1–25. Knorski, J. (1948). Conditioned Reflexes and Neuron Organization. Cambridge: Cambridge University Press. Knutson, B., Wolkowitz, O., Cole, S.W., Chan, T., Moore, E., Johnson, R., Terpstra, J., Turner, R.A., & Reus V.I. (1998). Selective alteration of personality and social behavior by serotonergic intervention. American Journal of Psychiatry, 155, 373–9. Kotler, P. (1973). Atmospherics as a marketing tool. Journal of Retailing, 49(4), 48–65. LeDoux, J. (1998). The Emotional Brain. New York: Simon and Shuster. LeDoux, J.E. (2000). Emotional circuits in the brain. Annual Review of Neuroscience, 23, 155–84.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
89
Lewis, P.A., Critchley, H.D., Rothstein, P., & Dolan, R.J. (2007). Neural correlates of processing valence and arousal in affective words. Cerebral Cortex, 17, 742–8. Loewenstein, G., Rick, S., & Cohen, J.D. (2008). Neuroeconomics. Annual Review of Psychology, 59, 647–72. Lowe, C.F. (1983). Radical behaviorism and human psychology. In G.C.L. Davey (Ed.), Animal Models of Human Behavior (pp. 71–93.) Chichester: Wiley. McGinn, C. (1991). The Problem of Consciousness. Oxford: Blackwell. McGinn, C. (1996). The Character of Mind ( 2nd edn). Oxford: Oxford University Press. McGinn, C. (2004). Consciousness and its Objects. Oxford: Oxford University Press. Mehrabian, A. (1979). Effect of emotional state on alcohol consumption. Psychological Reports, 44, 271–82. Mehrabian, A. (1980). Basic Dimensions for a General Psychological Theory: Implications for Personality, Social, Environmental, and Developmental Studies. Cambridge, MA: Oelgeschlager, Gunn & Hain. Mehrabian A., & de Wetter, R. (1987). Experimental test of an emotion-based approach to fitting brand names to products. Journal of Applied Psychology, 72, 125– 30. Mehrabian A., & Riccioni, M. (1986). Measures of eating-related characteristics for the general population: Relationships with temperament. Journal of Personality Assessment, 50, 610–29. Mehrabian A., & Russell J.A. (1974). An Approach to Environmental Psychology. Cambridge, MA: MIT Press. Mehrabian A., & Russell J.A. (1975). Environmental effects on affiliation among strangers. Humanitas, 11, 219–30. Morris, J.D., Klahr, N.J., Shen, F., Villegas, J., Wright, P., He, G., & Liu, Y. (2008). Mapping a multidimensional emotion in response to television commercials. Human Mind Mapping, 30, 789–96. Oliveira-Castro, J.M., Foxall, G.R., & James V.K. (2008). Individual differences in price responsiveness within and across food brands. Service Industries Journal, 28, 733–53. Oliveira-Castro, J.M., Foxall, G.R., & Schrezenmaier, T.C. (2006). Consumer brand choice: Individual and group analyses of demand elasticity. Journal of the Experimental Analysis of Behavior, 85, 147–66. Oliveira-Castro, J.M., Foxall, G.R., James V.K., Pohl, R.H.B.F., Dias, M.B., & Chang, S.W. (2008). Consumer-based brand equity and brand performance. Service Industries Journal, 28, 445–61. Osgood, C.E., Suci, G.J., & Tannenbaum, P.H. (1957). The Measurement of Meaning. Urbana, IL: University of Illinois Press. Osgood, C.E., May, W.H., & Miron, M.S. (1975). Cross-cultural Universals of Affective Meaning. Urbana, IL: University of Illinois Press. Panksepp J. (1998). Affective Neuroscience: The Foundations of Human and Animal Emotions. New York: Oxford University Press. Panksepp, J. (2005). On the embodied neural nature of core emotional effects. Journal of Consciousness Studies, 12, 158–84. Panksepp, J. (2007). The neuroevolutionary and neuroaffective psychobiology of the prosocial brain. In R.I.M. Dunbar, & L. Barrett (Eds), The Oxford Handbook of Evolutionary Psychology (pp. 145–162). Oxford: Oxford University Press. Paul, L.K., Lautzenhiser, A., Brown, W.S., Hart, A., Neuman, D., Spezio, M., & Adolphs, R. (2006). Emotional arousal in agenesis of the corpus callosum. International Journal of Psychophysiology, 61, 47–56.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
90
January 21, 2011
12:54
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Phelps, E.A. (2005). The interaction of emotion and amygdala: Insights from studies of the human amygdala. In L.F. Barrett, P.M. Niedenthal, & P. Winkielman (Eds), Emotion and Consciousness (pp. 51–66). New York: Guilford Press. Phelps, E.A. (2006). Emotion and cognition: Insights from studies of the human amygdala. Annual Review of Psychology, 57, 27–53. Politser, P. (2008). Neuroeconomics. New York: Oxford University Press. Price, C. (2005). Emotion. Milton Keynes: Open University. Quine, W.V.O. (1960). Word and Object. Cambridge, MA: MIT Press. Rachlin, H. (1995). Self-control: Beyond commitment. Behavioral and Brain Sciences, 18, 109–59. Rauch, S.L., Shin, L.M., Dougherty, D.D., Alpert, N.M., Orr, S.P., Lasko, M., Macklin, M.L., Fischman, A. J., & Pitman, R.K. (1999). Neural activation during sexual and competitive arousal in healthy men. Psychiatry Research Neuroimaging, 91, 1–10. Roberts, R.C. (1988). What an emotion is: A sketch. Philosophical Review, 97, 202–3. Robinson, T.E., & Berridge, K.C. (2003). Addiction. Annual Review of Psychology, 54, 25–53. Rolls, E.T. (1999). The Brain and Emotion. Oxford: Oxford University Press. Rolls, E.T. (2005). Emotion Explained. Oxford: Oxford University Press. Rolls, E.T. (2006). The neurophysiology and functions of the orbitofrontal cortex. In D.H. Zald, & S.L. Rauch (Eds), The Orbitofrontal Cortex (pp. 95–124). Oxford: Oxford University Press. Rolls, E.T., Grabenhorst, F., & Leonardo, F. (2009). Prediction of subjective affective state from brain activations. Journal of Neurophysiology, 101, 1294–308. Romero, S., Foxall, G.R., Schrezenmaier, T.C., Oliveira-Castro, J., & James, V.K. (2006). Deviations from matching in consumer choice. European Journal of Behavior Analysis, 7, 15–40. Rosenberg, A. (1988). The Philosophy of Social Science. Oxford: Clarendon. Rosenberg, A. (1992). Economics: Mathematical Politics or Science of Diminishing Returns? Chicago, IL: Chicago University Press. Russell, B. (1967). The Problems of Philosophy. Oxford University Press. (First published 1912). Russell, J.A., & Barrett, L.B. (1999). Core affect, prototypical emotional episodes, and other things called emotion: Dissecting the elephant. Journal of Personality and Social Psychology, 76, 805–19. Russell, J.A., & Mehrabian, A. (1976). Environmental variables in consumer research. Journal of Consumer Research, 3, 62–3. Russell, J.A., & Mehrabian, A. (1978). Approach-avoidance and affiliation as functions of the emotion-eliciting quality of an environment. Environment and Behavior, 10, 355–87. Ryle, G. (1949). The Concept of Mind. Oxford: Oxford University Press. Schoggen, P. (1989). Behavior Settings. Stanford, CA: Stanford University Press. Searle, J.R. (2000). Consciousness. Annual Review of Neuroscience, 23, 557–78. Seva, R.R., Duh, H.B.L., & Helander, M.G. (2010). Structural analysis of affect in the pre-purchase context. DLSU Business and Economics Review, 19, 43–52. Sherman, E., Mathur, A., & Smith, R.B. (1997). Store environment and consumer purchase behavior: Mediating role of consumer emotions. Psychology and Marketing, 14, 361–78. Skinner, B.F. (1945). The operational analysis of psychological terms. Psychological Review, 52, 270–7. Skinner, B.F. (1953). Science and Human Behavior. Macmillan, New York. Skinner, B.F. (1969). Contingencies of Reinforcement: A Theoretical Analysis. Englewood Cliffs, NJ: Prentice-Hall.
P1: OTA/XYZ P2: ABC c02 JWST046-Hodgkinson
January 21, 2011
12:54
Printer Name: Yet to Come
BRAIN, EMOTION, AND CONTINGENCY IN CONSUMER BEHAVIOUR
91
Skinner, B.F. (1971). Beyond Freedom and Dignity. New York: Knopf. Smith, T.L. (1994). Behaviour and its Causes: Philosophical Foundations of Operant Psychology. Dordrecht: Kluwer. Solomon, R.C. (1973). Emotions and choice. The Review of Metaphysics, 27, 20–41. Solomon, R.C. (2008). The philosophy of emotions. In M. Lewis, J.M. HavilandJones, & L.F. Barrett (Eds), Handbook of Emotions (3rd edn), (pp. 3–16). New York: Guilford Press. Spielberg, J.M., Stewart, J.L., Levin, R.L., Miller, G.A., & Heller, W. (2008). Prefrontal cortex, emotion, and approach/withdrawal motivation. Social and Personality Psychology Compass, 2, 135–53. Staddon, J.E.R. (Ed.) (1980). The Limits to Action: The Allocation of Individual Behavior. New York: Academic Press. Strawson, G. (1994). Mental Reality. Cambridge, MA: MIT Press. Sulloway, F.J. (2007). Birth order and sibling competition. In R.I.M. Dunbar, & L. Barrett (Eds), The Oxford Handbook of Evolutionary Psychology (pp. 297–310). Oxford: Oxford University Press. Sweeney, J.C., & Wyber, F. (2002). The role of cognitions and emotions in the musicapproach-avoidance behavior relationship. Journal of Services Marketing, 16, 51–69. Tai, S.H.C., & Fung, A.M.C. (1997). Application of an environmental psychology model to in-store buying behavior. The International Review of Retail, Distribution and Consumer Research, 7, 311–37. Taylor, S.E., Klein, L.C., Lewis, B.P., Gruenewald, T.L., Gurung, R.A.R., & Updegraff, J.A. (2000). Biobehavioral responses to stress in females: Tend-and-befriend, not fight-or-flight. Psychological Review, 107, 411–29. Toronchuk, J.A., & Ellis, G.F.R. (2010). Affective Neuronal Darwinism: The Nature of the Primary Emotional Systems. http://www.mth.uct.ac.za/∼ellis/ToronchuK% 20&%20Ellis,%202010.pdf (accessed 26 June 2010). Van Kenhove, P., & Desrumaux, P. (1997). The relationship between emotional states and approach or avoidance responses in a retail environment. The International Review of Retail, Distribution and Consumer Research, 7, 351–68. Wager, T.D., Feldman Barrett, L., Bliss-Moreau, E., Lindquist, K.A., Duncan, S., Kober, H. Joseph, J., Davidson, M., & Mize, J. (2008). The neuroimaging of emotion. In M. Lewis, J.M. Haviland-Jones, & L.F. Barrett (Eds), Handbook of Emotions (3rd edn), (pp. 249–71). New York: Guilford Press. Whalen, P.J. & Phelps, E.A. (Eds.) (2009). The Human Amygdala. New York: Guildford Press. Winkielman, P., Berridge, K.C., & Wilbarger, J.L. (2005). Emotion, behavior, and conscious experience. In L.F. Barrett, P.M. Niedenthal, & P. Winkielman (Eds), Emotion and Consciousness (pp. 335–62). New York: Guilford Press. Yani-de-Soriano, M., & Foxall, G.R. (2006). The emotional power of place: The fall and rise of dominance in retail research. Journal of Retailing and Consumer Services, 13, 403–16. Yani-de-Soriano, M., Foxall, G.R., & Newman, A. (in press). Interactive effects of the determinants of consumer response. Psychology and Marketing.
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
Chapter 3 LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE AND WORK ATTITUDES: CONCEPTUAL AND METHODOLOGICAL ISSUES David Chan School of Social Sciences, Singapore Management University, Singapore In industrial and organizational (I/O) psychology, we are often interested in systematic intra-individual changes in job performance (e.g., task performance, contextual performance) or work attitudes (e.g., organizational commitment, withdrawal behaviors). Indeed, many research areas are explicitly about the intra-individual changes and processes that unfold in various ways over time (e.g., learning and skill acquisition, newcomer socialization). Consistent with the focus on these intra-individual change phenomena, the past two decades of the I/O psychology literature has seen a sustained interest in conceptual and methodologic issues relating to the longitudinal assessment of various facets of changes over time (e.g., Bliese, Chan, & Ployhart, 2007; Chan, 1998a, 2011; Chan & Schmitt, 2000; Hofmann, Griffin, & Gavin, 2000; Hofmann, Jacobs, & Baratta, 1993; Wang & Chang, in press). There are various complexities in understanding changes over time including issues of levels of analysis, measurement error, multivariate modeling, and types of change in the focal variable. For example, the process of change may exist at one or more of multiple levels of analysis, such as at the individual, the group, or the organizational level. This raises fundamental construct validity issues (i.e., issues of composition models, see Chan, 1998b) such as whether the same or different constructs are being conceptualized and assessed at different levels, the functional relationships linking the constructs at the different levels, and whether the same or different processes of changes over time or interconstruct relationships are occurring at different levels. Changes over time may also exist in complex ways in cross-levels situations. For example, changes over time at one level may affect the changes over time or eventual outcome International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
94
January 21, 2011
12:56
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
at another level. Another type of cross-levels situations concerns changes over time in an inherently cross-levels construct such as person–group fit (a composite construct consisting of the lower level person component and the higher level group component), which raises issues of how different rates of change or different types of change occurring at different levels (or components) impact on the cross-levels (composite) construct. More fundamentally, any observed changes over time need to be decomposed into random fluctuations versus systematic changes in the focal variable. When systematic change over time exists, the trajectory of a variable may have time-varying correlates and the trajectory may affect or be affected by the trajectories of other variables, such that we need multivariate models that specify and test relationships linking changes in different focal variables. The type of change that the focal variable is undergoing may be changes in quantitative level, qualitative status, or a mixture of both. Finally, there may be between-group differences in one or more of the various facets of changes over time, and these groups may be observed groupings such as gender and culture groups or unobserved (or latent) groupings distinguishable by distinct characteristics of changes over time. Understanding the above complexities and the various facets of change over time, in terms of both the conceptual and methodologic considerations, is necessary in order to make adequate substantive inferences from the longitudinal assessment of changes in job performance, work attitudes, or other focal variables. The purpose of this chapter is to provide a state-of-the-art review of the conceptual and methodologic advances in understanding changes over time, with a focus on future research challenges and directions on substantive applications to studies of job performance and work attitudes.
MULTILEVEL ISSUES Many phenomena in I/O psychology research are inherently multilevel. With conceptual and methodologic advances in multilevel analysis (e.g., Chan, 1998b; Kozlowski & Klein, 2000; Morgeson & Hofmann, 1999), more studies are attempting to model multilevel phenomena. The bulk of the multilevel research in I/O psychology discusses the “traditional” type of multilevel data in which individuals are nested within groups. In modeling changes over time using longitudinal data, we are in fact dealing with a type of multilevel data in which the multilevel structure is less obvious. Longitudinal data are obtained from measurements repeated on the same individuals over time, and hence a multilevel structure is established with the repeated observations over time (Level 1) nested within individuals (Level 2). While the multilevel analysis of cross-sectional grouped data is concerned with inter-individual differences associated with group membership, multilevel analysis of longitudinal data is concerned with modeling intra-individual change over time. Although multilevel regression models can also be used to
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
95
analyze these changes over time (e.g., Bryk & Raudenbush, 1992), the issues of changes over time are often very complex and may involve facets of change over time (e.g., conceptual changes in the constructs, changes in calibration of measurement, various types of time-related error-covariance structures) that are not readily handled by multilevel regression models. In modeling change over time, we are primarily concerned with describing the nature of the trajectory of change and accounting for the inter-individual differences in the functional forms or parameters of the trajectories by relating them to explanatory variables. The explanatory variables may be in the form of experimentally manipulated or naturally occurring groups, time-invariant predictors, timevarying correlates, or the trajectories of a different variable. Latent growth modeling (LGM) and its extensions are well suited to address these issues. Chan (1998a) provided a detailed review of these issues and the application of LGM techniques, as well as an overview comparison between latent variable models and multilevel regression models. Developments in latent variable analysis, particularly structural equation modeling, have been successfully applied to modeling the complexities involved in a variety of these changes (see Chan, 1998a, 2002a,b, 2005; Singer & Willet, 2003). Duncan, Duncan, and Strycker (2006) provided a concise introduction to the technical issues involving analysis of data with complicated hierarchical structures combining intra-individual changes over time with two or more additional levels (e.g., individual, team, and organizational levels) and demonstrated how multilevel latent growth models can be applied to such data sets. However, despite the separate advances in multilevel research and longitudinal assessment of change, there is a lack of integration of the advances in these two areas and several important multilevel issues in modeling changes over time have not received sufficient, if any, attention. I classify these issues broadly into three categories: 1. Modeling changes over time at multiple levels; 2. Modeling cross-levels effects of changes over time; and 3. Modeling dynamic cross-levels constructs. These issues are explicated in the following sections. Modeling Changes Over Time at Multiple Levels Traditionally, the majority of the research on job performance and work attitudes (e.g., personnel selection, performance appraisal, job satisfaction, work motivation) has approached the focal constructs or phenomena under investigation from a micro perspective that focuses almost exclusively on individuallevel variables such as the individual’s abilities, job performance, and work perceptions. However, many constructs and phenomena of interest examined by I/O psychologists are in fact multilevel in nature involving multiple levels of
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
96
January 21, 2011
12:56
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
analysis such as the individual, group, and organization. Often, a higher level construct (e.g., team performance) is composed by aggregating (in one of several methods) the units at the lower level construct (e.g., each team member’s job performance). Correspondingly, the process of changes over time may exist at one or more of multiple levels of analysis, such as at the individual level and team level. The adequate examination of multilevel constructs and data sets involve addressing complex conceptual, measurement, and data analysis issues. Several researchers have developed useful organizing frameworks that help to clarify conceptualizations and decide on measurements or operationalizations of similar constructs at multiple levels, as well as identify the types of relevant evidence to support the multilevel hypotheses. Specifically, in the last two to three decades, many I/O psychologists have contributed to multilevel research by developing conceptual organizing frameworks (e.g., Chan, 1998b; Kozlowski & Klein, 2000; Morgeson & Hofmann, 1999; Rousseau, 1985) and many scholars have helped advance multilevel research by providing relatively nontechnical summaries and applications of existing multilevel analytical strategies (Bliese, 2000; Bryk & Raudenbush, 1992) which helped users in I/O psychology to apply appropriate techniques to their multilevel data sets. For a summary of these conceptual bases and analysis issues concerning multilevel analytical strategies, see Chan (2005). Composition models (Chan, 1998b; Rousseau, 1985) address fundamental construct validity issues by specifying the functional relationships linking the constructs at the different levels that reference essentially the same content but are qualitatively different at different levels (e.g., self-efficacy versus team efficacy). In Chan’s (1998b) typology of composition models, four of the five models (i.e., additive, direct consensus, referent-shift consensus, dispersion) have received much attention and have been applied in the past decade of multilevel studies. A common feature across these four composition models is that they are all focused on the static core attributes of focal constructs (e.g., efficacy perceptions), which describe some stable units or state of affairs at the individual or higher level. While these four models are fundamental and one or more of these models are necessary in composing the lower level construct to the higher level construct, they do not directly provide the essential conceptual basis for composing the process of changes over time at the lower level to the higher level. The fifth model in Chan’s typology, namely process composition, provides this basis. Process composition models are concerned with composing a process or mechanism from the lower level to the higher level, and are therefore well suited to examining changes over time at multiple levels in so far as changes over time are construed as processes or mechanisms that unfold over time. In a process composition model for changes over time, the change process or mechanism is first specified at the lower level explicating the essential or critical parameters and their interrelationships. The change process then is
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
97
composed to the higher level by identifying critical higher level parameters, which are higher level analogs of the lower level parameters, and describing interrelationships among higher level parameters, which are homologous to the lower level parameter relationships. To illustrate, two examples are provided below. An example of process composition of changes over time is a simple learning process in performing a task being composed from the individual level to the team level. At the individual level, the trajectory of change over time in task performance follows a functional form characterized by an initial period of linear increase in performance levels at a constant rate of change, an intermediate period of increase in performance levels at a decreasing rate of change, and a final maintenance period of no change in performance levels. Throughout the entire change period, the same construct of task performance is assumed to be measured and with the same precision, in the sense that there are no changes in the conceptual domain of the construct being measured nor changes in the calibration of the measurement. In other words, there is measurement invariance across time in that any difference in performance levels between two time points represents meaningful change and the magnitude of change may be directly interpreted as a change in the absolute level of task performance. In this example of simple process composition of performance changes over time, task performance at the team level is construed to change over time following a similar trajectory as that of the individual level, with the same assumptions of measurement invariance across time. That is, the individual level of performance changes over time is composed to the team level such that the critical parameters are analogous across the individual and team levels. The critical parameters are assumptions of measurement invariance across time and the functional form of the trajectory represented by the direction and rate of changes in performance levels over the entire change period. In terms of analysis, structural equation modeling may be used to test for measurement invariance across time at each of the two levels. After measurement invariance over time has been established, LGM or other techniques of longitudinal analysis (e.g., multilevel latent variable modeling) may be used to identify and test whether the functional forms of the change trajectory are similar or different across the two levels, depending on the substantive hypotheses concerning between-levels differences in the directions and rates of change in performance changes over time. A more complex example of process composition of changes over time is in the study of development of (i.e., changes over time in) efficacy perceptions about the team at both the individual and team levels. In research on work teams (e.g., Guzzo, Yost, Campbell, et al., 1993), efficacy perceptions at the team level is often a case of referent-shift consensus composition (Chan, 1998b). The composition starts with the individual-level construct of selfefficacy. Self-efficacy is defined as an individual’s belief and confidence in mobilizing his or her resources for successful task performance (an example
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
98
January 21, 2011
12:56
Printer Name: Yet to Come
I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
item is “I am confident that I can perform this task.”). A new form of the construct at the same level (i.e., individual level) is then derived by shifting the referent in the efficacy perception from the self to the team as a whole (an example item for the new form of the construct is “I am confident that my team can perform this task.”). The new construct, namely collective efficacy, is defined as the individual team member’s belief and confidence that the team can mobilize its resources for successful task performance. Note that collective efficacy is still at the original individual level of conceptualization (Guzzo, Yost, Campbell, et al., 1993). Within-group consensus (as indexed by within-group agreement of individuals’ perceptual scores) is used to justify the aggregation of individuals’ collective efficacy perceptions to represent the value of the higher level (i.e., group level) construct called team efficacy. In this example, assume that the researcher is examining how team members’ collective efficacy perceptions change over time, and is interested in describing the process in which the team progressively changes from the state of lack of within-group agreement of individual-level collective efficacy perceptions to the state of high within-group agreement. That is, the researcher wants to compose a team-level process of team efficacy emergence. To do so, the researcher first specifies an individual-level process describing how an individual develops collective efficacy perceptions. For simplicity, assume the researcher has a theory that development of collective efficacy is an integration process, moving from an initial state in which the individual’s distinct efficacy beliefs about the team’s ability in accomplishing distinct aspects of the task are unrelated or, at best, loosely interrelated through progressive states in which these separate beliefs become increasingly interrelated to the eventual state in which they become integrated into a single global belief. This integration process, which describes the nature of the changes over time in collective efficacy perceptions at the individual level, is composed to the higher level to specify the process of team efficacy emergence. That is, the researcher could specify team efficacy emergence as an integration process, moving from an initial state in which there is little agreement among individuals’ collective efficacy perceptions, through progressive states in which the level of agreement gradually increases, to the eventual state in which high agreement is achieved. Note that within-group agreement is a higher level analog of intra-individual correlation of collective efficacy beliefs of the team’s ability in accomplishing distinct aspects of the task. Similarly, the notion of increasing levels of withingroup agreement as a team progresses over time is analogous to the notion of increasing inter-correlations among distinct collective efficacy beliefs as an individual progresses over time. The initial and final states of the team also are analogous to the respective states of the individual. In short, critical parameters of the integration process at the individual level have higher level analogs that constitute the critical process parameters at the team level. In this team efficacy emergence example, the within-group agreement index is the higher level operational analog of the correlation coefficient. Note that
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
99
the within-group agreement index is used here as a dispersion measure (assessing team efficacy strength at multiple points in time) as opposed to a statistical criterion for aggregation. The integration process in team efficacy emergence can be construed as changes in team efficacy strength (as opposed to team efficacy level; i.e., intra-team changes in variance as opposed to intra-team changes in means). Hence, when moving from the source construct of collective efficacy to the higher level process of team efficacy emergence, a dispersion composition (from collective efficacy to team efficacy strength) precedes the process composition. As noted in Chan (1998b), a theory of the focal construct in a multilevel study may contain several composition forms. In terms of analysis, nested model comparisons using longitudinal confirmatory factor analysis may be used to test for changes over time in the factor structure of efficacy beliefs at each of the two levels. The hypothesized integration process of change over time would test for the fit of a longitudinal confirmatory factor analytic model (and compare it against competing models) that specified different theory-driven factor structures corresponding to initial, intermediate, and final time periods. Specifically, we expect a factor structure of distinct and uncorrelated or lowly correlated efficacy beliefs in the initial period, a factor structure of moderately correlated efficacy beliefs in the intermediate period, and factor structure of highly correlated efficacy beliefs or a single global factor of efficacy beliefs in the final period (for technical issues of analysis, see Chan, 1998a). The above conceptual and methodologic principles relating to process composition to model changes over time at multiple levels may be applied to a variety of substantive research areas in job performance and work attitudes, such as changes in levels and dimensionality of performance, organizational citizenship behaviors (OCBs), cohesion, withdrawal behaviors, cynicism, commitment, and organizational climate. Modeling Cross-levels Effects of Changes Over Time In addition to occurring at multiple levels, changes over time may occur in complex ways in cross-levels situations. There are two types of cross-levels situations that have not received sufficient attention in the literature. This sub-section and the next address these two types of situations, respectively. Almost all cross-levels situations in multilevel research refer to cross-levels effects in which the predictor or causal variable is at one level and the criterion or effect variable is at a different level (higher or lower level than the predictor or casual variable). Cross-levels effects involving changes over time are complex because the focal construct represented by the predictor/causal variable, the criterion/effect variable, or both, are not static. In fact, the values on the variable representing changes over time are not the absolute magnitude on the focal construct assessed by the static variable as indicated by the crosssectional measurement. Instead, the values representing changes over time are
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
100 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY growth parameters that describe the nature of the change trajectory including the rates of change (e.g., magnitude of the slope in a linear growth) or shape of the trajectory (e.g., quadratic function). In the vast majority of studies examining cross-levels effects that involve changes over time, the purpose has been to estimate the predictive or causal effect of a time-invariant (i.e., static) individual difference variable (e.g., cognitive ability or personality traits) on the inter-individual differences in intraindividual changes over time in the criterion or effect variable (e.g., Hofmann, Jacobs, & Baratta, 1993). The research strategy is focused on measuring the rate of change in the criterion/effect variable being tracked over time (e.g., inter-individual differences in the slope of job performance). That is, changes over time are treated as endogenous in the model and often at the lower level (intra-individual level represented by the repeated observations within an individual). The predictor/causal variable, on the other hand, is treated as exogenous and often at the higher level (individual level represented by the inter-individual differences on the trait construct). That is, the cross-levels effect is a downward effect represented in the model by a unidirectional path from the predictor at the higher level to the criterion (i.e., changes over time) at the lower level. This conceptualization fits nicely into the traditional hierarchical linear modeling (HLM) analytic framework used in multilevel research in which the individual (i.e., trait) is treated as a time-invariant predictor at the higher level (Level 2) accounting for (causing or predicting) the intraindividual changes over time occurring at the lower level (Level 1, which is nested under Level 2). In the HLM analytic framework, inter-individual differences in intraindividual changes over time is conceptualized as a growth parameter to be explained (predicted or caused) by other variables (i.e., treated as endogenous). The HLM technique is not suited to conceptualize the growth parameter representing these changes over time as an exogenous variable that explains (predicts or causes) other variables. Clearly, whether changes over time should be conceptualized as exogenous or endogenous should not be determined by what a chosen analytical technique can or cannot do, but should be determined by theory, as translated into conceptual model linking changes over time to other variables. Consider a theory of newcomer adaptation which states that while newcomers follow a linearly increasing trajectory in their intra-individual changes in job performance over time during the transition period (e.g., first 3 months in the organization), those who increased their performance at a faster rate (i.e., a higher slope of performance) will be assessed by their supervisors after the end of the transition period to have a higher potential to advance in the organization. In the conceptual model derived from this theory, the growth parameter representing intra-individual changes in performance over time is an exogenous variable that accounts for (causes or predicts) the supervisory assessment of potential. In this example, changes over time occur at the lower
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
101
(i.e., intra-individual) level and supervisory assessment of potential is the criterion at the higher (i.e., individual) level. That is, the cross-levels effect is an upward effect represented in the model by a unidirectional path from the predictor (i.e., changes over time) at the lower level to the criterion at the higher level. The LGM framework has the flexibility to represent changes over time as either exogenous or endogenous. Hence, the analytic framework is well suited to model upward cross-levels effects in which changes over time occurring at the intra-individual (lower) level account for the individual (higher) level variable. Specifically, the hypothesized LGM would specify a structural unidirectional path from the growth parameter to the individual-level criterion variable (in this example, supervisory assessment of potential). In the newcomer adaptation example, we could extend the theory by stating that newcomers with higher cognitive ability will be assessed by their supervisors after the end of the transition period to have a higher potential to advance in the organization due to their higher rate of change in performance over the transition period. In other words, the relationship between newcomer cognitive ability and supervisory assessment of potential is mediated by intra-individual changes in performance over time. In this conceptual model, there is a downward cross-levels effect represented by a unidirectional path from cognitive ability at the higher (individual) level to the performance changes over time at the lower (intra-individual) level, as well as an upward cross-levels effect represented by a unidirectional path from performance changes over time at the lower (intra-individual) level to potential at the higher (individual) level. To test this conceptual model, an LGM specifying these two structural unidirectional paths to represent the mediation effect could be fitted to the data. In addition to serving either as a predictor or a mediator, the growth parameter can serve as a moderator in a cross-levels effect situation. For example, we could extend the theory of newcomer adaptation to state that the strength of the positive relationship between newcomer impression management tendency (individual level) and supervisory assessment of potential (individual level) is moderated by newcomer performance changes over time (intra-individual level). In this example, a cross-levels effect occurs because the moderator is at a lower level than the level of the two variables with their relationship being moderated. In principle, this moderator effect can be tested by fitting an LGM in which the growth parameter interacts with the impression management variable to affect the potential assessment variable. In practice, testing this moderator effect may face certain analytic challenges due to technical difficulties associated with incorporating nonlinear functions involving continuous variables in latent variable models generally and latent growth models specifically, although there have been significant advances in methods for testing interactions involving latent variables (e.g., Joreskog & Yang, 1996; Wen, Marsh, & Hau, 2002). An alternative and easier way is to adopt a two-step approach to the analysis, by first using LGM to obtain the newcomer’s score on
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
102 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY the growth parameter (i.e., slope) and next, using moderated regression analysis, regress the criterion variable (i.e., potential) on the growth parameter, impression management, and the growth X impression management interaction term. This moderated regression tests the hypothesis that the strength of the relationship between impression management and potential is moderated by performance changes over time (i.e., the growth parameter). To summarize, cross-levels effects involving changes over time may occur in different ways. Although most commonly studied as such, the cross-levels effect need not always be a downward effect represented in the model by a unidirectional path from the predictor at the higher (individual) level to the criterion (i.e., changes over time) at the lower (intra-individual) level. As shown in this sub-section, cross-levels effects could be upward, represented by unidirectional paths from changes over time at the lower level to the criterion variables at the higher level. In addition, changes over time need not always be the criterion variable in the cross-levels effect relationship – it could be the predictor, mediator, or moderator. It is theoretically reasonable to expect that intra-individual changes over time in performance or work attitudes could have causal or predictive efficacy in accounting for individual level variables. By reconceptualizing intra-individual changes over time from a criterion variable to a predictor, mediator, or moderator variable, we are likely to open new and fruitful avenues for substantive research. Modeling Dynamic Cross-levels Constructs Another type of cross-levels situation in changes over time concerns dynamic cross-levels constructs. Unfortunately, discussions on this important aspect of change over time are virtually absent in the literature on longitudinal assessment. This somewhat surprising neglect needs to be addressed, particularly in I/O psychology where many study variables are inherently cross-levels constructs and the real-world phenomena that they represent outside the study are dynamic in nature (i.e., changes over time do occur). In I/O psychology, the prototypical examples of cross-levels constructs are person–environment (P-E) fit constructs such as person–job fit, person–group fit, and person–organization fit (for review of P-E fit constructs, see Edwards, 1994; Kristof, 1996). P-E fit constructs are inherently multilevel in nature. A cross-levels construct (Chan, 1998b), such as person–group fit, is a composite construct consisting of a lower level component construct (in this example, the person-level construct) and a higher level component construct (in this example, the group-level construct). A cross-levels construct is dynamic when one or both of the level construct changes over time. Dynamic cross-levels constructs are complex and they raise critical issues regarding how different rates of change or different types of change may occur at different levels (or components) and how these differential changes impact the cross-levels (composite) construct.
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
103
Theories and research on job performance and work attitudes, and more generally in the areas of recruitment, selection, classification, training and development, appraisal, and turnover, are inextricably linked to studies on P-E fit in so far as the investigation is focused on the match between the person and the work environment in which the person functions. In such studies, the nature of the P-E fit construct, as well as its predictive validity, is dependent on both the nature of the person constructs and the environment constructs in question. Clearly, any longitudinal changes in either the person level construct or environment level construct will have an impact on the P-E fit construct. Hence, dynamic cross-levels constructs such as dynamic P-E fit constructs pose important conceptual and methodologic challenges that need to be adequately addressed if we are to make substantive inferences from P-E fit studies. In any P-E fit study, the type of fit may be construed as complementary fit or supplementary fit. Complementary fit is concerned with the match between the nature of the needs or capabilities of the person and what the environment offers to or requires of the person. For example, the organization may demand time and ability, and the extent to which the employee supplies these resources affects complementary fit. Supplementary fit is concerned with the similarity in values, beliefs, and other characteristics between the person and the organization. For example, the extent to which employees with creative interests have the opportunity in the organization to engage in unstructured and unconventional activities affects supplementary fit. Changes over time in either the person or organization levels could affect the cross-levels P-E fit construct in various ways. Consider the situation of high complementary fit between a person’s cognitive ability and the ability demands required from the work environment. Assuming that the person’s cognitive ability remains constant over the time period in question, changes over time in the magnitude or type of work environment demands (e.g., the level of the ability demands of the environment increased or decreased considerably; new non-ability demands emerged in the environment) could lead to intra-individual changes in P-E fit (in this example, P-E fit decreases over time) even while the trait levels of individuals remain constant over time. This has direct implications and challenges for developing practical recommendations for recruitment and selection (using trait levels of individuals) when the empirical basis is constituted by the findings from cross-sectional (static) assessment of P-E fit. The problem gets more complicated when changes over time are occurring at both the person and environment levels, and especially when the rates of change and even nature of change over time (e.g., functional form of the trajectory, dimensionality of the construct) differ between levels. Similar issues apply to supplementary fit if the fit construct is dynamic. In short, when a P-E fit construct is in fact dynamic (i.e., changes over time), a static representation of the P-E fit construct obtained from the crosssectional assessment is likely to result in misleading substantive inferences and
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
104 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY practical recommendations. It is also important to note that although high P-E fit is generally predictive of positive outcomes, it is not true that fit is inherently adaptive nor is misfit inherently maladaptive. Construct-oriented studies need to be undertaken to address cutting edge research questions concerning when and how fit may have negative effects (e.g., through groupthink processes) and misfit may have positive effects (e.g., through innovative ideas), and these questions can only be adequately addressed by explicating the nature of the constructs and construct relationships involved, as well as the degree and type of changes over time that may occur in the lower level (person) and higher level (environment) components of the cross-levels fit construct. Static cross-levels constructs such as P-E fit constructs are typically analyzed using polynomial regressions (e.g., Edwards, 1994) or hierarchical multiple regressions to test the P-E interaction term representing the P-E fit construct (e.g., Chan, 1996). However, these techniques are not well suited to model dynamic cross-levels constructs because they were not developed to directly assess the various facets of intra-individual changes over time (Chan, 1998a). Given the lack of conceptual attention given to dynamic cross-levels constructs, it is not surprising that methodologic or data analysis discussions on longitudinal modeling have not explicitly discussed the assessment of dynamic cross-levels constructs. Fortunately, advances in LGM, with its multilevel, multivariate, and multiple group extensions, could provide a unified and flexible approach to model the various facets of changes in dynamic cross-levels constructs including both changes in the cross-levels construct and the correlates of these dynamic changes (i.e., causes or predictors of longitudinal changes in the crosslevel constructs and the impact that these changes have on other constructs). LGM could also be combined with longitudinal confirmatory factor analyses methods and measurement invariance analyses to assess changes in dimensionality over time for each level component of the cross-levels construct. For a review of LGM and its extensions, see Chan (1998a; 2002a) and Duncan, Duncan, and Strycker (2006).
MULTIVARIATE ISSUES Before one selects the appropriate technique for analyzing a longitudinal data set, the specific question about the change over time in the longitudinal process must be explicated (Chan, 1998a). These questions may be broadly classified into descriptive and explanatory questions. The descriptive question asks how the repeatedly measured unit of analysis (e.g., individual, group, organization) changes over time on one or more focal variables (e.g., job performance, group cohesion, organizational climate for safety). For example, in a study of changes in job performance over time, we may ask if the performance change is “reversible.” That is, does the trajectory of performance change
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
105
follow some monotonically increasing or decreasing (e.g., linear) functional form that represents an irreversible (at least within the longitudinal time period studied) change or some non-monotonic functional form (e.g., an “inverted U”) that represents reversible change over time? Another example of a descriptive question is whether the change in the focal variable is simply a direct quantitative change in magnitude (often referred to as alpha change) or a qualitative change in the conceptualization (often referred to as gamma change) of the construct of interest (Chan, 1998a; Golembiewski, Billingsley, & Yeager, 1976). For example, in newcomer adaptation research, an interesting question concerns whether organizational commitment is changing in strength only (alpha change) or it is changing in dimensionality (gamma change) over time. The descriptive question is concerned with the “what” or “how” of the change trajectory. The explanatory question focuses on the “why” by seeking to understand and predict the pattern of intra-unit (e.g., intra-individual) change over time described by (i.e., obtained from) the data. The variation to be explained or predicted here is typically the inter-individual differences in intra-individual changes over time. Consider the case where all individuals follow a positive linear trajectory of change in job performance but differ in the rate of change (i.e., slope). If this inter-individual variation in rate of change is systematic and not due to measurement error, then addressing the explanatory question may involve seeking to understand and predict the variation by incorporating one or more explanatory variables (e.g., cognitive ability) in the model of change and estimating their predictive validity. The explanatory variable can either be a time-invariant or timevarying predictor. It is evident from the above discussion, as well as the earlier discussion on multilevel issues, that understanding changes over time in a focal variable requires us to adopt a multivariate approach such that we can go beyond describing the intra-individual changes (in terms of the means and variances of the growth parameters associated with the functional form of the trajectory) to explaining these changes by identifying relevant predictor or criterion variables (time-invariant or time-varying) that are correlated with the inter-individual differences in the intra-individual changes. Indeed, in modeling change over time, the primary purposes are describing the nature of the trajectory of change and attempting to account for the inter-individual differences in the functional forms or parameters of the trajectories by relating them to explanatory variables that may be in the form of experimentally manipulated or naturally occurring groups, time-invariant predictors, time-varying correlates or the trajectories of a different variable. LGM and its extensions to examine multivariate and multiple group situations, which are implemented in a latent variable modeling framework, are well suited to address these issues. This section explicates the issues and advances in multivariate approaches to modeling changes over time and the next section examines multiple group issues.
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
106 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY To understand multivariate models of changes over time, we first need to have good grasp of the issues concerning how adequate univariate models can be specified to assess changes over time and subsequently combined to form multivariate models. LGM provides a unified and flexible framework for this purpose. Technical discussions of the data analytic issues are readily available in Chan (1998a). Hence, it suffices to provide a conceptual overview below to serve as the basis for discussing multivariate issues in modeling changes over time. I first introduce the LGM framework, beginning with the basic univariate growth model and how time-invariant predictors can be incorporated in the model, and then explicate the different types of multivariate growth models that can be specified to address more complex situations. Latent Growth Models Latent variable approaches are well suited for longitudinal modeling. They are highly flexible and powerful because a variety of latent variable models (i.e., structural equation models) can be fitted to the longitudinal data to describe, in alternative ways, the change over time. LGM, which is implemented using a latent variable approach, offers a direct and comprehensive assessment of the nature of true intra-individual changes and inter-individual differences in these changes. LGM also allows these differences to be related to individual predictors. An LGM can be elaborated into a multiple-indicator latent growth model, in which the focal variable of change is modeled as a latent variable represented by multiple indicators, thereby allowing both cross-sectional and longitudinal measurement errors to be modeled directly and assess whether the extent of distorting effects, if any, that these measurement errors have on the parameter estimates of true change. Technical details of LGM and multiple-indicator LGM are described in Chan (1998a). LGM represents the longitudinal data by modeling inter-individual differences in the attributes (i.e., parameters) of intra-individual changes over time (i.e., individual growth curves). In an LGM analysis, we can estimate the means and variances of the two growth parameters (intercept and slope factors) and examine if the two parameters are correlated with each other. The LGM analysis can also be used to examine associations between the growth parameters and predictor variables. The predictor could be time-invariant variables such as cognitive ability or personality variables. For example, in newcomer adaptation research, we can use LGM to predict initial status and rate of change in information seeking from proactive personality (Chan, 2000; Chan & Schmitt, 2000). The predictor could also be time-varying variables. When the predictor variable varies over time, it is possible to specify the predictor as a separate univariate LGM (i.e., to model the predictor’s change trajectory) and combine it with the original univariate LGM to form a multivariate growth model known as an associative model (see section below on multivariate latent growth models).
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
107
Multivariate Latent Growth Models To conceptualize and assess the relationships between different variables simultaneously as they change over time, different univariate LGMs can be combined to form a multivariate LGM. There are at least three types of multivariate LGMs: associative models, factor-of-curves models, and curveof-factors models (McArdle, 1988). The associative model is the simplest and most commonly studied type of multivariate model. It involves a direct estimation of the associations between the growth factors of the different univariate models. The factor-of-curves model and the curve-of-factors model extend the conceptualization of the relationships between the univariate models by describing the growth factors, in different ways, in terms of higher order latent growth factors. In each of the three multivariate models, one or more predictors can be included to estimate structural effects from predictors to latent growth factors. In the associative model combining two univariate LGMs representing variables A and B, respectively, the between-variable associations of the growth factors (i.e., InterceptA –InterceptB , InterceptA –SlopeB , SlopeA –InterceptB , SlopeA –SlopeB ) are directly estimated. Hence, by fitting an associative model, parameters from different change trajectories can be correlated to examine cross-domain associations (i.e., relationships between two focal variables being examined for intra-individual change over time). For example, in their study of newcomer adaptation, Chan and Schmitt (2000) specified associative models in which rate of change in proactivities (e.g., relationship building) was correlated with rate of change in adaptation outcomes (e.g., social integration). One or more predictors (e.g., personality traits) can also be included in the associative model, thereby allowing hypotheses regarding differential predictions (using the same individual predictor) of intra-individual change across domains can be tested. In Chan and Schmitt, proactive personality and previous transition experiences were incorporated in the associative model to simultaneously predict intra-individual changes in both proactivities and adaptation outcomes. Chan, Ramey, Ramey, et al. (2000) provide another illustration of how the associative model incorporating predictors of change may be used to examine complex cross-domain (i.e., multivariate) relationships. In this study, the authors examined the change trajectories of children’s social skills in home versus school settings, as well as family predictors of these changes. The study tracked 378 children at four time points, spaced at 12-month intervals over a 4-year period, from Kindergarten to Grade 3. Results showed systematic between-settings differences in children’s social skill development. The trajectory of social skills development at home had a different functional form compared with the trajectory at school. In addition, across settings, there were differential patterns of associations between growth parameters and individual predictors including family income, parent education, and child verbal skills.
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
108 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY The findings were invariant across gender groups. Substantively, the study obtained evidence for the context specificity of the children’s social skill development process. Although the precise nature of the context specificity of the development process and the contextual influences on the process was not addressed in the study, the application of the multivariate LGM framework provided a new direction to empirically examine children’s social skill development as a multifaceted process involving variables interrelated in a dynamic manner. The analysis provided a unified framework for structuring these dynamic relationships. This in turn allows researchers to systematically identify the antecedents, correlates, and consequents of the different aspects of the children’s social skill developmental process. For example, we can examine antecedents of change such as family or school characteristics, correlates of change such as child cognitive growth and changes in peer relations, and consequents of change including proximal outcomes such as child-subjective well-being and school achievement, and distal outcomes such as subsequent development of close relationships and personality development. The associative modeling framework adopted by Chan and Schmitt (2000) and Chan, Ramey, Ramey, et al. (2000) may be similarly applied to many areas of I/O psychology to examine multivariate relationships linking job performance, work attitudes, and other work-relevant variables. For example, in the study of dynamic performance, we can fit an associative model to examine whether the trajectory of intra-individual changes for contextual performance has the same or a different functional form as the trajectory for task performance. Predictors such as cognitive ability and personality traits may be incorporated in the associative model to assess if each predictor has similar or differential effects in predicting the growth parameters across the two performance dimensions. We can also examine antecedents of performance change such as previous work experiences and training interventions, correlates of performance change such as changes in self-efficacy and organizational commitment, and consequents of performance change such as subjective well-being and advancement in the organization. The factor-of-curves model combines univariate LGMs by specifying common higher order growth factors (second-order intercept and slope/shape factors) to account for the (first-order) growth factors of the univariate models, so that growth features that are common across univariate models as well as those that are specific to the univariate models are described. In a factorof-curves model describing linear growth, a second-order intercept factor is specified to account for the covariation between the intercept factors of the univariate models. Similarly, a second-order slope factor is specified to account for the covariation between the slope factors of the univariate models. The two second-order growth factors (intercept and slope) are allowed to covary. The factor-of-curves model is similar in logic to the familiar hierarchical factor model in confirmatory factor analysis, except that the factors in the factor-of-curves model are growth or chronomic (time-based) factors rather
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
109
than static common variance factors. The interpretation of the higher order growth factors, however, is conceptually similar to the higher order factors in confirmatory factor analysis in so far as they are higher order common factors postulated from the same measures (i.e., without including additional measures beyond the measures from which the lower order factors were derived) to account for the covariation of the lower order factors. The conceptual meaning of the second-order growth factors in the factor-of-curves model is dependent on the nature of the variables in the univariate models that were combined. For example, if the three univariate models being combined were modeling changes over time in use of alcohol, marijuana, and tobacco, respectively, then the common second-order growth factors may be interpreted as representing the intercept and slope of a higher order substance use construct (Duncan, Duncan, & Strycker, 2006). Similar to the associative model, predictors may be incorporated into the factor-of-curves model to account for the lower order and higher order growth factors. Clearly, the factor-of-curves model may be applied to many job performance and work attitudes variables. For example, in a study of the dynamics of OCBs, we can track changes in each of the five dimensions of OCBs (altruism, courtesy, civic virtue, compliance, sportsmanship) by first separately fitting a univariate LGM to each OCB dimension to examine the functional form of the trajectory and estimate the growth parameters. The five univariate LGMs may be combined to form an associative model to examine the pairwise between-dimensions covariations of the growth factors. With five univariate LGMs of OCB (each model having an intercept and a slope), there are altogether 40 between-dimensions factor covariations (and five within-dimensions factor covariations) to be estimated in the associate model. Although these 40 between-dimensions factor covariations may be used to describe the growth relationships across the five OCB dimensions and the 40 growth relationships between OCB dimensions can be compared in terms of their strength of association, it is difficult to provide a parsimonious account of the growth relationships among the five OCB dimensions due to the large number of parameter estimates of inter-factor relationships. Now, if all or most of the 40 between-dimensions covariations were significant and substantial, we would conclude that changes in the OCB dimensions were related to each other and a reasonable next question to ask would be whether a common OCB construct, with its intra-individual changes represented by higher order growth factors, could be postulated to account for the between-dimensions factor covariations. Accordingly, we can fit to the combined OCB data a factor-of-curves multivariate LGM of OCBs combining the five univariate LGMs. This is accomplished by specifying one second-order intercept factor and one second-order slope factor, with each factor causing the five corresponding first-order (intercept/slope) factors. The mean and variances of the second-order growth factors together describe the trajectory of the common OCB construct underlying the five OCB dimensions. Note that the OCB dimension-specific
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
110 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY (i.e., first-order) growth parameters are also estimated in the multivariate model. Hence, the factor-of-curves model of OCB allows us to assess changes over time that are specific to each OCB dimension, as well as changes over time that are common across OCB dimensions as represented by changes in the underlying single OCB construct. The 10 coefficients associated with the structural paths from the second-order growth factors to the corresponding first-order factors (five paths for intercepts and five paths for slopes) represent the extent to which the growth factor of the underlying common OCB construct could account for the corresponding growth factor of the respective OCB dimensions. In addition, predictors such as personality traits may be incorporated into the multivariate model to predict the second-order growth factors representing the common OCB construct, in addition to the first-order growth factors of the different OCB dimensions. Hence, by fitting the factorof-curves model of OCBs incorporating personality traits as predictors, we may be able to isolate the different aspects of OCBs that are associated with different personality traits and to different extent. The third type of multivariate LGM is the curve-of-factors model (McArdle, 1988), which is mathematically identical to what Chan (1998a) referred to as multiple-indicator LGM. Although mathematically identical, McArdle (1988) and Duncan, Duncan, and Strycker (2006) focused on the multivariate aspects of the model whereas Chan (1998) focused on the measurement errors and measurement invariance aspects of the model. The two different foci are explicated as follows. Using the substance use example described above, Duncan, Duncan, and Strycker (2006) combined the three univariate LGMs describing alcohol, marijuana, and tobacco use, respectively, into a curve-of-factors model. This is accomplished by first treating the three different substances within the same time occasion as multiple indicators of the time-specific latent factor (i.e., a common variance factor, not a growth factor) called substance use, and then using these first-order time-specific latent factor scores to form the trajectory of changes in substance use defined by the second-order growth factors (intercept and slope). Hence, unlike the associative model and the factor-of-curves model, both of which simultaneously specify the trajectories of different variables and examine the relationships between trajectories of different variables, the curve-of-factors model specifics the trajectory of one variable in which the intercept and slope factors are second-order factors derived from the time series of factor scores where each time point is a first-order factor measured by multiple indicators. Predictors may be incorporated into the model to predict the second-order factors (intercept and slope) or the first-order factors (i.e., the construct within a specific time point). Applying to the OCB example, the curve-of-factors model would specify that the five different OCB variables within each time point are multiple indicators of the same single OCB construct (common variance latent factor) within that time point, and the trajectory of this single OCB construct would be defined by the intercept and
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
111
slope factors (second-order factors) in the model. Predictors such as personality traits may be incorporated to predict the growth parameters of the OCB construct or the OCB construct within a specific time point. Note that in the curve-of-factor model, OCBs are construed as multiple indicators of a single OCB construct as opposed to measures of different distinct OCB constructs. Although the curve-of-factors model is referred to as a multivariate growth model by Duncan, Duncan, and Strycker (2006), it is multivariate only in the sense that there were multiple variables repeatedly measured over time. These different variables were combined into the curve-of-factors model by treating the different variables as multiples indicators of the same construct and it is the intra-individual change in only this construct that is being modeled. In contrast, for the associate model and the factor-of-curves model, the different variables repeatedly measured over time were modeled as distinct constructs undergoing intra-individual changes, as represented by the different univariate LGMs combined into the multivariate model. Rather than describing it as a multivariate growth model, Chan (1998a) construed the curve-of-factor model as a multiple-indicator LGM and emphasized the value of the model in addressing fundamental questions on changes over time relating to different types of measurement errors and different issues of measurement invariance over time. As noted by Chan (1998a; 2002a,b) and others (e.g., Duncan, Duncan, & Strycker, 2006), early work on LGM has not considered issues of measurement errors and measurement invariance. Chan (1998a) showed how LGM can incorporate measurement errors and measurement invariance concerns in the model specification by extending the LGM to a multiple-indicator LGM in which the focal variable of change is modeled as a latent variable assessed by multiple indicators (i.e., a curve-of-factors model) as opposed to a manifest variable, typically the case in prior work on LGM. The use of multiple indicators in an LGM allows both random and nonrandom measurement errors to be taken into account when deriving the intercept and slope/shape factors. The use of multiple indicators to assess the focal construct allows reliable (nonrandom) variance to be partitioned into true score common (construct) variance and true score unique variance. True score unique variance is nonrandom and it is that portion of variance in a measure that is not shared with other measures of the same construct. In LGM, the same measures are repeatedly administered over time. Hence, a failure to partition nonrandom variance into true construct variance and unique variance leads to distorted (inflated) estimates of true change in the focal construct over time. Because only scale/composite level but no item-level (multiple-indicator) information on the focal variable is used in the standard LGM, the standard LGM procedure does not provide the isolation of nonrandom error variance from reliable variance and it takes only random errors into consideration. The use of multiple-indicator LGM addresses the problem. To understand measurement invariance over time, it is useful to refer to the three types of change distinguished in Golembiewski, Billingsley, and Yeager
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
112 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY (1976): alpha, beta and gamma changes. Alpha change refers to changes in absolute levels given a constant conceptual domain and a constant measuring instrument. For example, if organizational commitment was adequately measured both at Time 1 and Time 2 in terms of reliability and validity such that the same construct was measured at both time points and with the same precision, then the difference in the commitment scores between the two time points represent an alpha change in organizational commitment and the change may be directly interpreted as a change in the absolute level of organizational commitment. We can meaningfully speak of alpha change only when there is measurement invariance of responses across time. Measurement invariance across time exists when the numerical values across time waves are on the same measurement scale. Measurement invariance could be construed as absence of beta and gamma changes. Beta change refers to changes in absolute level complicated by changes in the measuring instrument given a constant conceptual domain. Beta change occurs when there is a recalibration of the measurement scale. That is, in beta change, the observed change results from an alteration in the respondent’s subjective metric or evaluative scale rather than an actual change in the construct of interest. For example, because of the rater’s increased leniency in ratings over time, a rating of 6 given at Time 2 may be defined by the rater as was rating of 5 at Time 1. Gamma change refers to changes in the conceptual domain. Gamma change (i.e., change in the meaning or conceptualization of the construct(s) of interest) can take a variety of forms. For example, in the language of factor analysis, the number of factors (a factor representing a construct) assessed by a given set of measures may change from one time point to another. To illustrate, in a study of changes in performance over time, performance may undergo a type of gamma change represented by factorial integration of performance measurement so that performance components (factors) become increasingly interrelated over time such that performance at early time points are best represented as multiple distinct and relatively uncorrelated factors, at mid time points are best represented as multiple highly correlated factors and at later time points are best represented as a single factor. Chan (1998a) demonstrated how the fundamental questions on measurement errors, measurement invariance, functional forms of intra-individual changes, and other fundamental questions on change over time may be answered in an integrative two-phase latent variable analytical procedure that combines longitudinal means and covariance structures analysis and multipleindicator LGM. In Phase 1 of the procedure, longitudinal mean and covariance analysis, which is similar to longitudinal factor analysis except that both the indictor intercepts and factor means are also estimated, is used to examine issues of measurement invariance across time and across groups. Establishing invariance provides evidence that results of subsequent growth modeling constituting Phase 2 of the procedure are meaningful. By building invariance assessments as the first logical step to longitudinal modeling, this integrative
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
113
procedure contrasts with the analytical models that left untested the assumption of measurement invariance across time or groups. In addition to invariance assessments, Phase 1 of the procedure helps in the preliminary assessment of the basic form of intra-individual change by identifying the constraints on the patterns of true score (factor) means and variances over time. In Phase 2, multiple-indicator LGM is used to directly assess change over time by explicitly and simultaneously modeling the group and individual growth trajectories of the focal variable as well as their relationships to other time-invariant predictors and/or time-varying correlates (i.e., growth trajectories in a different domain). As explained in Chan (1998a), longitudinal mean and covariance analysis and multiple-indicator LGM together provide a unified framework for directly addressing the various fundamental questions on change over time.
MULTIPLE GROUP ISSUES An adequate longitudinal assessment of job performance or work attitude variables should be able to answer questions on whether the intra-individual changes over time are occurring in same or different ways between groups and, if different, in what specific ways the groups differ. A powerful advantage of the LGM method (univariate or multivariate) is that, because it is implemented in the structural equation modeling framework, it allows multiple groups to be assessed simultaneously to test for between-group equality or differences in specific parameters of change. That is, LGMs can be fitted simultaneously to different groups of individuals and multiple-group LGM analyses can be performed to test for across-groups invariance of one or more of the specified relationships in the LGM. The groups under comparisons could be experimental groups (e.g., randomly formed groups of participants assigned to different task conditions) or natural occurring groups such as male and female incumbents. The question of interest is whether a specific intra-individual change pattern found in one group is equal to, or differs from, in either magnitude or form, the intra-individual change pattern in a different group. For example, the growth trajectory representing changes in performance on a given job over time may differ in functional form between male and female incumbents. Alternatively, males and females may share the same functional form but they differ in the rate of change in job performance. As another example, employees who have completed a training program may undergo a type of gamma change represented by factorial integration of performance measurement so that performance components (factors) become increasingly inter-related over time whereas employees who completed a different training program may exhibit factorial invariance so that inter-correlations among performance components remain constant over time. Another between-groups comparison question concerns whether change is uni-path or multi-path in each group. Change can be
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
114 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY represented as proceeding in one single pathway or through multiple different pathways. Multiple paths occur when a detour from a single growth trajectory path is possible as individuals proceed from one time point to another. For example, when proceeding from Time 1 to Time 4 (through Times 2 and 3), assume some individuals follow a linear trajectory but others follow a quadratic trajectory. Using the multiple-group LGM approach, we would be able to examine if the growth trajectories of distinct groups of individuals being tracked over time follow the same or different functional forms. The multiple-group LGM is highly flexible as it has the capability to simultaneously model and compare across distinct groups the various specific facets of intra-individual change patterns. Within the population of interest, if the subpopulations characterized by different intra-individual change patterns are observable subgroup membership variables that are known a priori such as groupings by demographics (e.g., sex, ethnicity), then the longitudinal assessment, as explained above, is quite easily performed by applying multiple-group LGMs to isolate the data according to the subgroup membership variable. However, if the subpopulations characterized by different intra-individual change patterns are unobserved (i.e., latent) in the sense that subgroup membership (how many subgroups and which subgroup does an individual belong to) is not known a priori but latent and empirically derived from the individual’s values on a set of variables, then it is not possible to perform a straightforward multiple-group LGM analysis because there is no known subgroup membership variable. Fortunately, with recent methodologic advances in growth modeling, we can now model such unobserved heterogeneity in the population. Specifically, a class of longitudinal techniques known as general growth mixture modeling developed by Muth´en (2004) offers an inclusive framework that combines latent growth models and latent class models. This general framework allows the researcher to identify latent classes characterized by different patterns of latent growth. These mixture models are useful because they allow us, in a single integrated analysis, to identify unobserved groups of individuals with qualitatively different growth trajectories by establishing the number of latent subpopulations, the distinct intra-individual change patterns associated with the latent subpopulations, and assigning latent subpopulation membership to individuals. For discussions on technical issues and an empirical example on the growth mixture modeling method, see Wang and Chang (in press). Growth mixture modeling is a flexible framework that could be applied to many areas of study in I/O psychology and the logic of the technique allows us to open up new and fruitful avenues for future research such as identifying latent subpopulations with distinct intra-individual change patterns in various domains such as task performance, OCBs, and withdrawal behaviors. Finally, latent subpopulations and observed groupings may be combined in a single analysis to examine more complex multiple group issues. Specifically, the growth mixture modeling method for examining latent subpopulations
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
115
may be extended to fit data simultaneously to multiple observed (known a priori) groups. For example, we can examine whether the number and characteristics of latent subpopulations are the same or different across gender groups, a multiple-group growth mixture modeling analysis may be conducted to estimate growth mixture models simultaneously for the male and female groups. The logic for the analysis is similar to the standard multiple-group latent variable method in which specific parameters may be fixed to equal or allowed to vary freely across multiple observed groups to produce different multiple-group growth models for nested model comparisons to determine the most appropriate multiple-group model (Chan, 1998a). With appropriate cross-cultural theories, this integrative analytic method combining growth mixture modeling (to identify latent subpopulations) and observed multiplegroup latent variable modeling provides a powerful method to examine new substantive research questions on cross-cultural differences in intra-individual changes in terms of possible cross-cultural differences in latent subpopulations of intra-individual change patterns.
CONCLUSIONS For many decades in I/O psychology, predictor–criterion relationships have been described in terms of static models without much attention paid to the temporal aspects of the predictor–criterion constructs including what and how changes may occur over time. Consider the example of job performance models. An individual’s job performance may change over time in various ways (e.g., increase/decrease in level, changes in the number/nature of underlying dimensions) and these intra-individual changes are important for understanding, predicting, and evaluating job performance. For example, when performance changes over time either in terms of level or dimensionality, using a sample of job incumbents with varying levels of job tenure in a validation study could affect and confound estimates of validity and the interpretation of predictor–criterion relationships. When there exists between-group differences, either in terms of observed subgroup membership or latent subpopulations, a longitudinal assessment method that assume population homogeneity in intra-individual change patterns will lead to incomplete or even misleading substantive inferences. Advances in longitudinal analytical strategies, especially those that involve latent variable modeling as explicated in this article, allow us to make more direct and better connections linking theory, measurement, data, and interpretation. As shown in this article, when suitably applied, the analytical advances provide us both the conceptual basis and statistical method to hypothesize, test, and interpret intra-individual changes over time in the predictor and criterion variables in the context of multilevel, multivariate, and multiple group issues, which in turn allow us to derive adequate substantive implications and
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
116 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY make effective practical recommendations concerning job performance and work attitudes.
REFERENCES Bliese, P.D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K.J. Klein, & S.W.J. Kozlowski (Eds), Multilevel Theory, Research, and Methods in Organizations (pp. 349–81). San Francisco, CA: Jossey-Bass. Bliese, P.D., Chan, D., & Ployhart, R.E. (2007). Multilevel methods: Future directions in measurement, longitudinal analyses and nonnormal outcomes. Organizational Research Methods, 10, 551–63. Bryk, A., & Raudenbush, S.W. (1992). Hierarchical Linear Models. Thousand Oaks, CA: Sage. Chan, D. (1998a). The conceptualization of change over time: An integrative approach incorporating longitudinal means and covariance structures analysis (LMACS) and multiple indicator latent growth modeling (MLGM). Organizational Research Methods, 1, 421–83. Chan, D. (1998b). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83, 234–46. Chan, D. (2000). Conceptual and empirical gaps in research on individual adaptation at work. In C. Cooper, & I. Robertson (Eds), International Review of Industrial and Organizational Psychology. Chichester, UK: John Wiley. Chan, D. (2002a). Latent growth modeling. In F. Drasgow, & N. Schmitt, Measuring and Analyzing Behavior in Organizations: Advances in Measurement and Data Analysis (pp. 302–49). CA: Jossey-Bass. Chan, D. (2002b). Longitudinal modeling. In S. Rogelberg (Ed.), Handbook of Research Methods in Industrial and Organizational Psychology (pp. 412–30). Malden, MA: Blackwell Publishers, Inc. Chan, D. (2005). Multilevel research. In F. Leong, & J. Austin (Eds), The Psychology Research Handbook (2nd edn), (pp. 401–18). Thousand Oaks, CA: Sage. Chan, D. (2011). Advances in analytical strategies. In S. Zedeck (Ed.), APA Handbook of Industrial and Organizational Psychology (volume 1), (pp. 85–113). Washington, DC: APA. Chan, D. Ramey, S., & Ramey, C., & Schmitt, N. (2000). Modeling intraindividual changes in children’s social skills at home and at school: A multivariate latent growth approach to understanding between-settings differences in children’s social skills development. Multivariate Behavioral Research, 35, 365–96. Chan, D., & Schmitt, N. (2000). Inter-individual differences in intra-individual changes in proactivity during organizational entry: A latent growth modeling approach to understanding newcomer adaptation. Journal of Applied Psychology, 85, 190–210. Duncan, T.E., Duncan, S.C., & Strycker, L.A. (2006). An Introduction to Latent Variable Growth Curve Modeling (2nd edn). Mahwah, NJ: LEA. Edwards, J.R. (1994). The study of congruence in organizational behavior research: Critique and a proposed alternative. Organizational Behavior and Human Decision Processes, 58, 51–100. Golembiewski, R.T., Billingsley, K., & Yeager, S. (1976). Measuring change and persistence in human affairs: Types of change generated by OD designs. Journal of Applied Behavioral Science, 12, 133–57. Guzzo, R.A., Yost, P.R., Campbell, R.J., & Shea, G.P. (1993). Potency in groups: Articulating a construct. British Journal of Social Psychology, 32, 87–106.
P1: OTA/XYZ P2: ABC c03 JWST046-Hodgkinson
January 21, 2011
12:56
Printer Name: Yet to Come
LONGITUDINAL ASSESSMENT OF CHANGES IN JOB PERFORMANCE
117
Hofmann, D.A., Griffin, M.A., & Gavin, M.B. (2000). The application of hierarchical linear modeling to organizational research. In K.J. Klein, & S.W.J. Kozlowski (Eds), Multilevel Theory, Research, and Methods in Organizations (pp. 467–511). San Francisco, CA: Jossey-Bass. Hofmann, D.A., Jacobs, R., & Baratta, J. (1993). Dynamic criteria and the measurement of change. Journal of Applied Psychology, 78, 194–204. Joreskog, K.G., & Yang, F. (1996). Nonlinear structural equation models. The Kenny–Judd model with interaction effects. In G.A. Marcoulides, & R.E. Schumacker (Eds), Advanced Structural Equation Modeling: Issues and Techniques (pp. 57–88). Hillsdale, NJ: LEA. Kozlowski, S.W.J., & Klein, K.J. (2000). A multilevel approach to theory and research in organizations: Contextual, temporal, and emergent processes. In K.J. Klein, & S.W.J. Kozlowski (Eds), Multilevel Theory, Research, and Methods in Organizations (pp. 3–90). San Francisco, CA: Jossey-Bass. Kristof, A.I. (1996). Person-organization fit: An integrative review of its conceptualizations, measurement, and implications. Personnel Psychology, 49, 1–49. McArdle, J.J. (1988). Dynamic but structural equation modeling of repeated measures data. In R.B. Catell, & J. Nesselroade (Eds), Handbook of Multivariate Experimental Psychology (2nd edn), (pp. 561–614). New York: Plenum. Morgeson, F.P., & Hofmann, D.A. (1999). The structure and function of collective constructs: Implications for research and theory development. Academy of Management Review, 24, 249–65. Muth´en, B. (2004). Latent variable analysis: Growth mixture modeling and related techniques for longitudinal data. In D. Kaplan (Ed.), Handbook of Quantitative Methodology for the Social Sciences (pp. 345–68). Newbury Park, CA: Sage Publications. Rousseau, D.M. (1985). Issues of level in organizational research: Multi-level and crosslevel perspectives. In B.M. Staw, & L.L. Cummings (Eds), Research in Organizational Behavior (pp. 1–7). Greenwich, CT: JAI Press. Singer, J.D., & Willett, J.B. (2003). Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York: Oxford University Press. Wang, M., & Chan, D. (in press). Mixture latent Markov modeling: Identifying and predicting unobserved heterogeneity in longitudinal qualitative status change. Organizational Research Methods. Wen, Z., Marsh, H.W., & Hau, K.T. (2002). Interaction effects in growth modeling: A full model. Structural Equation Modeling, 9, 20–39.
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
Chapter 4 ESTIMATING THE RELATIVE IMPORTANCE OF VARIABLES IN MULTIPLE REGRESSION MODELS Dina Krasikova and James M. LeBreton Department of Psychological Sciences, Purdue University, West Lafayette, IN, USA
Scott Tonidandel Department of Psychology, Davidson College, Davidson, NC, USA Organizational scholars continue to be interested in examining relative importance of variables in multiple regression analysis (cf. Eby, Durley, Evans, et al., 2006; Luo, 2005; McAllister, Kamdar, Morrison, et al., 2007; Schleicher, Venkataramani, Morgeson, et al., 2006). Although the search for the proper measure of importance has been conducted for decades (e.g., Engelhart, 1936), variable importance in multiple regression contexts has traditionally been examined using bivariate correlation coefficients, standardized regression coefficients (i.e., beta-weights), squared standardized regression coefficients (i.e., squared beta-weights), or product measures (i.e., the products of the correlation and standardized regression coefficients; Budescu, 1993; Johnson & LeBreton, 2004). However, these measures of relative importance are problematic when researchers are testing the importance of correlated variables, which is often the case in the organizational sciences (Johnson & LeBreton, 2004). More specifically, when used as measures of relative importance in regression models with correlated predictors, these indices do not correctly partition criterion variance (LeBreton, Ployhart, & Ladd, 2004) and therefore provide distorted estimates of predictor importance. To address this statistical deficiency, Budescu (1993) proposed general dominance analysis and Johnson (2000) introduced relative weight analysis. These two statistical approaches have become widely accepted by an audience of organizational researchers and practitioners. Multiple extensions of Budescu’s International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
120 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY and Johnson’s original work on relative importance were proposed in the subsequent decades (e.g., Azen & Budescu, 2003, 2006; Azen & Traxel, 2009; LeBreton & Tonidandel, 2008; Tonidandel & LeBreton, 2010; Tonidandel, LeBreton, & Johnson, 2009). We believe that a single source synthesizing our knowledge about relative importance analysis accumulated over years of research is much needed. Therefore, the purpose of this chapter is to provide an integrative review of the extant literature on relative importance analysis with the emphasis on dominance analysis and relative weight analysis. The chapter is structured as follows. After examining the concept of variable importance in a regression context, we overview importance measures traditionally used in organizational research and discuss their theoretical and practical limitations. Then, we focus on the two accepted techniques designed to measure importance of correlated predictors – dominance analysis and relative weight analysis – and discuss recent developments in relative importance research. We conclude with suggestions for future work.
VARIABLE IMPORTANCE IN MULTIPLE REGRESSION Prediction and Explanation as Applications of Multiple Regression Models Multiple regression is widely used in organizational research and practice (Austin, Scherbaum, & Mahlman, 2002; Johnson, 2000) for the purposes of prediction and explanation (Courville & Thompson, 2001; Johnson & LeBreton, 2004; Pedhazur, 1997). When researchers and practitioners rely on multiple regression for primarily predictive purposes, they are concerned with maximizing the predictive power of the regression model and looking for a set of predictors that are able to explain the largest amount of variance in the criterion (e.g., Kerlinger & Pedhazur, 1973). As Courville and Thompson (2001) noted, “When our research application is purely predictive, in a sense interpretation may be irrelevant. We may very much desire an accurate prediction, but we may not care why our predictive rule works” (Courville & Thompson, 2001: 232). Such an application of regression strikes of dust bowl empiricism and thus, not surprisingly, the process of model selection is often exploratory (Azen & Budescu, 2003). This process involves the following steps: 1. Specifying the regression equation for the current sample where data are available for the set of predictors and the criterion variable. 2. Estimating regression coefficients for the predictors. 3. Applying the same regression coefficients in similar samples to predict criterion scores. Often, researchers rely on the formal model selection methods that help identify the model with the greatest predictive ability from the subset of the available
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
121
alternatives (Kutner, Nachtsheim, Neter, et al., 2005). The predictive power of the model is typically measured by R2 indicating the proportion of variance in the criterion accounted for by the linear combination of predictors. In contrast, researchers relying on multiple regression for primarily theory testing or explanatory purposes are interested in examining the contribution of each predictor in explaining variance in the criterion relative to other predictors (Johnson & LeBreton, 2004; Kerlinger & Pedhazur, 1973). In such instances, prediction remains important, but the emphasis is on understanding of why the regression equation “works.” Thus, interpretation of the regression model becomes the primary focus of researchers. As Johnson and LeBreton (2004) noted, “In this case, we are interested in the extent to which each variable contributes to the prediction of the criterion . . . interpretation is the primary concern, such that substantive conclusions can be drawn regarding one predictor with respect to another” (Johnson & LeBreton, 2004: 239). When regression is used for explanatory purposes the focus shifts from searching for the most predictive model to interpreting relationships among predictor variables and the criterion and understanding to what extent each predictor variable drives the overall prediction. It should be noted that these two applications of regression are not mutually exclusive and are often used together (Azen & Budescu, 2003). This is especially true in organizational sciences where researchers are typically interested in both identifying the most predictive model and evaluating the usefulness of predictors in explaining criterion variance (LeBreton, Hargis, Griepentrog, et al., 2007). However, regardless of how explanation is used, independently or as a follow-up to the prediction stage, it involves the analysis of each predictor’s importance in the model. On the Meaning of Variable Importance in Regression Analysis Evaluating the importance of variables in a multiple regression is a complicated and multifaceted process involving a host of factors including issues related to the construct validity of the predictors, cost associated with data collection, fairness, organizational goals and values, and participant (and organizational) acceptance (LeBreton, Hargis, Griepentrog, et al., 2007). Organizational scientists often turn to statistical criteria to help guide inferences concerning variable importance (e.g., Society for Industrial and Organizational Psychology, 2003) such as incremental importance and the relative importance of variables in the multiple regression (LeBreton, Hargis, Griepentrog, et al., 2007). Incremental importance Assessment of incremental importance allows researchers to evaluate the unique contribution a variable makes in explaining criterion variance, above and beyond other variables in the model (i.e., importance is defined in terms
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
122 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY of a variable’s “usefulness”; Darlington, 1968). Incremental importance can be measured as change in R2 (R2 ) indicating the amount of additional variance in criterion accounted for by including a new variable into a model already containing a set of predictors. Alternatively, incremental importance of a variable is sometimes examined in terms of the significance of the regression coefficient obtained during a simultaneous entry regression analysis. Understanding a variable’s incremental importance is useful when we need to determine whether the new variable is redundant with the other variables in the model (i.e., whether it captures unique variance in the criterion; Cronbach & Gleser, 1957; Sechrest, 1963). Inferences concerning incremental importance are influenced by variable intercorrelations and the order in which variables are entered into the model. Specifically, the proportion of criterion variance explained by both the new variable and the existing set of variables is automatically attributed to the latter (LeBreton, Hargis, Griepentrog, et al., 2007; Tabachnick & Fidell, 2007). Thus, the total contribution this variable makes to the predicted criterion variance is likely to be higher than its R2 in the presence of other correlated predictors. In other words, small increments in R2 may mask more meaningful contribution of the predictor if this predictor is taken alone, and important predictors may appear less so when we rely solely on incremental importance analysis (LeBreton, Hargis, Griepentrog, et al., 2007). Also, if the order of variable entry changes, R2 associated with the particular variable may be quite different, depending on at what step this variable is added to the model. Thus, incremental importance analysis is useful only when we are specifically concerned with the additional, incremental contribution of the predictor beyond the set of existing predictors and/or we have a very specific a priori ordering of the variables. In sum, when our purpose is to compare predictors in terms of their overall contribution to explaining criterion variance, incremental importance analysis may provide misleading conclusions. Relative importance may provide a more accurate assessment of a variable’s overall explanatory contribution. Relative importance Johnson and LeBreton (2004) defined relative importance as “the proportionate contribution each predictor makes to R2 , considering both its direct effect (i.e., correlation with the criterion) and its effect when combined with the other variables in the regression equation” (i.e., beta-weight; Johnson & LeBreton, 2004: 240). This definition highlights the important properties of the relative importance analysis that distinguish it from the analysis of incremental importance. First, the former provides for a global assessment of a variable’s contribution to the total predictive value of the regression model considering its predictive power “in isolation from and in combination with other variables” (LeBreton, Hargis, Griepentrog, et al., 2007: 481). This property of relative importance analysis permits researchers to obtain answers to such
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
123
questions as: What variable(s) is(are) driving the R2 ? How does my new variable contribute to R2 ? How can I interpret the contribution of all my variables, including my new variable, to R2 ? How can I decompose the total predictive value of my regression model into the constituent variables? Does the relative contribution of my new variable completely dominate the relative contribution of the other variables? Second, newer measures of relative importance were specifically designed to address issues of importance in the face of correlated predictors. Incremental importance analysis and relative importance analysis examine different but complementary aspects of variable importance in multiple regression models. LeBreton, Hargis, Griepentrog, et al. (2007) suggested that organizational researchers would benefit from examining both incremental and relative importance which taken together may provide a more complete picture of the predictor usefulness in regression analysis. We now turn to the discussion of measures of relative importance historically used by organizational researchers.
HISTORICAL MEASURES OF RELATIVE IMPORTANCE Given the widespread interest in assessing variable importance in regression models, methodologists have been searching for the appropriate indices of relative importance (for a review of the history of relative importance research see Johnson & LeBreton, 2004). Historically, bivariate correlations, standardized regression coefficients (beta-weights), and product measure weights have been used by researchers. Each approach is limited in its ability to determine relative importance. Bivariate correlations (rxy ) and squared bivariate correlations (rxy 2 ) are inappropriate for evaluating relative importance because, by definition, they inform us about the bivariate relationship between each predictor and the criterion and ignore information about the other predictors included in the regression model. Use of rxy and rxy 2 to index relative importance would only be appropriate when predictors are orthogonal. In this case, proportions of criterion variance explained by the predictors do not overlap and rxy 2 becomes the true measure of relative importance (Johnson, 2000; Johnson & LeBreton, 2004). Beta-weights (β x ) and squared beta-weights (β x 2 ) are the most commonly used measures of relative importance (Darlington, 1990). Similarly to bivariate correlations, they are inappropriate for assessing relative importance when predictors are intercorrelated. In regression analysis, the overlapping criterion variance explained by correlated predictors is typically credited to the predictor with the largest bivariate correlation. More generally, collinearity among the predictors can lead to unstable estimates of variable importance. Thus, the
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
124 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Table 4.1
Y X1 X2 X3 X4
Problems with squared betas as measures of relative importance when predictors are intercorrelated
1.00 0.40 0.40 0.35 0.35
1.00 0.60 0.60 0.60
Correlation Matrix
Relative Importance (β x 2 )
1.00 0.60 0.60
0.0386 0.0386 0.0051 0.0051
1.00 0.60
1.00
magnitude of the regression coefficient attached to the predictor may provide misleading information regarding the overall contribution of the predictor into the model. For example, it is possible that a predictor having large positive bivariate correlation with the criterion will be assigned a near-zero or uninterpretable negative regression weight due to its overlap with the other predictors (Courville & Thompson, 2001; Darlington, 1968; Thomas, Hughes & Zambo, 1998). Regression coefficients provide information about the incremental importance of a variable, not relative importance. Finally, similarly to correlation coefficients, standardized regression coefficients are only appropriate for estimating relative importance when predictors are orthogonal. In this case, beta-weights are equal to bivariate correlation coefficients, and squared betaweights sum up to R2 and can be unambiguously interpreted as measures of relative importance. Table 4.1 presents some hypothetical data along with estimates of relative importance obtained using squared beta-weights. Here, we regress Y on four correlated variables and obtain estimates of importance. Based on squared beta-weights, we would infer that variables X 1 and X 2 are roughly seven times more “important” than X 3 and X 4 . A simple examination of the bivariate correlations reveals very small differences among the variables. Thus, the differences in relative importance are being driven by two factors: 1. predictors are intercorrelated; 2. the variables with the largest bivariate correlations are being given too much predictive credit while those with the smaller correlations are being given too little credit. This phenomenon is further illustrated in Table 4.2. In this example, we have slightly modified the pattern of criterion-related validity and variable intercorrelations. Relying on squared beta-weights we would infer that X 1 and X 2 are roughly 16 times more “important” than X 4 and 672 times more important than X 3 . Collinearity contracts the relative importance of variables X 3 and X 4 . Thus, betas and squared betas provide true estimates of relative importance only when the predictors are orthogonal, which is seldom the case in organizational research and practice.
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
125
VARIABLES IN MULTIPLE REGRESSION MODELS Table 4.2
Problems with squared betas and product measure weights as measures of relative importnace when predictors are intercorrelated Relative Importance
Y X1 X2 X3 X4
1.00 0.40 0.40 0.30 0.30
1.00 0.70 0.60 0.70
Correlation Matrix
βx2
rxy * β x
1.00 0.60 0.70
0.0672 0.0672 0.0001 0.0043
0.1037 0.1037 0.0026 –0.0197
1.00 0.30
1.00
Another traditional estimate of relative importance is the product measure. Product measure weights (Bring, 1996; Hoffman, 1960) are the multiplicative products of zero-order correlations and standardized regression coefficients (rxy * β x ). Product measure weights have at least one major advantage over bivariate correlations and beta-weights – they sum to the model R2 and may be interpreted as relative effect sizes even when the predictors are correlated. When predictors in the regression model are orthogonal, product measure weights yield the same relative importance estimates as squared bivariate correlations and squared beta-weights. Product measure weights share the limitations of both measures. Like bivariate correlations, product measure weights may indicate low predictor importance when predictor–criterion correlation is very low but the variable nonetheless adds predictive value to the model (e.g., if the variable is a suppressor; Cohen & Cohen, 1983). Like regression coefficients, product measure weights frequently have near-zero values when the corresponding predictor–criterion correlations are of substantial magnitude (Darlington, 1968). In other words, when one component of the product measure weight (β x or rxy ) is low, the product measure weight ignores the magnitude of another component (Johnson & LeBreton, 2004). This contradicts the definition of relative importance stating that an index of relative importance should consider the predictive value of the variable both in isolation and in the presence of other variables (Johnson & LeBreton, 2004). Another problem with product measure weights is that they may yield uninterpretable (i.e., negative) estimates of importance (Johnson & LeBreton, 2004). This problem is illustrated by estimating the product measure weights for the data presented in Table 4.2. Here we see that variable X 4 has negative importance. Thus, according to the product measure, X 4 explains negative variance in the criterion variable. Given the limitations of the historical measures of relative importance, we recommend that researchers rely on more recent advancements in estimates of relative importance such as general dominance weights and relative weights.
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
126 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
MODERN RELATIVE IMPORTANCE STATISTICS Dominance Weights Dominance weights are defined as average usefulness or average contribution of a predictor to the model R2 across all possible subset regressions (Azen & Budescu, 2003; Budescu, 1993). General dominance weights (Cj ) are calculated by averaging predictor’s squared semi-partial correlations (R2 ) obtained in all possible subset regression models. For example, in order to obtain a general dominance weight for X1 , the following steps should be taken (LeBreton & Tonidandel, 2008: 330): 1. Calculate R2 for X1 when X1 is added to the model with no predictors. R2 in this case will be equal to r 2 YX1 , the squared zero-order correlation between X1 and Y. 2. Calculate R2 for X1 when X1 is added to the model with only X2 . In this case, R2 will be equal to r 2 Y(X1. X2) , the squared first-order semi-partial correlation between X1 and Y , after controlling for X2 . 3. Calculate R2 for X1 when X1 is added to the model with only X3 . R2 will be equal to r 2 Y(X1. X3) , the squared first-order semi-partial correlation between X1 and Y , after controlling for X3 . 4. Calculate R2 for X1 when X1 is added to the model with both X2 and X3 . At this step, R2 will be equal to r 2 Y(X1. X2, X3) , the squared second-order semi-partial correlation between X1 and Y , after controlling for both X2 and X3 . 5. Average values of R2 obtained in steps 1–4 to obtain the general dominance weight for X1 . General dominance weights for X2 and X3 are computed analogously. Because the variables’ general dominance weights sum up to the model R2 , each individual weight may be interpreted as a relative effect size (LeBreton, Hargis, Griepentrog, et al., 2007). On a related point, it is possible to calculate rescaled general dominance weights obtained by dividing corresponding individual weights by the model R2 (i.e., Cj /R2 ). Such rescaled weights represent the proportion (or percentage, if these weights are further multiplied by 100) of total variance in the criterion attributed to each predictor. Rescaled dominance weights are often useful when communicating with managers or executives who may find statistical discussions confusing or difficult to follow. That is, explaining to an executive that some variable (e.g., extraversion) accounts for 35% of the predicted variance in the criterion (e.g., sales performance) is going to be better received than a discussion of validity coefficients, standardized regression coefficients, product measure weights, and/or increments in R2 (LeBreton, Hargis, Griepentrog, et al., 2007). In addition to evaluating magnitudes of relative importance weights, dominance analysis permits examining patterns of dominance. Azen and Budescu
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
127
(2003) identified three patterns of dominance: complete dominance, conditional dominance, and general dominance. Complete dominance occurs when a variable yields a larger increment in R2 than another variable across all subset regression models containing both variables (Azen & Budescu, 2003; Budescu, 1993). Thus, if the regression model includes four predictors X1 , X2 , X3 , and X4 , X1 is said to “completely dominate” X2 if X1 has a larger squared semipartial correlation compared with X2 across all possible models that include both of those variables. Alternatively, X1 might dominate X2 in some models (e.g., in models that include only X3 or X4 ) but not in others (e.g., in models that include both X3 and X4 ). This pattern of dominance is referred to as conditional dominance and is said to occur when “the average additional contribution within each model size is greater for one predictor than the other” (Azen & Budescu, 2003: 136), where model size is the number of variables included in the subset regression model. Finally, examination of general dominance simply involves estimating weights described above (i.e., examining the average squared semipartial correlation across all possible subset regression models). Dominance analysis is used to identify variables that tend to “outperform” other variables across the various regression models. Dominance analysis has a number of advantages over the traditional measures of relative importance: 1. it was specifically developed for the use with correlated predictors; 2. it provides estimates of relative importance that sum to the model R2 ; 3. its weights can be rescaled and easily communicated to organizational decision makers; 4. it partitions predicted variance in a manner that is consistent with contemporary definitions of relative importance; and 5. it permits researchers to examine various patterns of dominance. However, obtaining dominance weights may become computationally intensive. Specifically, because a dominance analysis requires all subset regression models, the number of models rapidly increases with the increase of the model size. For example, a dominance analysis containing 20 predictors involves computing 1 048 575 regressions. When researchers have to deal with large regression models, they may benefit from conducting a relative weight analysis. Relative Weights Johnson (2000) described a procedure for deriving relative weights (RWj , or epsilon weights, εj ) using a variable transformation approach (e.g., Gibson, 1962; Green, Carroll, & DeSarbo, 1978; Johnson, 1966). His approach involves creating a new set of uncorrelated predictors (ZXk ) that are maximally related to the original set of correlated predictors (Xj ). Both ZXk and Xj are then used to produce importance estimates. For example, if a criterion variable
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
128 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY λ11
X1
λ12
Zx1
β1
λ13 λ21 λ22
X2
λ23
β2 Zx2 β3
λ31 X3
Y
λ32
Zx3
λ33 Figure 4.1 tors.
Calculation of univariate relative weights for the regression model with three predic-
is regressed on three variables, the computation of the relative weight for X1 will involve the following steps (see also Figure 4.1): 1. Derive a set of k orthogonal weights ZXk that are maximally related to the set of j original predictors Xj . 2. Obtain a set of standardized regression coefficients β k by regressing the criterion variable Y on the set of the newly created orthogonal predictors ZXk . 3. Obtain a set of standardized regression coefficients λjk by regressing Xj on the set of the newly created orthogonal predictors ZXk . 4. Compute relative weights by summing the products of squared standardized regression coefficients β k 2 and λjk 2 obtained at steps 2 and 3. Thus, the relative weight for the first predictor (X1 ) will be calculated as ε1 = β 1 2 λ11 2 + β 2 2 λ12 2 + β 3 2 λ13 2 . Relative weights for the other predictors in the model are calculated in a similar manner. Additionally, rescaled relative weights indicating the proportions of criterion variance accounted for by each predictor can be computed by dividing relative weights by the model R2 (i.e., RWj /R2 ). The primary limitation of the relative weight analysis is that it is typically conducted on the full model containing all predictors. This does not allow for identifying pairwise patterns of variable importance analogous to complete and conditional patterns of dominance. However, relative weights have multiple advantages over other measures of relative importance as the weights: 1. are specifically developed for use with correlated predictors; 2. sum to the R2 and thus can be used as estimates of relative effect sizes;
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
129
3. are easily rescaled into proportions or percentages when communicating with managers or executives; and 4. conducting a relative weight analysis is more computationally efficient than conducting a dominance analysis. Thus, both dominance analysis and relative weight analysis are designed to deal with predictor collinearity (using an all subset regression approach and a variable transformation approach, respectively) and both furnish estimates that sum to the R2 . The question then becomes, to what extent dominance weights and relative weights converge and how they relate to the other measures of relative importance. Convergence of Dominance Weights and Relative Weights Johnson (2000) examined patterns of convergence and divergence among different estimates of relative importance using as an example a regression model with ratings of various traits (“ratee technical proficiency, dependability, friendliness, obnoxiousness, and the extent to which the ratee is seen as a show-off”; Johnson, 2000: 10) predicting peer-ratings of overall job performance. Results revealed substantial convergence among general dominance weights and relative weights. Both the magnitudes of relative importance estimates and the pattern of variable importance obtained as a result of dominance analysis and relative weight analysis were virtually identical. However, estimates yielded by these newer techniques deviated substantially from the results obtained using squared correlations, squared betas, and the product measures. Similar conclusions were reached by LeBreton, Ployhart, and Ladd (2004) who performed a large Monte Carlo comparison of the existent relative importance methods (i.e., dominance weights, relative weights, squared correlations, squared betas, and product measures) under a variety of conditions that often complicate multiple regression analysis. The authors manipulated such factors as criterion-related validity of predictors, collinearity among predictors, and total number of predictors in a regression model, and examined the impact of these factors on the consistency in predictor rank orders obtained using different indices of relative importance. Kendall’s tau correlation was used as a measure of convergence among relative importance methodologies. Across the experimental conditions, dominance weights demonstrated the greatest convergence with relative weights and the lowest convergence with beta-weights. The predictor rank orders furnished via dominance weights had the highest (above 0.90) Kendall’s tau correlation with those obtained via relative weights across levels of criterion validity, magnitudes of predictor collinearity, and various numbers of predictors included in the regression model. Predictor rank orders obtained in dominance analysis and relative weight analysis remained relatively consistent across the levels of the manipulated factors. Only small decreases in convergence were observed with the increase in mean
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
130 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY criterion validity, predictor collinearity, and number of predictors included in the regression model. Other indices of relative importance (i.e., squared correlations, squared betas, and product measures) demonstrated lower and less stable convergence with dominance weights. Specifically, the discrepancies between dominance weights and squared correlations, squared betas, and product measures increased with an increase in the level and dispersion of criterion-related validity, an increase in the level and dispersion of variable intercorrelations, and an increase in the number of variables included in the model. After a thorough review of their findings, LeBreton, Ployhart, and Ladd (2004) advised organizational researchers to rely on dominance analysis and relative weight analysis when examining relative importance, especially when they are dealing with at least moderate levels of R2 and have more than three predictors that are moderately correlated with each other. Given the high levels of convergence between general dominance weights and relative weights, the choice between the two depends on several issues (LeBreton, Hargis, Griepentrog, et al., 2007; LeBreton & Tonidandel, 2008). If researchers are interested in establishing patterns of complete or conditional dominance, then dominance analysis is the only choice. If researchers are seeking to establish relative importance for a large number of predictors, then they are advised to use relative weights because dominance analysis can become computationally burdensome with a large number of predictors. If researchers simply seek to obtain estimates of relative importance for a modest number of variables (e.g., 10 or less) then either dominance analysis or relative weight analysis should suffice.
RECENT DEVELOPMENTS IN RELATIVE IMPORTANCE RESEARCH In the past few years, relative importance statistics have been extended for use with the more complex variations of multiple regression analysis. These new developments include estimating the significance of relative importance estimates, and estimating variable importance in multivariate and logistic regression models. Below, we discuss these newer development before offering some concluding comments and suggestions for future research. Examining Significance of Relative Importance Estimates In addition to comparing the importance of a particular variable relative with other predictors in the model, a researcher may also wish to claim that a particular variable is meaningful. For a predictor to be meaningful, it must contribute to R2 over and above chance levels. To address this question, individuals have typically relied on the statistical significance of regression
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
131
coefficients. Courville and Thompson (2001) demonstrated that the statistical significance of regression weights is a flawed metric for evaluating whether a variable is a meaningful predictor because of this statistic’s inability to accurately partition shared variance. In fact, in a survey of articles published in the Journal of Applied Psychology over an 11-year span, their re-analysis of published data revealed numerous instances where authors incorrectly claimed that particular variables were not meaningful because of a mistaken reliance on the statistical significance of regression coefficients. Because of the limitations associated with regression weights for making such a determination, Tonidandel, LeBreton, and Johnson (2009) suggested that a better metric for these judgments would be the statistical significance of relative importance weights. Tonidandel, LeBreton, and Johnson (2009) outlined a procedure that uses bootstrapping to evaluate the statistical significance of relative weights because the sampling distribution of relative weights is not known. The logic of their approach involves evaluating whether the relative weight associated with a particular predictor is significantly different from the relative weight produced by a predictor that is known to be unimportant (i.e., a randomly generated variable). If a predictor’s relative weight is significantly different from the relative weight produced by a randomly generated variable, then the predictor can be said to be meaningful (i.e., contribute to R2 beyond chance levels). The specific steps for evaluating the statistical significance of a relative weight are the following: 1. Add a randomly generated variable to the original data set. 2. Use nonparametric sampling with replacement to create a large number (e.g., 1000) of bootstrapped data sets of the original data. 3. Calculate the relative weights for all predictors and the randomly generated variable in each of the bootstrapped data sets. 4. Calculate the differences between predictor relative weights and the relative weight for the randomly generated variable. 5. Construct a confidence interval around each difference between relative weights. These confidence intervals can be constructed using the bias corrected accelerated (BCa) method of obtaining confidence intervals described in detail by Efron and Tibshirani (1993). The null hypothesis is that the difference between the relative weight of the predictor and the randomly generated variable is zero. If a particular relative weight is to be deemed as statistically significant, the confidence interval around the difference between the relative weight of the meaningful variable and the relative weight produced by a random variable will not include zero. Tonidandel, LeBreton, and Johnson (2009) conducted a large-scale simulation study to evaluate this procedure. Their main conclusions were that the significance tests of relative weights: (i) provide overly conservative Type I error protection; (ii) provide meaningful levels of statistical power under most
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
132 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY circumstances; and (iii) display little evidence of bias. Given the limitations of regression weights and their corresponding test for significance, these authors recommend that researchers conduct significance testing of relative weights as a supplement to the more commonly performed significance tests of correlation and regression coefficients. This approach was recently implemented by Tonidandel, Braddy, and Fleenor (manuscript in preparation) in a study examining the relative importance of four managerial skill dimensions for predicting leader effectiveness. In an initial regression analysis, only one of the four managerial skill dimensions had a statistically significant regression coefficient. In contrast, all four managerial skills were associated with statistically significant relative weights. In the regression analysis, because of the correlations among the four skills, most of the shared variability among the skill dimensions tended to be assigned to a single predictor, thus masking the true contribution of the other three skills. A likely, but incorrect, conclusion based on the statistical significance of the regression weights is that only one of the four skill dimensions was meaningful. The statistical significance of the corresponding relative weights, on the other hand, more accurately identified all four skill dimensions as important predictors of managerial effectiveness. A similar approach can be used to evaluate whether the relative importance estimates of two predictors are significantly different from one another. For example, one might want to argue that a particular predictor is significantly more important than another predictor in the model, or one might wish to claim that a particular predictor is significantly more important in one sample than in another sample. Addressing such questions requires application of a bootstrapping procedure. However, instead of computing the difference between the relative weight of a predictor and the relative weight of a randomly generated variable, these questions require calculating the difference between the relative weights of two meaningful predictors. Johnson (2004) introduced a bootstrapping procedure that produces confidence intervals around the difference between relative weights from two predictors within a single sample or relative weights from the same predictor across different samples, while Tonidandel, LeBreton, and Johnson (2009) offered some updated recommendations. Estimating Predictor Importance in Multivariate Regression Models Azen and Budescu (2006) and LeBreton and Tonidandel (2008) extended general dominance weight analysis and relative weight analysis, respectively, to multivariate multiple regression designs that attempt to model relationships between a set of j predictor variables and a set of q criterion variables. Such designs are often encountered in the organizational sciences where researchers are studying criterion spaces that are inherently multidimensional (e.g.,
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
133
counterproductive work behaviors, organizational commitment, transformational leadership, job satisfaction). Multivariate dominance analysis Multivariate dominance analysis is analogous to its univariate counterpart in the sense that both univariate and multivariate dominance weights measure importance of each predictor taken individually and in the presence of other predictors, and are calculated using the all possible subset regressions method (Azen & Budescu, 2006). The critical difference between the two types of dominance analysis is the measure of linear association between predictors and criteria that underlies computation of dominance weights. Calculation of univariate dominance weights requires determining the contribution of each predictor in explaining variance in the single criterion and therefore relies on the use of the univariate measure of linear association (i.e., R2 ). Multivariate dominance analysis is used with models including more than one criterion. Therefore, calculation of multivariate dominance weights involves decomposition of variance in the q-dimensional criterion space (rather than decomposition of the R2 based on a single-array criterion variable) into the proportions of variance accounted for by each predictor and relies on the use of a multivariate analog of R2 . Azen and Budescu (2006) recommended multivariate measures of linear association – R2 XY and P 2 YX – that can be used in multivariate dominance analysis. These measures are bounded (i.e., vary between 0 and 1, the lower and the upper bounds of model fit), invariant to linear transformations of the variables, do not decrease with the inclusion of additional predictors into the model, are sensitive to correlations among the criterion variables, and therefore can be considered analogous to the univariate measure of linear association (i.e., R2 ) and used for computing multivariate dominance weights (Azen & Budescu, 2006). These authors noted that P 2 YX is most appropriately used with asymmetric research questions (i.e., when the direction of prediction in the regression model matters), whereas R2 XY can be used with both symmetric and asymmetric research questions (i.e., when the prediction direction may be reversed and reciprocal effects among the two sets of variables may be considered). Once the multivariate analog of R2 is chosen, the importance of each predictor is estimated as in the univariate case. Change in R2 XY or P 2 YX associated with the newly added predictor is first calculated for each subset regression, and then all values of R2 XY or P 2 YX are averaged across models to obtain a multivariate dominance weight for this predictor. Azen and Budescu (2006) caution researchers against inferring multivariate dominance from multiple univariate dominance analyses. The authors demonstrated that, similarly to other univariate procedures that may yield spurious conclusions about multivariate relationships (cf. Tabachnick & Fidell, 2007),
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
134 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY “univariate dominance does not necessarily imply multivariate dominance” (Azen & Budescu, 2006: 168). Multivariate relative weight analysis Multivariate relative weight analysis is based on the same logic as its univariate counterpart but requires a few additional steps to account for the collinearity among the criterion variables. In multivariate regression models, both predictors and criteria are intercorrelated, therefore multivariate relative weight analysis requires that a variable transformation (and back translation) approach is applied to both sets of variables. LeBreton and Tonidandel (2008) provided a complete matrix algebra derivation of multivariate relative weight analysis; however, the basic process unfolds as follows: 1. Create a new set of orthogonal variables ZYp that are maximally related to the original set of criterion variables Yq . 2. Obtain a matrix of correlation coefficients (Yqp ) linking the original criterion variables Yq to the orthogonolized criterion variables ZYp . 3. Calculate a matrix of regression coefficients (β jq ) by multiplying the inverse of the matrix obtained at step 2 and the matrix of validities between the original sets of predictors Xj and criteria Yq (rxy ). 4. Obtain a matrix of correlation coefficients (λjk ) linking the original predictors (Xj ) to the new uncorrelated predictors ZXk . This step is analogous to computing λjk in the univariate version of relative weight analysis described earlier in this chapter. 5. Calculate a matrix of regression coefficients (β kq ) by multiplying the inverse of the matrix obtained at step 4 and the matrix of regression coefficients (β jq ) obtained at step 3. 6. Calculate a multivariate relative weight (µj ) for each of the predictors by first multiplying squared λjk and squared β kq , and second averaging the obtained values across criterion variables. The path diagram depicting the sequence of steps involved in the multivariate relative weight analysis for a model with three correlated predictors is shown in Figure 4.2. This approach was used by Dalal, Baysinger, Brummel, et al. (manuscript in preparation) to examine relative importance of job attitudes (employee engagement, job satisfaction, organizational commitment, job involvement, perceived organizational support, and work centrality) and trait affect in explaining variance in three forms of overall employee’s contribution to the organization – employee’s task performance, organizational citizenship behavior (OCB), and counterproductive work behavior (CWB). More specifically, the authors used multivariate relative weight analysis to determine the extent to which: (i) the most recently proposed job attitude – employee engagement – is redundant with other attitudes in predicting overall job performance; and (ii) attitudes
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
135
VARIABLES IN MULTIPLE REGRESSION MODELS
λ11
X1
λ12 λ21 λ22 λ23
γ11
λ32
Y1
γ12
γ13 γ21
β22
Zx2
Zy2
β23
γ22 γ23
Zx3
Y2 γ31
β31
λ33
Figure 4.2 dictors.
Zy1
β21
λ31 X3
β12 β13
λ13
X2
β11
Zx1
β32 β33
Zy3
γ32
Y3
γ33
Calculation of multivariate relative weights for the regression model with three pre-
add value in predicting criteria above and beyond trait affect. Multivariate relative weights obtained in this study indicated that negative affect was the most important predictor of overall job performance (the composite of three performance facets), followed by job satisfaction and employee engagement. Thus, the results demonstrate that job satisfaction and employee engagement make greater contribution to predicting overall job performance than most of the other attitudes. The results also indicate that attitudes (job satisfaction and employee engagement) and negative affect are not redundant in predicting the criterion – both emerge as important predictors when included in the model simultaneously. In addition to providing raw relative weights, multivariate relative weight analysis allows computing rescaled multivariate relative weights that indicate the proportions of variance in the criterion space accounted for by each predictor and can be interpreted as effect sizes. Rescaled weights are computed as a ratio of the corresponding relative weight and the multivariate measure of linear association P 2 YX (µj /P 2 YX ; LeBreton & Tonidandel, 2008). Multivariate dominance weights and multivariate relative weights have a number of advantages. First, both measures have been shown to yield virtually identical estimates of multivariate relative importance (LeBreton & Tonidandel, 2008). Second, multivariate dominance weights and multivariate relative weights tend to furnish unbiased sample-based estimates of importance (i.e., estimates that converge with the corresponding population values). Finally, multivariate measures of relative importance are easily interpretable in contrast to the other available alternatives (e.g., canonical coefficients; Tabachnick & Fidell, 2007). Estimating Predictor Importance in Logistic Regression Organizational scholars have also become increasingly interested in criteria that may not meet the distributional assumptions of ordinary least squares
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
136 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY (OLS) regression. However, both dominance analysis and relative weight analysis were originally developed for use with OLS regression and thus are not suitable for use with alternative criteria such as dichotomous outcomes. To meet this need, Azen and Traxel (2009) and Tonidandel and LeBreton (2010) presented modifications of traditional dominance analysis and relative weight analysis to handle criterion variables that do not meet the assumptions of OLS regression but would be appropriate for a logistic regression model. Logistic dominance analysis Applying dominance analysis to logistic regression relies on the same principles as traditional dominance analysis and requires conducting an all subsets regression and examining the average change in R2 for each variable over all the subsets. The critical difference between the two is that when applying dominance analysis to logistic regression one must rely on a logistic regression analog of R2 . Azen and Traxel (2009) describe four suitable replacements for the coefficient of determination when conducting logistic regression (R2 M , R2 N , R2 E , and R2 (y,ˆy) ). Once a metric is chosen, one simply needs to conduct an all-subsets logistic regression analysis and examine the average change in the chosen logistic regression R2 analog that results from adding a predictor to all possible subsets of the remaining predictors. Questions of general dominance, conditional dominance, and complete dominance can then be addressed with the resulting weights that are produced by this approach. Logistic relative weight analysis Applying relative weight analysis to logistic regression also requires a slight modification of the traditional procedure. Recall that one of the steps for obtaining relative weights involves computing a set of standardized regression coefficients β k by regressing the criterion variable Y on the set of the newly created orthogonal predictors ZXk . In a logistic relative weight analysis, this step is no longer applicable as the coefficients that would be generated are not suitable for a categorical criterion. Instead, one must obtain a set of linking coefficients similar to β k that are appropriate for logistic regression. Tonidandel and LeBreton (2010) describe how these logistic regression analogs of the traditional standardized regression coefficients identified by Menard (2004) can be obtained and used in place of β k in a logistic relative weight analysis. These fully standardized logistic regression weights can be used to link the original variables to the criterion to produce relative weights that represent the contribution of each of the original predictors in explaining the categorical criterion. For a more technical discussion of the logistic relative importance analysis, the interested reader is directed to the original papers cited above.
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
137
CONCLUSIONS AND SUGGESTIONS FOR FUTURE RESEARCH Organizational researchers have frequently been interested in examining relative contribution of predictors in multiple regression models. However, traditionally used measures of relative importance (bivariate correlations, squared bivariate correlations, beta-weights, squared beta-weights, and product measures) have been shown to provide inaccurate estimates of importance in the presence of correlated predictors (LeBreton, Ployhart, & Ladd, 2004). Dominance weights (Budescu, 1993) and relative weights (Johnson, 2000) were designed to address the need of organizational scholars in statistically meaningful measures of relative importance. The purpose of this chapter is to review the various ways variable importance has been defined and measured. After reviewing the traditional methods that have been used to estimate relative importance, we have described two modern relative importance statistics and explained how they yield virtually identical results. These two statistical techniques, dominance analysis and relative weight analysis, have emerged as the preferred methods for estimating relative importance and are especially useful when researchers rely on regression analysis for explanatory and theory-testing purposes (LeBreton, Hargis, Griepentrog, et al., 2007). Dominance analysis and relative weight analysis have also been updated in recent years to permit inferences concerning statistical significance, and to accommodate multivariate regression models and logistic regression models. These are important advancements, but we believe there are at least three other avenues for the further development of the relative importance methodology: estimating importance of predictors in regression models with higher order terms, determining predictor importance in multilevel models, and using relative importance analysis in MANOVA designs.
Determining Relative Importance for Models with Interactions and Other Higher Order Effects Testing for significance of interaction effects and higher order, non-linear terms are two popular applications of regression. For example, it has been demonstrated that moderated multiple regression is frequently used in organizational research to test for presence of interactions among predictors (Cortina, 1993). Also, due to researchers’ increased attention to studying dynamic criteria (Hofmann, Jacobs, & Baratta, 1993), investigating temporal relationships among variables (Mitchell & James, 2001), and examining person’s affect, cognitions, and behaviors using experience sampling methods (Beal & Weiss, 2003), analytic techniques that allow detecting trends in longitudinal data (e.g., polynomial regression and latent growth curve modeling) have become widely used. However, up to now, both dominance analysis and relative weight
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
138 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY analysis have been predicated largely on the use of a first-order regression model. Future research should seek to understand how to decompose the R2 in regression models containing interactions and other higher order terms.
Determining Relative Importance for Multilevel Models With the increased interest in examining organizational phenomena at multiple levels of organizational hierarchy (Klein, Dansereau, & Hall, 1994), multilevel modeling has become a very popular analytic approach (Peugh & Enders, 2005). Within the past few decades, it has been freqeuntly used for analyzing hierarchical data either obtained from lower level organizational units nested within higher level organizational units or collected from subjects over multiple time points (Hofmann, 1997; Raudenbush & Bryk, 2002). In multilevel modeling, similar to OLS regression, effects of predictors on outcome variables are reflected in regression coefficients. However, due to the nested nature of data, multilevel regression coefficients reside at different levels of analysis. Therefore, both dominance analysis and relative weight analysis that are based on the traditional OLS regression model cannot be applied to examining predictor importance in a multilevel context. Thus, research is necessary to explore how to determine the relative importance of variables when they are measured at different levels of analysis.
Extending Relative Importance Analysis to MANOVA Designs Researchers are also frequently interested in examining the effects of multiple treatments (experimental conditions) on multidimensional criteria (e.g., facets of job satisfaction, dimensions of cognitive ability; LeBreton & Tonidandel, 2008) or sets of multiple intercorrelated criteria (Stone-Romero, Weaver, & Glenar, 1995). MANOVA is recommended for use with such research questions that involve examining differences across groups on a set of outcome variables (DeShon & Morris, 2002). Relative importance analysis appears to be a useful supplement to MANOVA when predictors are to be compared with each other on their contribution to explaining variance in a linear combination of outcome variables. Although univariate relative importance analysis has been recently extended for use with multivariate models (i.e., Azen & Budescu (2006) proposed multivariate dominance analysis, LeBreton & Tonidandel (2008) introduced multivariate relative weight analysis, and Huo & Budescu (2009) demonstrated how dominance analysis can be used with canonical correlation), the need in relative importance analysis for use with MANOVA remains unaddressed. We believe that the general principles of relative importance (i.e., decomposing predicted variance and identifying the most important predictors) could also be extended to MANOVA designs. Historically, interpreting the relative
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
139
importance of criterion variables in a MANOVA design has been accomplished using: 1. univariate ANOVAs, which are known to be problematic when used to make inferences regarding multivariate relationships; 2. stepdown analysis, which requires strong a priori theory concerning the hierarchical arrangement of the predictor variables; or 3. discriminant function analysis, which relies on the use of regression coefficients to evaluate variable importance and therefore shares the limitations of beta-weights that have been discussed above. Given that none of the traditionally used procedures appear to be adequate for estimating variable importance in a MANOVA context, the basic statistical models underlying relative importance in regression models need to be extended for use in MANOVA. Although the above three areas represent fruitful areas for future research, others will undoubtedly discover new directions and new problems to be overcome. We are encouraged by the interest organizational researchers have in relative importance analyses, and hope that this chapter will serve as a useful review for those seeking to learn more about these statistical procedures.
REFERENCES Austin, J.T., Scherbaum, C.A., & Mahlman, R.A. (2002). History of research methods in industrial and organizational psychology: Measurement, design, analysis. In S.G. Rogelberg (Ed.), Handbook of Research Methods in Industrial and Organizational Psychology (pp. 3–33). Malden, MA: Blackwell Publishers. Azen, R., & Budescu, D.V. (2003). The dominance analysis approach for comparing predictors in multiple regression. Psychological Methods, 8, 129–48. Azen, R., & Budescu, D.V. (2006). Comparing predictors in multivariate regression models: An extension of dominance analysis. Journal of Educational and Behavioral Statistics, 31, 157–80. Azen, R., & Traxel, N. (2009). Using dominance analysis to determine predictor importance in logistic regression. Journal of Educational and Behavioral Statistics, 34, 319–47. Beal, D.J., & Weiss, H.M. (2003). Methods of ecological momentary assessment in organizational research. Organizational Research Methods, 6, 440–64. Bring, J. (1996). A geometric approach to compare variables in a regression model. American Statistician, 50, 57–62. Budescu, D.V. (1993). Dominance analysis: A new approach to the problem of relative importance of predictors in multiple regression. Psychological Bulletin, 114, 542–51. Cohen, J., & Cohen, P. (1983). Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences (2nd edn). Hillsdale, NJ: Lawrence Erlbaum. Cortina, J.M. (1993). Interaction, nonlinearity, and multicollinearity: Implications for multiple regression. Journal of Management, 19, 915–22.
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
140 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Courville, T., & Thompson, B. (2001). Use of structure coefficients in published multiple regression articles: β is not enough. Educational and Psychological Measurement, 61, 229–48. Cronbach, L.J., & Gleser, G.C. (1957). Psychological Tests and Personnel Decisions. Urbana, IL: University of Illinois Press. Darlington, R.B. (1968). Multiple regression in psychological research and practice. Psychological Bulletin, 69, 161–82. Darlington, R.B. (1990). Regression and Linear Models. New York: McGraw Hill. DeShon, R.P., & Morris, S.B. (2002). Modeling complex data structures: The general linear model and beyond. In S.G. Rogelberg (Ed.), Handbook of Research Methods in Industrial and Organizational Psychology (pp. 390–411). Malden, MA: Blackwell Publishers. Eby, L.T., Durley, J.R., Evans, S.C., & Ragins, B.R. (2006). The relationship between short-term mentoring benefits and long-term mentor outcomes. Journal of Vocational Behavior, 69, 424–44. Efron, B., & Tibshirani, R.J. (1993). An Introduction to the Bootstrap. New York: Chapman & Hall. Engelhart, M.D. (1936). The technique of path coefficients. Psychometrika, 1, 287–93. Gibson, W.A. (1962). Orthogonal predictors: A possible resolution of the Hoffman– Ward controversy. Psychological Reports, 11, 32–4. Green, P.E., Carroll, J.D., & DeSarbo, W.S. (1978). A new measure of predictor variable importance in multiple regression. Journal of Marketing Research, 15, 356–60. Hoffman, P.J. (1960). The paramorphic representation of clinical judgment. Psychological Bulletin, 57, 116–131. Hofmann, D.A. (1997). An overview of the logic and rationale of hierarchical linear models. Journal of Management, 23, 723–44. Hofmann, D.A., Jacobs, R., & Baratta, J.E. (1993). Dynamic criteria and the measurement of change. Journal of Applied Psychology, 78, 194–204. Huo, Y., & Budescu, D.V. (2009). An extension of dominance analysis to canonical correlation analysis. Multivariate Behavioral Research, 44, 688–709. Johnson, J.W. (2000). A heuristic method for estimating the relative weight of predictor variables in multiple regression. Multivariate Behavioral Research, 35, 1–19. Johnson, J.W. (2004). Factors affecting relative weights: The influence of sampling and measurement error. Organizational Research Methods, 7, 283–99. Johnson, J.W., & LeBreton, J.M. (2004). History and use of relative importance indices in organizational research. Organizational Research Methods, 7, 238–57. Johnson, R.M. (1966). The minimal transformation to orthonormality. Psychometrika, 31, 61–6. Kerlinger, F.N., & Pedhazur, E.J. (1973). Multiple Regression in Behavioral Research. New York: Holt, Rinehart & Winston. Klein, K.J., Dansereau, F., & Hall, R.J. (1994). Levels issues in theory development, data collection, and analysis. Academy of Management Review, 19, 195–229. Kutner, M.H., Nachtsheim, C.J., Neter, J., & Li, W. (2005). Applied Linear Statistical Models (5th edn). New York: McGraw Hill. LeBreton, J.M., Hargis, M.B., Griepentrog, B., Oswald, F.L., & Ployhart, R.E. (2007). A multidimensional approach for evaluating variables in organizational research and practice. Personnel Psychology, 60, 475–98. LeBreton, J.M., & Tonidandel, S. (2008). Multivariate relative importance: Extending relative weight analysis to multivariate criterion spaces. Journal of Applied Psychology, 93, 329–45. LeBreton, J.M., Ployhart, R.E., & Ladd, R.T. (2004). A Monte Carlo comparison of relative importance methodologies. Organizational Research Methods, 7, 258–82.
P1: OTA/XYZ P2: ABC c04 JWST046-Hodgkinson
January 21, 2011
12:58
Printer Name: Yet to Come
VARIABLES IN MULTIPLE REGRESSION MODELS
141
Luo, Y. (2005). How important are shared perceptions of procedural justice in cooperative alliances? Academy of Management Journal, 48, 695–709. McAllister, D.J., Kamdar, D., Morrison, E.W., & Turban, D.B. (2007). Disentangling role perceptions: How perceived role breadth, discretion, instrumentality, and efficacy relate to helping and taking charge. Journal of Applied Psychology, 92, 1200–11. Menard, S. (2004). Six approaches to calculating standardized logistic regression coefficients. American Statistician, 58, 218–23. Mitchell, T.R., & James, L.R. (2001). Building better theory: Time and the specification of when things happen. Academy of Management Review, 26, 530–47. Pedhazur, E.J. (1997). Multiple Regression in Behavioral Research: Explanation and Prediction (3rd edn). New York: Holt, Rinehart & Winston. Peugh, J.L., & Enders, C.K. (2005). Using the SPSS Mixed procedure to fit crosssectional and longitudinal multilevel models. Educational and Psychological Measurement, 65, 717–41. Raudenbush, S.W., & Bryk, A.S. (2002). Hierarchical Linear Models: Applications and Data Analysis Methods (2nd edn). Thousand Oaks, CA: Sage. Schleicher, D.J., Venkataramani, V., Morgeson, F.P., & Campion, M.A. (2006). So you didn’t get the job . . . now what do you think? Examining opportunity-to-perform fairness perceptions. Personnel Psychology, 59, 559–90. Sechrest, L. (1963). Incremental validity: A recommendation. Educational and Psychological Measurement, 23, 153–8. Society for Industrial and Organizational Psychology (SIOP) (2003). Principles for the Validation and Use of Personnel Selection Procedures (4th edn). Bowling Green, OH. Stone-Romero, E.F., Weaver, A.E., & Glenar, J.L. (1995). Trends in research design and data analytic strategies in organizational research. Journal of Management, 21, 141–57. Tabachnick, B.G., & Fidell, L.S. (2007). Using Multivariate Statistics (5th edn). Boston: Pearson. Thomas, D.R., Hughes, E., & Zambo, B.D. (1998). On variable importance in linear regression. Social Indicators Research, 45, 253–75. Tonidandel, S., & LeBreton, J.M. (2010). Determining the relative importance of predictors in logistic regression: An extension of relative weights analysis. Organizational Research Methods, 13, 767–81. Tonidandel, S., LeBreton, J.M., & Johnson, J.W. (2009). Determining the statistical significance of relative weights. Psychological Methods, 14, 387–99.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
Chapter 5 EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS Rosalind Searle The Open University, Walton Hall, Milton Keynes, MK7 6AA, UK
Antoinette Weibel University of Konstanz, 78457 Konstanz, Germany
Deanne N. Den Hartog Amsterdam Business School, University of Amsterdam, 1018 TV, Amsterdam, The Netherlands
As business leaders, trust in us is at an historic low. We should be alarmed. We should use this as a wake-up call for reform. And we must embark upon an urgent journey to restore this diminished trust in business. (Edelman, 2009: 1) We have to recognize that we face more than a deficit of dollars right now. We face a deficit of trust – deep and corrosive doubts about how Washington works that have been growing for years. To close that credibility gap we have to take action on both ends of Pennsylvania Avenue – to end the outsized influence of lobbyists, to do our work openly, to give our people the government they deserve. (President Obama in his inaugural address, January 2010)
INTRODUCTION Trust in organizations, as the two quotes above highlight, is currently a fragile resource. Yet trust has been identified as critical to the success of organizations, with higher trust leading to improved cooperation and coordination in the workplace, lower conflict, and enhanced performance (e.g., Dirks & Ferrin, 2002; Zaheer, McEvily, & Perrone, 1998). Some authors argue that trust is becoming the key organizing principle for business (McEvily, Perrone, & Zaheer, 2003) as new forms of organizations demand novel means of International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
144 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY coordinating tasks and promoting cooperation. Such new forms of organizing increase inter-dependency through flatter hierarchies, more self-organization, team-based structures, learning-centered community organizations, and the emergence of relational contracting (for a recent overview see Snow, Fjeldstad, Lettl, et al., 2010). In order to be successful, all of these new forms of organizing require trust. Given the above observations, it is not surprising that over the past 20 years interest in researching the antecedents and consequences of trust has flourished in different disciplines within the social sciences. The purpose of this chapter is to review the empirical and conceptual literature on employee trust within the context of work organizations. We focus on what employee trust means, what drives its development, how trust processes unfold over time, and, finally, the impact employee trust has on organizations. Although trust is also important in the relationships individuals or organizations have with other stakeholders (e.g., customers, suppliers), our focus here is on employees’ trust within firms. We include both individual employees’ trust of specific other organizational members (e.g., trust in a specific colleague or trust in the leader) as well as individual employees’ trust in generalized organizational entities (e.g., trust in management or in the organization as a whole). We begin by reviewing different views on the nature of trust. Next, we focus on three core elements in the literature: antecedents of trust, trust processes, and the most salient direct and indirect consequences of trust. In terms of antecedents, we consider both individual differences (in particular, the propensity to trust) and contextual factors and processes (e.g., leadership, human resource (HR) practices, justice, and organizational control) that can affect employee trust. Next, we consider the dynamic nature of trust and how to repair it when broken. We then show how trust affects important behaviors (e.g., knowledge sharing and citizenship) and organizational outcomes (e.g., innovation and performance). In conclusion, we outline an agenda for future research.
THE NATURE OF TRUST Trust “entails a state of perceived vulnerability or risk that is derived from individuals’ uncertainty regarding the motives, intentions, and prospective actions of others on whom they depend” (Kramer, 1999: 571). Thus, trust only becomes an issue when an employee is dependent on, and vulnerable to, the actions of another party – be it the supervisor, the work group, or the employer. In trust research, the trusting party is often called “trustor,” while the object of trust is termed “trustee.” To trust means to accept risks: the trusting employee has little to gain but much to lose if trust is misplaced (e.g., Deutsch, 1960). For example, the decision to share a piece of information with a supervisor makes the employee dependent. A trusting decision may be based on misplaced trust and can lead to negative consequences. In such situations of dependence, employees are more likely to accept vulnerability if
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
145
they have positive expectations of the intentions or behavior of the other party involved (Rousseau, Sitkin, Burt, et al., 1998). This combination of positive expectations and vulnerability within a dependent relationship is pivotal for trust (Luhmann, 1988). As the prominence of risk and vulnerability varies between contexts, trust is bound not to be a salient consideration in each and every situation. For example, when employees have control over how another party (e.g., a fellow team member) behaves, they are less vulnerable. Hence, trust is less of an issue. Similarly, if there is little to lose when the trusted party does not behave as desired, the risk may be so small that trust is simply not relevant. Trust Beliefs In line with the aforementioned definition which highlights how trust revolves around perceived vulnerability and risk in a dependent relationship (e.g., Kramer, 1999; Lewicki, Tomlinson, & Gillespie, 2006), much of the literature focuses on employees’ trust as a confident set of beliefs an employee holds regarding a trustee (such as their leader) and the relationship with this person. These perceptions lead the employee to have positive expectations about the outcome of dealings with this trustee. Such a confident set of beliefs often arises from an assessment of the trustworthiness of the trustee (Mayer, Davis, & Schoorman, 1995). Although previous work specified fewer (e.g., Deutsch, 1960) or more dimensions (e.g., Butler, 1991), the most influential model of trust beliefs was developed by Mayer, Davis, and Schoorman (1995). It focuses on ability, benevolence, and integrity of the trustee as main components of trust beliefs. Ability beliefs concern the perceived competence of the trustee. Benevolence beliefs reflect the care and concern of the trustee for the well-being of the trusting employee. Integrity beliefs focus on whether the trustee is likely to adhere to moral principles and codes of behavior. Although most authors follow this ability, benevolence, and integrity model, this convergence does not reflect a true consensus as researchers often include very different facets under the generic labels of “ability,” “benevolence,” and “integrity” (McEvily & Tortoriello, 2005). For example, some conceptualize integrity as congruence between words and deeds (Simons, 2002), while others regard it as having a set of shared core values (Long & Sitkin, 2006). Although initially trust research has been able to move ahead through adopting a pragmatic approach, which has glossed over differences in the details behind these labels, as this field matures it may now be important to go back in order to tease out these differences. It is very likely that trust beliefs vary with the context of trust relationships. For example, reliability, so far treated as a subcomponent of integrity, might be the essential trust driver in more calculative relationships and also in high-risk contexts. Most work on trust beliefs focuses on the interpersonal level, that is, on individuals trusting specific other people (such as their leader, a co-worker, and so on) and the role that ability, benevolence, and integrity have there.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
146 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Yet, in the face of the current global economic crisis, trust in organizations has become an important topic. In response, several authors have started to use the ability, benevolence, and integrity model to capture beliefs around the trustworthiness of organizations (e.g., Gillespie & Dietz, 2009; Schoorman, Mayer, & Davis, 2007). Trust in the organization, of course, differs from trust in a specific person by its referent; it is not so clear what individuals are referring to when they say they trust their employer or the organization. Giddens (1990: 34), for example, related trust in organizations to “reliability and faith in the correctness of abstract principles” and similarly (Carnevale, 1995: xi) defines it as the “faith that an institution will be fair, reliable, competent, and non-threatening.” As such, perceptions of organizational trustworthiness hinge on the “collective characteristics of an administrative organization and top management group which are not reducible to features of individual actors and which ensure some continuity of activities and direction when those actors change” (Whitley, 1987: 133). Therefore, trust beliefs referring to organizations may be tied to a more lasting impression than interpersonal trust beliefs, emphasizing the role of cues signaling reliability and continuity. Attributes of trust, such as predictability, may thus be of greater significance for organizational than interpersonal level trust beliefs. However, this has not yet been tested. In addition, ability, benevolence, and integrity beliefs are also likely to be important because in order to accept vulnerability, cues that signal an organization’s good intentions and competence also matter (Choudhury, 2008). In line with this, Searle, Den Hartog, Weibel, et al. (in press) found that employees’ trust in their employing organization related both to the perceived ability of the organization and the perceived benevolent and high-integrity intentions of the organization. Other Views of Trust The dominant view of trust as a psychological state associated with a confident set of beliefs; however, Dietz and Den Hartog (2006) note that some attention has also focused on individuals’ decision to trust, using a rational choice perspective or behavioral decision making view (e.g., Cook, Yamagishi, Cheshire, et al., 2005; Hardin, 1993; Kramer, 2006). In addition, some authors have examined how trust might be expressed as an act (e.g., Gillespie, 2003). These three positions (beliefs, decisions, action) can be considered as a progression from viewing trust only as an internal private perception, through to trust as expressed in observable behavior. The latter behavioral perspective is controversial, however, as it blurs boundaries between trust and its outcomes. Also, the connection between the visible actions and trust is not always clear. For example, in behavioral economics cooperation is frequently used as a proxy for trust (e.g., Bohnet & Zeckhauser, 2004; Kosfeld, Heinrichs, Zak, et al., 2005); however, while trust and the confident belief that the other party will reciprocate may be one reason to cooperate, there are also other reasons why
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
147
cooperation could occur, such as the moral position that one ought to cooperate. Discerning which actions are indeed trust-based and which occur for other reasons still requires a measurement of trust beyond the action itself. Types of Trust Several authors suggest that different types of trust exist, reflecting the type of the relationship between the trusting employee and the other party. For example, researchers distinguish between thin and thick trust (e.g., Nooteboom & Six, 2003), competence- and goodwill-trust (e.g., Sako, 1992), or calculative-based and relationally-oriented trust (Kramer, 1999; Lewicki & Bunker, 1996; Rousseau, Sitkin, Burt, et al., 1998). Central to most such models is the distinction between economic exchange-based rational trust and community exchange-based social/emotional trust (Lewicki, Tomlinson, & Gillespie, 2006; McAllister, 1995; Rousseau, Sitkin, Burt, et al., 1998). These models echo developments in other areas of organizational psychology, such as in social exchange, justice, psychological contract, and organizational citizenship behavior (OCB) literature, where the type of relationship has been identified to be an important intervening or shaping variable (Shore, Bommer, Rao, et al., 2009; Song, Tsui & Law, 2009). Relationship-sensitive trust models assume that trust co-varies with the underlying relationship and that the nature of the relationship can be described along a continuum ranging from shallow to deep relationships. A shallow relationship rests on economic-exchange norms where favors are merely reciprocated. The relationship is constrained in the range and frequency of contact between the parties (e.g., Clark & Mills, 1993; Fiske, 1992). In contrast, a deep relationship rests on a communal exchange norm where help is offered based on a concern for the welfare of the other, and the relationship is broad in range and number of contacts between the parties (Sheppard & Sherman, 1998). Different trust types are linked to different depths of relationships. For example, shallow relationships are linked to calculative trust, which is based on a pure cost–benefit calculation. The employee’s decision to trust the other party then hinges on the idea that, in this transaction, it is in the other party’s interest not to betray them (Hardin, 1993). Incentives for the trustee not to betray trust may be due to such negative actions being visible, and therefore resulting in sanctions or forestalling future opportunities for profit generation, because betrayal would lead ultimately to a cessation of cooperation with the trusting employee. Employees may thus trust each other purely based on a calculation of the low possibilities of betrayal. However, there is some dispute as to whether calculative trust should be seen as trust in another party at all. Many researchers contend that when such trust is directed towards the sanctioning system or towards incentives of the situation, trust beliefs concerning the trustee are not required and thus there is no need for interpersonal trust in this situation (James, 2002; Kramer, 1999).
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
148 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Linked to deep relationships, and thus at the other end of the relationship continuum, lies identification-based trust. This form of trust exists because the parties fully understand each others’ needs and share central values. Identification-based trust develops into a strong mutual understanding that operates in a broader variety of contexts and situations. However, some researchers argue that identification-based trust is close to blind or unquestioning trust (Dukerich, Kramer, & McLean Parks, 1998). Questions again arise as to whether this should be regarded as trust as defined above because, in the context of deep relationships, there is often a lack of felt vulnerability or risk. If the trusting employee has complete confidence (rightfully or not) in the trustee, then a defining criterion of trust is missing. Between the extremes of calculative and identification-based trust lie two other forms of trust: knowledge-based and relational-based trust (Rousseau, Sitkin, Burt, et al., 1998). Similar to calculative trust, knowledge-based trust involves some calculation or reasoned prediction. Knowledge-based trust is deduced from a common interaction history and derived from past behavior that enables the trusting employee to predict trustee behavior (Lewicki & Bunker, 1996). Knowledge-based trust focuses on the predictability of trustees, through exposure to their competencies and character. For example, employees may feel they can trust a co-worker to perform a task well, because they have seen this co-worker act competently in similar situations. Knowledge-based trust is about knowing the interaction partner well enough so their behavior becomes predictable. In contrast to calculative trust, which focuses on whether the trusting employee is likely to be betrayed given the incentives for cheating in a specific situation, knowledge-based trust is firmly built on perceptions of the other party’s trustworthiness. Thus, trust beliefs become the focal point. Although relational-based trust also emerges over time via direct and indirect contact with the other party, a different source of the perceived intentions not to betray them is central (Deutsch, 1960). Relational-based trust can only evolve if the trustee is perceived to have an intrinsic interest in the relationship with the trusting employee. The focus thus lies on the social orientation toward other people (Kramer, 1999). As both parties interact, positive expectations of each other’s trustworthiness expand and a shared concern about each other’s welfare emerges. Over time, emotions enter the relationship due to the reciprocated care and concern between the parties (McAllister, 1995). In other words, the bases of relational-based trust are more affective and social than rationally calculated and as such more focused on community concerns than on economic exchange. These different relationship-sensitive trust types hold at the inter-personal level; however, less is known about different levels of trust in collectives such as work organizations. If employees can be said to have a relationship with their organization, similar issues will play a role for trust in organizations. That is, previous personal experience with the firm and observing the treatment of colleagues (cf. Bandura, 1986) help build trust levels that go beyond
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
149
calculative trust forms. Knowledge-based trust in the organization would then focus more on economic exchange, and the competence and predictability of the organization to deliver on their promises (“Do I get what I deserve?”), whereas relational-based trust in the organization highlights shared values and identification with the organization, felt care, and a sense of partnership. This hypothesis has yet to be tested, but is in line with the distinction between transactional versus relational psychological contracts (e.g., Conway & Briner, 2009; Rousseau, 1995).
ANTECEDENTS OF TRUST After considering the nature of employee trust, an important question is what leads employees to trust their leaders and their organization. In this section, we discuss the main drivers of trust beliefs, decisions, and actions that have been outlined in the literature to date. We first describe the role of individual differences and most notably the propensity to trust, before turning our attention towards contextual factors affecting employee trust, including: leadership, HR practices, justice, and control. Individual Differences Research on the individual difference antecedents of trust has mostly focused on personality and the predisposition to trust, but recent research suggests other individual differences, such as status, may also play a part. Disposition to trust Some individuals trust more readily than others. Rotter (1967) was the first to identify individuals’ readiness or propensity to trust as a personality attribute. He defined trust propensity as a relatively stable generalized expectancy about the trustworthiness of others, leading to the formation of trusting relationships with others and an acceptance of vulnerability. This personality factor is referred to variously as generalized trust (Stack, 1978), disposition to trust (Kramer, 1999), or trust propensity (Mayer, Davis, & Schoorman, 1995) and forms as a sub-factor of the agreeableness dimension of the five factor model of personality (Costa & McCrae, 1992). The existence of a trusting personality dimension is important, as this implies personal experience at work is not the only shaper of employees’ trust levels. Both disposition and experience drive trust. Here, we examine the empirical evidence on three distinct effects of disposition to trust. First, we address the impact of this disposition on trust even when trustworthiness information about the other party is available. Second, research suggests it affects perceptions of others in general as those with a higher disposition to trust have a more positive and less suspicious view of
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
150 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY others. Finally, it influences individuals’ trusting and cooperative behaviors, with high trust propensity individuals displaying such behaviors more frequently than low trust propensity individuals. One of the most important influences of propensity to trust is on the development and maintenance of trust. Kee and Knox (1970) argue that trust beliefs of the trusting employee are shaped by dispositional trust, even where previous experience with a transaction partner is available. Lewis and Weigert (1985) suggest that the propensity to trust may drive and shape the “cognitive leap” of trust, beyond the level that previous experience alone would warrant, and argue that trust predisposition has a significant and independent impact on trust even in the presence of trustworthiness information. These theoretical propositions have been examined empirically and a recent meta-analysis has shown that trust propensity is indeed significantly related to interpersonal trust once dimensions of trustworthiness are controlled for (Colquitt, Scott, & LePine, 2007). A similar effect of propensity to trust has also been found in the case of employees’ trust in their employer: employees’ propensity to trust had a significant effect on trust in the organization beyond the perceptions of the ability and benevolent intentions of the firm (Searle, Den Hartog, Weibel, et al., in press). The influence of dispositional trust beyond first-hand trusting experience seems to be particularly relevant in ambiguous situations (Gill, Boies, Finegan, et al., 2005) or where an individual is dealing with unfamiliar actors (Bigley & Pearce, 1998). In ambiguous and novel contexts, situational cues will tend to be weaker. This leaves more individual discretion and therefore individual differences factors, such as personality, have a greater impact in determining behavior. Thus, contexts in which propensity to trust was found to be relevant to trust behaviors include reorganizations, cross-functional team working, and joint ventures (McKnight, Cummings, & Chervany, 1998). Few studies, however, have included trust propensity as an intervening variable in examining ambiguous context, an exception is Yakovleva, Reilly, and Werko (2010) in their examination of trust in virtual contexts. Whereas in conventional work environments trust between co-workers is important, they found in virtual and geographically dispersed contexts its significance is magnified. Their field study looked at dyads in a new product development team and found trust propensity had a greater influence on trust in virtual dyads than for those operating face-to-face. Research also suggests that individuals with a high propensity to trust have a more positive and less suspicious view of others (Rotter 1967, 1980). One suggested mechanism for this effect is a filter altering the interpretations of others’ actions (Govier, 1994). Parks, Henager and Scamahorn (1996) in their dilemma game study showed that those with high propensity to trust are more sensitive to signs of trustworthiness, while conversely those with a low trust propensity are more sensitive to signs of betrayal. They suggest that these two groups attend to information in different ways, with those high on
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
151
trust propensity requiring repeated evidence of untrustworthiness in order to become untrusting, while those with low propensity needed further proof of trustworthiness before they could trust. Yakelova, Reilly, and Werko (2010) in their field study confirmed that those high on trust propensity have more favorable perceptions of others’ trustworthiness, particularly of others’ benevolence and integrity. In addition, they revealed that low trusting individuals often became exploitative if they experienced their opponents behaving cooperatively over a long period. These studies suggest that differences in trust propensity may significantly alter the types of cues to which employees attend. Importantly, high trust propensity does not appear to be synonymous with gullibility (Rotter, 1980), but reflects the tendency of the individual to hold a more positive perspective of others and their intentions. Hence, high trust propensity individuals are not naive. Indeed, Colquitt, Scott, and Lepine (2007) suggest that the construct of trust propensity could be regarded as reflecting differences in sensitivity towards events, particularly where issues of justice arise. Filtering and attending to different types of information is argued to predispose those high on trust propensity to be less sensitive to information about adverse situations (Govier, 1994). This suggests propensity to trust may act as a buffer in contexts of organizational trust breakdown. The final impact of trust propensity concerns individuals’ cooperation, with those high on propensity to trust demonstrating more frequent cooperative behaviors. Typically, high trust individuals are found to display more honest and compliant behavior, which includes less cheating (Rotter, 1967, 1980; Stack, 1978) and are more prone to norms of reciprocity (Berneth & Walker, 2009). For example, Parks, Henager, and Scamahorn (1996) found high trust propensity individuals respond to calls for cooperation, but are not easily cued to react competitively. In contrast, those low on trust propensity do not respond when asked to become more cooperative, but react when asked to become completive. Another study suggests that high trust propensity individuals are more sensitive to their surroundings by moderating their consumption during times of scarce resources, while the consumption patterns of low trust propensity individuals remain unchanged (Brann & Foddy, 1987). As a result of this higher level of cooperative behavior, these trusting individuals have more resources and information available to them, which assists in their decision making and performance (Burt, 1992; Wayne, Shore, & Liden, 1997). Further individual difference factors Although researchers have identified other individual differences that may be important antecedents in studying trust in organizational contexts, as yet they have received relatively little attention. For example, recent questionnairebased research across a range of European firms, with employees at different organizational levels, indicates that those higher in their organization showed more trust in their organization and perceived it as more trustworthy (Searle,
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
152 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Den Hartog, Weibel, et al., in press). One reason for this higher perception of trust may be that those in senior management levels have more influence over decision making within an organization. This increased decision latitude might also enhance their perceptions of personal control and reduce their sense of vulnerability. In addition, senior management is often privy to greater volumes and depth of information than those further down the hierarchy. Such information may provide additional insights, helping individuals to contextualize and more fully understand the organization. Other individual level factors that have emerged as drivers of trusting beliefs include education, experience, and disciplinary background (e.g., Yakovleva, Reilly, & Werko, 2010). Yet to date the influence of these individual factors have neither been analyzed theoretically in a systematic fashion nor studied empirically. However, it is likely that factors that are shown to impact risk perceptions – such as gender, age, and cultural background (Cohn, Macfarlane, Yanez, et al., 1995; Gustafson 1998) – will also influence employees’ initial trust beliefs, and their subsequent attention to ongoing cues, signals, and information from within the organization. Organizational Antecedents of Trust Organizational factors that have been shown to be positively related to employee trust include leadership, human resource management (HRM) practices, organizational justice, and control mechanisms. Leadership Leadership has been defined as the ability of an individual to influence, motivate, and enable others to contribute toward the effectiveness and success of the organizations of which they are members (House, Hangers, Hanges, et al., 2004) or related as “the process of influencing the activities of an organized group in its efforts toward goal setting and goal achievement” (Stogdill, 1950: 3). To be able to exert such influence effectively and without the need for constant monitoring or for forceful coercion, trust between leaders and followers seems crucial. The GLOBE study, which focused on cultural differences in implicit leadership theories, suggests several trust-related qualities universally endorsed as important for leaders; for example, in all 60 countries involved in this study, an outstanding leader was expected to be a confidence builder, good at team building and communicating, who is decisive and intelligent as well as trustworthy, just, and honest (Den Hartog, House, Hanges, et al., 1999; House, Hanges, Hanges, et al., 2004). Trust in leaders has positive effects within organizations. For example, in a study of restaurants, Davis, Schoorman, Mayer, et al. (2000) found that employees’ trust in the general manager of their restaurant was related to improved financial performance and reduced employee turnover. Similarly, in
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
153
their study in a service context, Grant and Sumanth (2009) showed that trust in leaders was positively related to employees’ prosocial motivation and behavior. In their meta-analytical review, Dirks and Ferrin (2002) found that trust in leaders was positively related to enhanced job performance, organizational citizenship behavior, commitment, satisfaction, and reduced turnover intentions. Although to some extent trust in leaders may be role-based i.e., trust in the leader based not on person-specific knowledge about their capabilities or intentions, but on the role or office they hold and their training or background in that role (e.g., Kramer, 1999), the leader’s personal characteristics and actions are also likely to affect trust. Leader’s trustworthiness in the eyes of followers relies on the intentions that followers attribute to the leader (e.g., benevolence and integrity versus malevolence and dishonesty) and the leader’s perceived competence and ability (i.e., the extent to which the leader is seen as capable). So, trust in the leader grows if leader behavior signals these trustworthiness cues (e.g., Sweeney, 2010). Related research demonstrates that leader behaviors, such as transformational leadership and leader–member exchange, show strong relationships with trust in the leader (e.g., Cunningham & MacGregor, 2000; Den Hartog, 2003; Podsakoff, MacKenzie, Moorman, et al., 1990). Different leader behaviors relate to trust. Arguably the most classic distinction in leader behavior focuses on task versus relationship oriented leader behavior, also labeled initiating structure versus consideration (Bass, 1990; Fleishman & Harris, 1962; for a recent meta-analysis of studies of the effectiveness of these dimensions see Judge, Piccolo, & Ilies, 2004). Consideration includes leader behavior indicating mutual trust, respect, and a certain warmth and rapport between leader and subordinate. This form of leader behavior is closely related to relationship-based or affective forms of trust (McAllister, 1995). A leader’s initiating structure, in contrast, describes task-related behavior in which the leader organizes, directs, and defines group activities and seems to relate more to knowledge-based trust. In the last few decades, transformational, visionary, and charismatic models have dominated research on leadership. Essentially, these models contend that effective leaders articulate an attractive vision and behave in ways that reinforce the values inherent in that vision, which in turn inspires followers to transcend their own self-interests for the sake of the collective community. Followers become highly committed to the goal of the collective and perform beyond expectations (Bass, 1985, 1997). As noted above, such leadership is based on a strong bond of relational (or at times even identification-based) trust. Studies show that transformational leadership relates positively to trust in the leader (e.g., Podsakoff, Mackenzie, Moorman, et al., 1990) and spills over in more generalized trust in management and co-workers (e.g., Den Hartog, 2003). Research also found that trust mediates the relationship between transformational leadership and OCB (e.g., Podsakoff, Mackenzie, Moorman, et al., 1990). Transformational leadership is usually contrasted with transactional leadership, with
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
154 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY the former characterizing leader–follower relations as a series of (implicitly economic) exchanges between leaders and followers where followers perform as expected in exchange for promised rewards, in which more knowledge-based trust occurs as followers need to trust their leader to keep their word, treat them fairly, and provide them with the promised rewards. The meta-analysis by Dirks and Ferrin (2002) shows that indeed transactional leadership also relates to trust in the leader (albeit somewhat less strongly than transformational leadership). However, they could not distinguish between different forms of trust such as relational versus knowledge-based in relation to these styles. More recently, ethical and authentic styles of leadership have received attention. Brown, Trevino, and Harrison (2005: 120) define ethical leadership as “the demonstration of normatively appropriate conduct through personal actions and interpersonal relationships and the promotion of such conduct to followers through two-way communication, reinforcement and decision-making.” Authentic leadership (e.g., Avolio and Chan, 2008; Avolio, Gardner, Walumbwa, et al., 2004; Gardner, Avolio, Luthans, et al., 2005) is broader and focuses on how leader behaviors can draw upon and promote both positive psychological capacities and a positive ethical climate, as well as foster greater self-awareness, an internalized moral perspective, balanced processing of information, and relational transparency (e.g., Walumbwa, Avolio, Gardner, et al., 2008). Both authentic and ethical leadership are expected to enhance trust and initial research supports this proposition. For example, positive relationships were found of ethical leadership with trust in the leader (e.g., Brown, Trevino, & Harrison, 2005) and trust in management (Den Hartog & De Hoogh, 2009). Kalshoven, De Hoogh, and Den Hartog (in press) developed a multidimensional measure differentiating between seven different ethical leader behaviors (fairness, power sharing, people orientation, integrity, role clarification, ethical guidance, and concern for sustainability) and found that all seven dimensions related positively and significantly to trust in the leader. Kalshoven and Den Hartog (2009) proposed that group prototypicality and trust sequentially mediate the relationship between ethical leadership and leader effectiveness. The group prototype forms an ideal representation of the group’s identity prescribing appropriate attitudes and behaviors. Ethical leaders are role models and thus likely to be seen as prototypical. In turn, prototypes are more trusted and effective. A field study reported by Kalshoven and Den Hartog (2009) supported this mediational model. Trust is not always a salient concern in the leader – follower relationship For example, Brockner, Siegel, Daly, et al. (1997) studied conditions under which the relationship between employees’ trust in and support for organizational authorities is more or less pronounced. They argue that when outcomes are favorable for an employee, issues of trust are not salient or critical in determining support for leaders. Receipt of favorable outcomes does not raise issues of leader trustworthiness, because the outcomes themselves demonstrate that the leader is reliable and can be counted on to act in ways desired by the
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
155
employee. In contrast, when outcomes are unfavorable, trust becomes more salient and more critical for such support to occur. The results of their field studies supported their expectation that trust was more strongly related to support for an authority when outcomes were unfavorable. Brockner, Siegel, Daly, et al. (1997) conclude that trust seems to help overcome potential adverse reactions of employees to decisions with unfavorable outcomes and their study shows trust in an authority such as a leader can be important, but is not always salient. Human resource managment policies and practices HRM policies and practices are concerned with the management of employment relationships in the organization and argued to be relevant for the development of employee trust (Robinson & Rousseau, 1994; Searle & Skinner, In press; Whitener, 1997). Such policies can be seen as statements of intent, and to employees the nature of their implementation or delivery forms a measure of the extent to which management’s intentions are genuine and can be trusted (e.g., Skinner, Saunders, & Duckett, 2004). While the content of HR policies is typically the domain of HR professionals, their delivery is often the responsibility of line managers. For example, Wright and Nishii (in press) distinguish HR practices as intended from those that line managers implement, and in turn they distinguish implemented practices from the ones individual employees perceive. At each stage distortion is possible, thus in the end there may be a disjunction between the intended policies as conceived by the HR function and employees’ post-delivery perceptions. Hence, the content and manner of implementation of HRM practices can both be important for trust (e.g., Perrone, Zaheer, & McEvily, 2003; Shapiro, 1987). In this section we first briefly review how singular HRM practices affect trust and then explore in more detail how bundles of practices relate to employee trust in representatives of the organization, such as managers, or in the organization as a whole. Next, we look at the employment relationship of employees at the individual level by focusing on the psychological contract in relation to trust. Finally, we highlight the role of the HR function in enhancing or reducing employees’ trust in the organization. Early research on trust and HR often looked at trust in line managers and concentrated on distinct areas of HR policy, including performance appraisal (e.g., Dobbins, Platz, & Houston, 1993; Earley, 1986; Folger & Konovsky, 1989; Mayer & Davis, 1999), and reward and compensation (Pearce, Branyiczki, & Bigley, 2000). For example, a quasi-experimental study by Mayer and Davis (1999) showed that improving performance management processes to address employees’ concerns and create a more acceptable process increased trust in top management. These studies how trust relates to accuracy and fairness of HR policies, but as Whitener (2001) emphasizes the content of the policies also influences the development of employees’ trust in the organization.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
156 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Another area that has received some research attention is the role of socalled high involvement work systems in employee trust. High involvement work systems (HIWS) can be seen as a specific set of practices designed to improve communication flow, foster empowerment and participation, and encourage employees to invest both tangibly, as well as emotionally, in their employer (Vandenberg, Richardson, & Eastman, 1999). Examples of practices that are typically included in HIWS are information sharing and employee participation, job security, performance management, and training and development (e.g., Boselie, Dietz, & Boon, 2005; Delery, 1998; Huselid, 1995). Such HIWS relate strongly to employee trust in the organization. For example, Appelbaum, Bailey, Berg, et al. (2001) found that manufacturing employees’ experiences of such a system positively influenced trust in the organization as well as commitment, satisfaction, and stress. Gould-Williams (2003) in his public sector study showed such practices predicted both employee trust in colleagues, managers, and the organization. HRM practices can signal the trustworthiness of the organization (Searle, Den Hartog, Weibel, et al., in press). In addition, these practices clarify and make more predictable for employees what is required to progress in the organization and what they will receive in return for investing effort (Tzafrir, 2005). For example, giving high job security signals organizational benevolence and care to employees (Iles, Mabey, & Robertson, 1990), and also reduces vulnerability for employees, especially in more difficult times. Family friendly policies suggest to employees that the organization is concerned about employee well-being (Grover & Crooker, 1995). This can be a sign of benevolence. The flexibility provided by such practices also aids employees in managing their performance risks: they know that during a family emergency time can be taken off, which can be made up later (Guest, 2002; Perry-Smith & Blum, 2000). Another way of analyzing the relationship of HRM and trust is through considering research on the psychological contract (Conway & Briner, 2009; Guest & Conway, 2002; Robinson, 1996; Rousseau, 1989, 1995; Rousseau & McLean Parks, 1993). The psychological contract describes the employee’s relationship with the organization. A popular definition of the psychological contract is “individual beliefs, shaped by the organization, regarding terms of an exchange agreement between the individual and the organization” (Rousseau, 1995: 9). Beliefs here refer to an employee’s interpretation of implicit and explicit promises made (Conway & Briner, 2009). Rousseau (1989) linked psychological contracts and trust by proposing that unmet employee expectations would ultimately lead to a perceived breach of the psychological contract of which a loss of trust was one unfortunate consequence. Robinson (1996) empirically demonstrated the intertwined nature of psychological contract breaches and trust. In her longitudinal study, she showed that high initial trust in an employment relationship reduced the likelihood of employees’ perceiving psychological contract breach. Over time, unmet expectations eroded
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
157
employee trust in the organization, but where initial trust was lower the reduction was more pronounced. HRM is linked to the psychological contract in several ways (see Conway & Briner, 2009). HRM practices and bundles of practices influence employees’ psychological contract perceptions. For example, the inclusion of more practices appears to indicate to employees (and employers) that more promises have been made between the parties, while inadequate provision is linked to contract breaches (Guest, 1998; Guest & Conway, 2002). This suggests a joint role of HRM and psychological contract processes in developing employee trust: offering new employees a lot of practices or a HIWS may lead to high expectations and thus also to high initial trust of the new employees in their employer. A subsequent failure to deliver these practices adequately, however, can lead to experiencing breaches and a loss of trust. A final issue is the position of the HRM professional. Often the HRM function sits uneasily between the demands of managers and employees (Caldwell, 2003), being charged by the organization with finding satisfactory solutions to meet the demands of both parties. Although normatively committed to trustbuilding models of employment relations, HRM professionals might often paradoxically be central in designing and implementing trust-reducing practices, such as redundancy (e.g., Allen, Freeman, Russell, et al., 2001; Spreitzer & Mishra, 1997). Hence, this group of professionals perform a complex role in the development, implementation, and management of policies, which may variously help to build and erode trust in organizations. We now turn to a related antecedent that also acts to sensitize employees to trust issues: justice. Organizational justice Being treated justly and fairly is related to trust. Lewicki, Wiethoff, and Tomlinson (2005) suggest that trust and justice maybe related in three ways: 1. Justice may act as an antecedent to trust; 2. Justice might be an outcome of trust; or 3. The two may co-evolve. The majority of research focuses on justice as an antecedent of trust. The justice literature distinguishes at least three different types of justice – distributive, procedural, and interactive justice (cf. Colquitt, Conlon, Wesson, et al., 2001). All of these forms are related to trust (see e.g., Cohen-Charash & Spector, 2001; Lewicki, Tomlinson, & Gillespie, 2005). Distributive justice is concerned with perceptions of fairness of outcomes (Adams, 1963). Two meta-analytic studies show that distributive justice is linked to trust in the supervisor (Cohen-Charash & Spector, 2001; Dirks & Ferrin, 2002). This result is somewhat surprising as early studies linked distributive justice with attitudes towards specific outcomes such as pay
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
158 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY satisfaction, rather than with more general assessments such as trust perceptions of the manager or the organization (e.g., Folger & Konovsky, 1989). Yet, distributive justice creates the experience of fulfilled expectations. This implies it can contribute to the experience- or history-based cognitive evaluations of trustworthiness of the party involved in distributing the outcomes and thus strengthen trust in that party. Employee trust not only depends on perceptions of the fairness of allocations of outcomes, but also on perceptions of the procedures used to arrive at such decisions, in other words: procedural justice (Cohen-Charash & Spector, 2001). For example, Folger and Konovsky (1989) found that fair performance appraisals positively affected employees’ trust beliefs relating to their supervisor. This positive relationship between procedural fairness and trust beliefs has been supported longitudinally (Kernan & Hanges, 2002) and across different contexts (Konovsky & Cropanzano, 1991; Pillai, Schriesheim, & Williams, 1999). A recent study unraveled the effect of procedural justice on different types of trust beliefs. Frazier, Johnson, Gavin, et al. (2010) demonstrated that procedural justice mainly affects perceptions of supervisors’ benevolence and integrity, while it seems to have no influence on perceived ability. In addition to affecting trust beliefs, procedural justice is also linked to trust-based cooperative behaviors as research shows this form of justice is a strong predictor of OCB (Cohen-Charash & Spector, 2001; Konovsky & Pugh, 1994; Podsakoff, MacKenzie, Paine, et al., 2000). Interactive justice is based on perceptions of how organizational decisions are communicated and enacted by management (Bies & Moag, 1986). It refers to the day-to-day interactions between managers and employees and thus acts as a constantly updated source of trustworthiness perceptions (Frazier, Johnson, Gavin, et al., 2010). In a qualitative study, Saunders and Thornhill (2003) found that interactive justice, and more precisely the opportunity to ask questions and to participate in decision making, was the strongest justice predictor of employees’ trust in their supervisor. However, it remains unclear whether interactive justice uniquely contributes to trust in the supervisor as shown by Dirks and Ferrin (2002), or whether its effect on trust is irrelevant once procedural and distributive justice are controlled for as Colquitt, Conlon, Wesson, et al. (2001) demonstrate. More research on the unique effect of different types of justice on distinct types and levels of trust is needed to resolve this question. Unique effects of interactive justice have been shown in the area of trust-related behaviors such as organizational citizenship (Colquitt, Conlon, Wesson, et al., 2001) and knowledge-sharing (Kim & Mauborgne, 2003). In this way interactive justice certainly indirectly boosts trust relationships in companies. Two new lines of research promise future insights into the relationship of justice and trust. First, proponents of the multifoci view of justice argue that the perceived source of justice (i.e., whether justice is attributed to the organization, the group, or the supervisor) is likely to be another powerful driver
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
159
of justice-related attitudes and perceptions such as trust (Cropanzano, Byrne, Bobocel, et al., 2001). For example, Rupp and Cropanzano (2002) demonstrate that procedural justice can be attributed either to the organization or to the supervisor, and that depending on this attribution, employees are more likely to show cooperative behavior towards that specific target. Similarly, Frazier, Johnson, Gavin, et al. (2010) describe how the source of justice and the subject of trust are closely intertwined. The second line of research follows from the suggestion of several authors that the interplay of justice and trust can be better understood if the underlying quality of the relationship between supervisor and employee is taken into account (Lewicki, Wiethoff, & Tomlinson, 2005; McAllister, 1995; Tyler & Blader, 2000). Intrinsically motivated communal relationships can be distinguished from self-interest-based exchange relationships based on tit-for-tat exchanges (Clark & Mills, 1993). Extending this to trust holds promise as trust studies have linked communal relationships to affective (McAllister, 1995) and relational trust (Lewicki & Bunker, 1996), while exchange relationships are more aligned with cognitive (McAllister, 1995) and knowledge-based trust (Lewicki & Bunker, 1996). Extrapolating from these findings we suggest that perceived justice might act as an indicator of the quality of relationships. Distributive justice is likely to signal transactional exchange relationships and should thus be clearly linked to cognitive and knowledge-based trust. Procedural and interactive justice, on the other hand, are clearer indicators for communal relationships. Procedural and interactive justice should thus be strongly related to affective and relational trust. Control We now turn to a fiercely disputed “antecedent” of trust: organizational control. Some argue that control is an important antecedent to trust (e.g., Coletti, Sedatole, & Towry, 2005; Sitkin, 1995), while others tend to view control as a detriment for trust-building (e.g., Ghoshal & Moran, 1996; Schoorman, Mayer, & Davis, 2007). Under what conditions distinct types of control relate positively or negatively to employee trust still remains unclear (Costa & Bijlsma-Frankema, 2007). To disentangle the relationship between control and trust, it is first important to define organizational control, as some controversies in the literature seem to stem from a failure to define control clearly and to differentiate views of what organizational control entails. Next, after summarizing the key issues in the ongoing debate on trust and control, we offer some contingencies to this relationship, which might resolve the remaining controversies. Organizational control is defined as a process by which the organization (and manager) regulates, or adjusts, the behavior of employees in the direction of organizational objectives (Cardinal & Sitkin, in press; Challagalla & Shervani, 1997). Control systems form a configuration of formal and
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
160 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY informal control mechanisms (Sitkin, Cardinal, & Bijlsma-Frankema, 2009). Here, we focus primarily on the formal elements of a control system as much of the debate in the literature has centered on the effects of formal control on trust. In an early conceptualization, Edwards (1979) distinguished three generic types of control mechanisms, namely standard specification, measurement and evaluation, and sanction and rewarding (see also Eisenhardt, 1985; Kirsch, 2004). Standard specification essentially entails the definition and specification of goals or means, or both. It involves mechanisms of defining outcomes, behavior patterns, and process enactment. Measurement and evaluation are designed to implement organizational standards and goal attainment through information exchange. Evaluation mechanisms include monitoring, evaluation of progress, and feedback. Finally, rewards and sanctions provide incentives for employees to achieve organizational goals. Mechanisms for contingent rewards and punishments include monetary incentives, praise, promotion, and acts of rebuke. In relation to trust, most researchers have focused either on the effects of the degree of regulation as the outcome of standard setting or on the impact of monitoring on interpersonal trust. Some researchers, often drawing on the human resources tradition (e.g., Argyris, 1964; McGregor, 1960) and sociological theories (e.g., Fox, 1974; Shapiro, 1987; Zucker, 1986), propose that control systems signal distrust (Argyris, 1952; Ghoshal & Moran, 1996; Strickland, 1958), raise relational detachment (Thompson & Warhurst, 1998) and strengthen a divisive “us versus them” perspective between general management and employees (Bijlsma-Frankema, Sitkin, & Weibel, 2007). Others, from a more macro-oriented or controlling/accounting background, argue that formal control may positively affect trust as formal control systems obligate manager and employees equally to adhere to a common set of rules which should raise trust in organizational role incumbents such as managers (e.g., Coletti, Sedatole, & Towry, 2005; Sitkin, 1995). The results of studies based on these different arguments contradict each other as both positions receive some support. A way to reconcile the different views and findings is to separate the effects of formal control on trust-related perceptions from its effect on trust-based decisions or actions (for a similar approach see Ferrin, Bligh, & Kohles, 2007). When we consider trust perceptions, attribution theory provides a helpful frame for assessing the influence of formal control by the manager on trust perceptions of the employee. Attribution theory, in general, analyzes individuals’ causal inferences derived from observation of the behavior of others in everyday life (Heider, 1958; Kelley, 1973; for an overview see Martinko, Douglas, & Harvey, 2006). Attributions – such as positive or negative expectations of the trustees’ intentions – form the basis of the decision to trust (Kramer, 1996). Recent attribution research suggests that, in addition to the evaluation of whether an event is caused by the situation or by the
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
161
intentional act of a person, “relationship quality” should also be taken into account (Eberl, 2004). Individuals distinguish an instrumental from an intrinsic relationship motivation. The latter – like communal relationship motivation – is based on the idea that the relationship itself creates emotional satisfaction. Hence, control that signals an intrinsic relationship motivation can strengthen trust, whereas signals of controllers’ instrumental relationship motivation would weaken trust. In this way, the effect of managerial control behaviors is dependent on the message these behaviors convey. For example, the effect of evaluation-based monitoring on trust differs from the effect of need-based monitoring, with the latter carrying clear signals of intrinsic relationship motivation (McAllister, 1995). Falk and Kosfeld (2006) have tested the effect of evaluation-based monitoring in an experimental setting and show that such a “controlling intent” ruptures trust relationships. BijlsmaFrankema and Van de Bunt (2003) found that need-based monitoring, with a clear aim to provide feedback and support to employees, enhances trust in managers. In analyzing the effect of managerial formal control on trust-based decisions and actions, the risk-reduction function of formal control needs to be considered. By standardizing and limiting available options, formal controls reduce the opportunity for violations, and thus enable a leap of trust, that is a decision to trust to be made, even when prior trust is low (Das & Teng, 1998). However, there is an important caveat to this positive relationship; the trustee needs to be able to show trust-based actions rather than just forced cooperation. For example, if control is too tight and the degree of regulation is too high, cooperation beyond compliance is hardly achievable. This suggests a curvilinear relationship between control and trust-based cooperation, which is in line with findings from related areas (Schoorman, Mayer, & Davis, 2007). For example, the ethical compliance literature shows that strictly control-based ethics programs often produce the unintended effect of employees sticking rigidly to rules rather than actively reflecting and acting upon ethical issues (Weaver, 1999). Similarly, research on standards-contingent rewards (commission or merit pay) reveals that, depending on the intensity and salience of these rewards, negative consequences tend to arise including employees narrowing their focus and concentrating only on objectives that are measured (Weibel, 2007; Weibel, Rost & Osterloh, 2010).
TRUST PROCESSES The dynamic nature of trust over time is mentioned in many trust models, either from development to decline (Currall & Epstein, 2003) or from low to high trust in a deepening relationship (Lewicki & Bunker, 1995). In this section, we review the studies of trust processes. First, we describe the limited research available on trust progression within organizations. We
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
162 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY look at the potential self-reinforcing character of trust (“trust begets trust”) as well as on cycles of distrust. Second, we review the important work on trust repair that focuses on whether and how trust can be repaired if broken or damaged. Trust and Distrust Dynamics While trust tends to be seen as fragile (e.g., Child & Mollering, 2003), the dominant cross-sectional research designs in the field are inherently weak in identifying dynamics in trust. Research on trust patterns and shifts over time is scarce, despite the often mentioned importance of understanding the development stages of trust within an organizational context and thus much of what is discussed in this section is conceptual in nature and still needs to be validated empirically (Ferrin, Bligh, & Kohles, 2008; Lewicki, Tomlinson, & Gillespie, 2006; Serva, Fuller, & Mayer, 2005). When studying trust-building or trust-dissolving processes, a first concern is the initial level of trust in relationships. Some authors hold that trust tends to be relatively high at the start of interactions (McKnight, Cummings, & Chervany, 1998; Robinson, 1996), while others suggest that trust starts low and increases over time (Lewicki & Bunker, 1995). Although empirical research into this matter in an organizational setting is limited, there is some reason to believe that high initial levels of employee trust are more often the “default” initial condition. One reason is that most new employees tend to be attracted to the firm they join and at least to some extent self-select into their new jobs and employers, suggesting they come in with positive expectations (e.g., Searle & Billsberry, In press). Also, as noted earlier, disposition has a role in trust building in first encounters and research suggests that in individualistic countries, trust in strangers tends to be relatively high, at least higher than in collectivistic countries (Huff & Kelley, 2003; Yamagishi, 2003). Hence, at least for individualistic countries, high or at least moderate initial levels of trust seem most likely. Other drivers of high initial trust are close-knit dependencies in networks (Coleman, 1975), previous positive reputations (Burt & Knez, 1995), common group memberships (Dukerich, Kramer, & McLean Parks, 1998), and strong institutional bases of trust such as effective structural safeguards, laws, and regulation (Shapiro, 1987). At the heart of the theoretical work on trust dynamics lays the potential self-reinforcing character of trust (e.g., Ferrin, Bligh, & Kohles, 2008). Self-reinforcement implies that unintended consequences of actions or beliefs lead to the re-creation or amplification of original conditions (Crozier, 1964; Masuch, 1985; Merton, 2010). In its simplest form, a self-reinforcing cycle can be described as a state of reciprocal causality between two variables, although more variables may be involved. Unlike a self-correcting cycle, a selfreinforcing cycle is a deviation, which becomes more distorted over time. In this way an initial (positive or negative) deviation in one variable produces
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
163
a similar deviation in another, producing an amplification effect, resulting in an upward or downward spiral (Lindsley, Brass, & Thomas, 1995). In trust research, these types of spirals are discussed as escalations of trust or distrust often due to a self-fulfilling prophesy, in which expectations (or assumed expectations) affect behavior in such a way that the original conceptions of reality become true (Darley & Fazio, 1980). Within trust research there is wide recognition of the reciprocal nature of trust; implicit within this conception is the idea that “trust lubricates cooperation, and cooperation itself breeds trust” (Nahapiet & Ghoshal, 1998: 255). Ferrin, Bligh, and Kohles (2008) propose that there are three routes as to how trust could beget trust. First, they argue that trust perceptions could spark the partners’ trust perceptions through the communication efforts between the parties. Second, cooperation (trust-based or not trust-based) might lead to a reciprocated cooperation via a tit-for-tat strategy fed by a general norm of reciprocity. Finally, they argue that the trusting employee’s positive expectations of the transaction partner can result in trust-based behaviors and that these behaviors in turn affect the trustworthiness perceptions that party holds. In a dyadic simulation study, Ferrin, Bligh, and Kohles tested all possible relations and concluded that “an individual observes another’s cooperative behavior, next develops a conclusion about the other’s trustworthiness based on that observation, and then performs a reciprocation behavior based on that conclusion” (Ferrin, Bligh, & Kohles, 2008: 171). Similarly, an experiment by Serva, Fuller, and Mayer (2005) showed that behavioral consequences drive reciprocal trust. They also found that some behaviors generally thought to be strong signals for trust, such as delegation, negatively affected trustworthiness perceptions. As in the case of control, the intentions of the delegating party appear to have a role. Further research is required to decipher precisely when, and why, specific behaviors are regarded as a signal of trust in the complexity of real life situations. While simulation studies and experiments enable researchers to control drivers of trust dynamics, they lack the context and richness of longitudinal field studies. Ethnographic and qualitative studies might be better suited to the identification of patterns and shifts in trust over time, and in providing insight into which behaviours matter. Distrust, too, can trigger a self-reinforcing cycle; however, with partially different drivers from those for trust cycles (Ghoshal & Moran, 1996; Zand, 1972). Distrust rests on confident negative expectations, which result from the perceived value incongruencies between the parties (Lewicki, McAllister, & Bies, 1998; Sitkin & Roth, 1993). In general, A will distrust B if B is perceived not to share, or even betrays, the core values of A. Hence, in the vicious cycle (distrust begets distrust), actions of another party are negatively construed as involving value incongruence and the ensuing reactive counter-behavior is then similarly construed in value incongruence terms (Bijlsma-Frankema, Sitkin, & Weibel, 2007). In their field study, Bijlsma-Frankema, Sitkin, and Weibel showed vicious cycles of distrust are driven by distrust-based behaviors, by
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
164 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY avoidance of interaction, and by amplification of value incongruence. Hence, as is the case with trust, distrust-based behaviors promote a re-evaluation of the perceived intentions of the other party. Unlike trust, however, such cycles are more difficult to stop because a typical consequence of rising reciprocal distrust is the evasion of interaction with the other party (March & Olsen, 1975), which prevents a re-evaluation of the situation. Finally, research has also investigated the so-called vicious cycle of bureaucracy with its negative effect on trust-based relationships (Argyris, 1957; Blau, 1963; Ghoshal & Moran, 1996). Viewed from a trust perspective, this can be seen as a special form of the distrust cycle. For example, the classic work by McGregor (1960) describes how distrusting managers (who operate on the basis of theory X assumptions) rely on strict managerial control and sanctions. These distrust-based behaviors are said to trigger negative feelings of being distrusted and negative reciprocity. Research on the undermining effect of restrictive controls shows that these “X-based beliefs and behaviors” by managers alter the intentions of the employee in terms of their intrinsic motivation to cooperate. This even implies such X-behaviors undermine the foundation of an employee’s trustworthiness (Deci, Koestner, & Ryan, 1999; Frey, 1997; Frey & Jegen, 2001). This so-called crowding-out, or corruption, effect was shown to occur when individuals think their behavior is under the control of extrinsic factors, for example, performance-contingent rewards (Weibel, Rost, & Osterloh, 2010) or strong evaluative monitoring (Falk & Kosfeld, 2006). Trust Repair The well-documented current financial crisis has revealed prominent examples of organizational level failures, such as the UK’s Royal Bank of Scotland (Banking Times, 2010; BBC, 2010; The Guardian, 2009), but other well-publicized unethical incidents that shake trust in organizations have also occurred, not least the Enron case (Gillespie & Dietz, 2009). Yet, our knowledge of how to repair trust once broken is limited. Trust violation and repair have been conceptualized and studied both at the individual and organizational levels (e.g., Ferrin, Kim, Cooper, et al., 2007b; Gillespie & Dietz, 2009; Ren & Gray, 2009; Sitkin & Roth, 1993). Implicit in the research is an assumption that trust is positively valued in organizations and, therefore, its restoration is important. In this section we first explore definitions of trust violation and repair. Next, we examine four distinct approaches to breach and repair and, finally, we highlight some significant differences between trust breach at an interpersonal level and organizational level violations. In defining trust repair, there is some agreement. For example, researchers mainly concur that repair requires an earlier perceived violation, reducing current levels of trust in the other party (Dirks, Lewicki, & Zaheer, 2009; Kim, Dirks, & Cooper, 2009; Lewicki & Bunker 1996). Employees may perceive a violation because they realize that their dependency has been exploited or that
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
165
the organization or the supervisor as failed to fulfill perceived obligations. Yet, trust violations can emerge without an obvious infraction occurring between parties; for example, the trusting employee may witness a colleague being so poorly treated by the manager that they no longer trust this manager, without having been badly treated themselves. Perceptions of a breach can thus spill over to those not directly involved (Kim, Ferrin, Cooper, et al., 2004). There are also some issues that are still debated. For example, although authors agree repair focuses on the transition from the current negative to a more positive position before trust can be re-established (Sitkin & Stickel, 1996), what the aim of restoration should be is less clear. Should trust repair focus mainly on re-establishing a workable exchange relationship? Or is a much deeper level of forgiveness needed, as, for example, in cases where violation has led to extreme affective reactions such as vengeance and retaliation (e.g., Bies & Tripp, 1996)? Also, cognitive, behavioral, and affective aspects of repair processes can be distinguished (e.g., Dirks, Lewicki, & Zaheer, 2009), but we need to uncover whether the full range of components is needed for authentic trust repair. There are several processes that form the focus of work on interpersonal trust repair: attributional (e.g., Tomlinson & Mayer, 2009), social equilibrium (e.g., Ren & Gray, 2009), structural (e.g., Nakayachi & Watabe, 2005), and temporal processes (e.g., Kim, Dirks, Cooper, et al., 2006; Lewicki & Bunker, 1996; Ren & Gray, 2009). We describe each of these four in turn below. The dominant approach to trust repair focuses on attributions. This approach suggests that whether and how trust violations need to be addressed will depend on how the breach is framed and attributed to the perpetrator (e.g., Gillespie & Dietz, 2009; Kim, Dirks, Cooper, et al., 2006; Rhee & Valdez, 2009; Tomlinson & Mayer, 2009). Here, trust repair is conceptualized as a cognitive process. For example, Tomlinson and Mayer (2009), building on Weiner (1986), suggest that trust breakdown involves three key causal attributions – locus of causality, controllability, and stability – all of which need to be dealt with in trust repair. Also applying an attributional lens, Kim, Ferrin, Cooper, et al. (2004) demonstrated the importance of the type of violation for repair. In a scenario study, they compared two types of response to a breach: apology and an admission of culpability versus the denial of responsibility. Response effectiveness was contingent on the type of violation: a competencebased violation was better appeased by an apology, whereas denial formed the more effective response for integrity breaches. However, a caveat to these findings is that if the mistrusted party was guilty, apologizing always appeared to be a better strategy (Kim, Ferrin, Cooper, et al. 2004). In contrast, social equilibrium models understand the consequences of trust breaches as a disequilibrium in both the relationship and a social context (e.g., Bottom, Gibson, Daniels, et al., 2002; Reb, Goldman, Kray, & Cropanzano, 2006). In other words, trust breaches occur if offenders violate victims’ expectations of how they should be treated given their relative standing in the social context (Ren & Gray, 2009). Trust repair is then focused on the social
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
166 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY processes to restore the equilibrium. Building on Goffman (1967), Dirks, Lewicki, and Zaheer (2009) suggest that appropriate social rituals, such as penance and punishment, might restore the equilibrium. Related to this approach, studies of the counterproductive behavioral consequences of trust violations, such as revenge and deviant outcomes (Bies & Tripp, 1996, Tripp, Bies, & Aquino, 2002) and non-cooperative behaviors (Bottom, Gibson, Daniels, et al., 2002), highlight how firms with high justice climates tend to have better procedures that deal formally with violations. This makes forgiveness and reconciliation more likely (Aquino, Tripp, & Bies, 2006, Tripp, Bies, & Aquino, 2007). In contrast, in firms without formal channels individuals are suggested to be far more inclined to seek their own redress after violations (Tripp, Bies, & Aquino, 2007). A third approach to repair highlights the role of formal structures, such as contracts and covenants, which make elements of future exchange explicit and thus reduce the likelihood of further breaches (e.g., Gillespie & Dietz, 2009; Nakayachi & Watabe, 2005; Sitkin & Roth, 1993). Structural factors can be separated into two components: legal remedies, which include incentives, penalties or monitoring (Sitkin, 1995); and social mechanisms, which might include creating obligations between the parties. Finally, some authors focus on the temporal stages of repair (e.g., Lewicki & Bunker, 1995). This perspective emphasizes the necessity of following a sequence of stages (e.g., Kim, Dirks, Cooper, et al., 2006; Kim, Ferrin, Cooper, et al., 2004; Ren & Gray, 2009). Dirks, Lewicki, and Zaheer (2009), for example, outline four stages: 1. 2. 3. 4.
Developing an understanding of initial trust before breakdown; Identifying the breakdown or violation event(s); Engaging in efforts to repair the broken trust; and Post-repair.
Although most stage models focus on the interpersonal level, their utilization for the repair of trust in organizations is also being considered. For example, Gillespie and Dietz’s (2009) conceptual framework of organizational level trust repair commences with the importance of an immediate response, including both a verbal acknowledgement of the incident (e.g., faulty products or inappropriate conduct of role incumbents) and regret, but also action against those currently known causes. The second stage involves a more in-depth diagnosis of the problem, leading to reform, with two distinct interventions: a verbal apology, plus internal actions to attend to issues identified in the earlier diagnosis. The final stage is an evaluation of the repair. Gillespie and Dietz hypothesize the initial two stages might be significant to regulate distrust in the firm, whereas the latter two are concerned with demonstrating the organization’s trustworthiness. Additional complications emerge, however, when considering repair at an organizational level. For example, it is often difficult to discern precisely which
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
167
stakeholders are affected by a breach (Mahon, 2002). A firm may have multiple direct and indirect targets for repair. Also, the firms’ repertoire of potential responses is far greater than that of any single individual involved in an interpersonal trust breakdown. For example, aside from an apology or denial, they can also scapegoat the previous management team and thus become almost a new entity by replacing perceived perpetrators with less tainted individuals (Tschannen-Moran & Hoy, 2000). In addition, a greater range of organizational system components can be utilized in repair. These can include both external dimensions of governance and public reputation, and more internally focused factors, such as strategy, leadership and management practice, culture, and structure and policies (Gillespie & Dietz, 2009). For example, in the recent global financial crisis, repair effort occurred simultaneously on two fronts: through individual organizations changing components, such as staff members and structures, but also through attempts to restore trust in the sector as a whole by external regulation. The impact of different efforts on trust might differ. Finally, it may be important to take into account the firm’s perceived intention in the original trust breach (e.g., was the organization out to deceive clients?), and in managing the fallout (e.g., is the perception that they just want to cover up or that they genuinely care?). Empirical work on these dynamics of repair is still relatively scarce, and to date we have little evidence on whether the proposed progressions through each stage are necessary for repair at individual or organizational levels. In addition, the dynamic interactions between the parties involved in terms of action and re-action have also not received sufficient attention. However, access to field data in this area is often denied as firms are reluctant to admit trust-breach related problems. Instead, trust breaches and repair strategies might be studied with archival data. Employees’ reactions might be simulated with scenario techniques as demonstrated by Kim, Ferrin, Cooper, et al. (2004). We now turn to the most important consequences of trust in organizations.
OUTCOMES OF EMPLOYEE TRUST Trust is claimed to have a number of key benefits for organizations and their members (Kramer, 1999) and in line with these claims studies show that trust directly affects performance in and of organizations (for overviews see Colquitt, Scott & Lepine, et al., 2007; Dirks & Ferrin, 2001, 2002). In addition, trust also indirectly boosts performance as it fosters desirable work-related behaviors and a cooperative climate that in turn are conducive to (organizational) performance. Performance Two recent meta-analytic studies have analyzed how employee trust is related to individual performance. Dirks and Ferrin (2002) showed that employee trust in leaders is weakly related to job performance, whereas employee trust
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
168 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY in the organization does not seem to affect job performance. Colquitt, Scott, and Lepine (2007) also addressed trust in managers and found a moderate effect of employee trust on individual level job performance. They also evaluated whether the specific measurement of trust moderates the relationship between employee trust and job performance. Three popular types of trust measurement were distinguished: trust as positive expectations (e.g., Cook & Wall, 1980), trust as willingness to be vulnerable (e.g., Mayer & Davis, 1999), and direct trust ratings (e.g., Brockner, Siegel, Daly, et al. 1997). Colquitt, Scott, and Lepine (2007) showed that the effect of trust on performance is robust across these different measures; that is, the type of measurement did not moderate this relationship. A number of studies have examined the effects of employee trust on unit performance. The findings are inconsistent. Although Dirks (2000) found that teams in the NCCA basketball league who trusted their leader significantly outperformed teams with low trust in their leader. Most other studies of the relationship between trust and unit performance have not found significant effects (e.g., Dirks & Ferrin, 2002). Often, these non-findings are explained by the fact that the impact of trust on unit performance is less direct. Trust is thought to influence group and organizational level performance positively via its effects on more proximate behavioral and attitudinal concepts. We review some of these effects in the following sections. Another way to explain why trust has seldom been found to affect unit performance is that research has yet to explore this relationship from different angles. For example, Deutsch-Salamon and Robinson (2008) argue that organizational performance is more likely to be influenced by employees’ perceptions of being trusted by management than by employee trust in management. In their longitudinal study of 88 retail stores, they showed that where employees are trusted the organization’s sales and customer service performance increases. Knowledge Sharing and Innovation The relationship between employee trust, learning, knowledge-sharing, and innovation has a long history (e.g., Argyris, 1964). Trust has been found to be an antecedent to both pupils’ (Rigby, Deci, Patrick, et al., 1992) and employees’ (Gagne & Deci, 2005) intrinsic motivation to learn. For example, in a field experiment, Black and Deci (2000) found that chemistry students who had a supportive and trust-based relationship with their instructors developed higher internalized motivation for learning and gained better grades than fellow students who did not feel as supported and trusted. Most of the research to date on the relationship between trust and willingness to learn was conducted in educational contexts and although in the work context it still awaits rigorous testing, replication of these results seems likely, as Osterloh and Frey (2000) propose.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
169
Trust also positively affects knowledge exchange. It enhances knowledge exchange in inter-organizational networks (Dyer & Nobeoka, 2000), in knowledge communities (Brown & Duguid, 1991), across teams (McDermott, 1999), and more generally inside companies (Collins & Smith, 2006; Huemer, Von Krogh, & Roos, 1998; Husted & Michailova, 2002). There are two reasons for this positive effect of trust. First, trust enhances employees’ willingness to share knowledge (Andrews & Delahaye, 2000). Second, trust is also seen as a core element of a knowledge-supportive organizational culture (Von Krogh, 1998). Supportive trust-based cultures enhance cooperation, which in turn foster the sharing of knowledge (McNeish & Mann, 2010). Evidence also suggests a direct link between trust and innovation processes (Newell & Swan, 2000). Innovation can offer tangible savings and benefits for organizations, including reducing lead times to market and the identification of information about often hidden problems. Innovation is often linked to problem-solving processes and thus how organizations identify and utilize information about errors relates to innovation. Trust fosters these benefits because it encourages employees not to fear revealing their errors (Edmondson, 2004). In particular, employee trust generates access to information on mistakes which less trusting employees might keep hidden. For example, in her study of innovation within the health context, Edmondson (2004) showed how staff became more willing to identify mistakes where they felt safe and could trust that their errors would not be punished harshly. Trust made it possible for this often hidden information to emerge, enabling teams to become more aware, identify problems, and find novel solutions to the issues. Discretionary Behaviors Discretionary behaviors such as extra-role activities, contextual performance, and OCB are an often suggested consequence of employee trust (Konovsky & Pugh, 1994; McAllister, 1995; Pillai, Schriesheim, & Williams, 1999; Podsakoff, Mackenzie, Paine, et al., 1990). In their meta-analysis, Dirks and Ferrin (2002) found that trust in leadership is significantly linked to all facets of OCB, with a somewhat more pronounced impact of trust on conscientiousness, altruism, and sportsmanship. Similar strong effects of trust on discretionary behaviors were shown in meta-analytic studies (Colquitt, Scott, & Lepine, 2007; Podsakoff, Mackenzie, Paine, et al., 2000). In their overview, Mayer and Gavin (2005), too, note that each study on trust and OCB they evaluated clearly demonstrated a positive relationship between these two variables. Yet, a number of issues still merit some further attention. First, the influence of trust on discretionary behaviors may currently be underestimated, as trust is likely to be an important mediator for several other variables studied as antecedents of discretionary behaviors. Second, trust may also act as a moderator for some factors relating to discretionary behaviors. We discuss both the potential mediating and moderating role of trust below.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
170 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY The willingness of employees to show discretionary behavior such as OCB is often explained in terms of social exchange. Employees are willing to show these behaviors trusting that their manager, colleagues, or employer will reciprocate in time. Most researchers drawing on social exchange theory to explain discretionary behavior thus view trust as an important – although mostly untested – mediator variable. More precisely, discretionary behaviors are proposed to be an outcome of a communal, rather than a transactional, exchange relationship. Konovsky and Pugh (1994) were among the first to test whether relational-trust should be seen as a signal for the type of exchange relationship between employee and employer. They propose that procedural fairness feeds into relational-trust beliefs, signaling a community exchange condition which then is reflected in employees’ willingness to show discretionary behaviors. Their empirical study indeed demonstrates this hypothesized mediating effect of trust. Other evidence confirms the reciprocal pattern of such communitybased relationships. Trust in management and perceptions of managers’ trustworthiness is influenced by employees’ discretionary behavior (Chiaburu & Lim, 2008; Frazier, Johnson, Gavin, et al., 2010), but also how much employees perceive their managers trust them (Korsgaard, Brodt, & Whitener, 2002; Lester & Brower, 2003). As indicated above, trust has also been proposed as a moderator in relation to discretionary behavior. For example, Coyle-Shapiro (2002) suggested that trust moderates the impact of expectations of psychological contract fulfillment on discretionary behaviors. In a longitudinal field study she demonstrated that employees who trusted their employers were more willing to show discretionary behavior because they perceived a low risk of contract breach. In a series of laboratory experiments, Bradner and Mark (2002) showed that where the relationship between the parties is more distant, fewer discretionary behaviors are displayed. Trust was found to moderate the effects of distance on relationships, helping to overcome the negative impact of distance. However, Yakovleva, Reilly, and Werko (2010) studying teams in an organizational field setting, observed the exact opposite effect: in distant teams the effect of trust on discretionary behaviors was less strong than in co-located teams. They suggest that their findings can be explained by the fact the trust is less salient in virtual teams as felt vulnerabilities are likely to be lower for non-co-located teams. Motivation to Act Another important outcome of trust is autonomous motivation. Selfdetermination theory proposes that autonomy, support, and caring, which we labeled earlier as relational trust, is a fundamental prerequisite for autonomous motivation to develop (Gagn´e, 2003; Grant & Sumanth, 2009). Autonomous motivation is internally regulated, based on the satisfaction an individual derives from involvement in particular activities. Activities are valued for their own sake and are undertaken without any reward except the
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
171
activity itself. Autonomous intrinsic motivation refers to a satisfying flow of activity (Csikszentmihalyi, 1991). Examples include skiing, playing a game, or solving interesting puzzles. This type of motivation impacts on performance as it is linked to creativity, innovation, and discretionary behaviors (e.g., Amabile, 1979; Gagn´e, 2003). Autonomous prosocial motivation implies taking the wellbeing of others into account without expecting a reward (Grant & Sumanth, 2009). The welfare of the community enters into the preferences of the individual. Examples of prosocially motivated activities include: volunteering for an animal shelter, or fundraising for a social organization (cf. Frost, Osterloh, & Weibel, 2010). Prosocial motivation, too, is an important antecedent to performance as it has been linked to cooperation, helping, and OCB (Finkelstein & Penner, 2004). Proponents of self-determination theory, Ryan and Deci (2000) argue that we are naturally inclined to be intrinsically and prosocially motivated, given proper nurturing. Such nurturing comes about if individuals feel they can freely choose to pursue an activity (satisfying the need for autonomy), when they master an activity (satisfying the need for competence), and when they feel connected to and supported by important reference persons (satisfying the need for relatedness). Trust is linked to all three facilitating conditions. Trust directly satisfies employees’ need for relatedness. In an intervention study, Deci, Connell, and Ryan (1989) demonstrated that training supervisors to demonstrate their trust in employees and to build employees’ trust by showing care and empathy towards the employees resulted in a more favorable work climate and strengthened perceptions of relatedness. Trust of the supervisor and employees’ trust in the supervisor are also likely to nurture autonomy and competence needs. Several studies have demonstrated that trust is a precondition for managers to delegate decision rights and thus to grant employees autonomy (Leana, 1986; Whitener, Brodt, Korsgaard, et al., 1998). Spreitzer and Mishra (1999) in a survey study show that managerial trust is linked to employee participation, and that employee participation is linked to unit level performance. The possible mediating effect of intrinsic motivation, however, was not tested in this study. Finally, trust is also likely to satisfy employees’ need for competence. Trust has been found to heighten perceptions of self-efficacy (Hsu, Ju, Yen, et al., 2007) and trusted employees feel more empowered (Gomez & Rosen, 2001). Hence, trust should also affect positively competence experience. We thus expect trust has a positive impact on autonomous motivation through different routes. However, some of these relationships have yet to be tested empirically. Attitudinal Impact Trust also affects job-related attitudes. Trust has long been demonstrated to impact upon job satisfaction (for meta-analytic results see Dirks & Ferrin,
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
172 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY 2002). For example, Discroll (1978) surveyed the faculty of a liberal arts college and showed that trust in university officials had a significant impact on overall job satisfaction. Most job satisfaction studies have emphasized cognitive aspects of satisfaction at the expense of affective components. Hence, not surprisingly, Dirks and Ferrin (2002) found that above all cognitive and knowledge-based trust positively relates to job satisfaction. More recently, however, Yang and Mossholder (2010), adopting a multi-foci, multi-basis approach to the assessment of job satisfaction not only confirmed the importance of the impact of knowledge-based trust in the organization on job satisfaction, but demonstrated that in addition affective, relational-trust in the supervisor also has an effect on job satisfaction. The affective components of job satisfaction are also captured in the more encompassing concept “well-being.” Trust has been found to be an important antecedent to general well-being. For example, Helliwell and Wang (2010) demonstrate by analyzing the Gallup World Poll and the cycle 17 of the Canadian General Social Survey a strong effect of general and interpersonal trust on life-satisfaction and well-being. Unfortunately, however, although proposed conceptually (Chughtai & Buckley, 2007), to our knowledge no empirical study has investigated whether well-being in the workplace is an important outcome of employee trust, nor have there been studies differentiating between the effects of affective and cognitive trust on well-being. Organizational commitment is another suggested attitudinal consequence of trust (e.g., Aryee, Budwar, & Chen, 2002; Pillai, Schriesheim, & Williams, 1999; Tan & Tan, 2000). Commitment can broadly be defined as “a force that binds an individual to a course of action that is of relevance to a particular target” (Meyer & Herscovitch, 2001: 301). Three forms of organizational commitment have been postulated: affective (emotional attachment, desire to remain), normative (felt obligation to remain), and continuance (need to remain due to loss of investments or lack of alternatives) commitment (e.g., Meyer, Bobocel, & Allen, 1991). Trust has been linked above all to affective commitment. For example, a recent multi-foci study identified how trust in both co-workers and the organization were positively associated with affective commitment (Tan & Lim, 2009). Yet, some researchers have noted that trust might also be closely connected to normative commitment (Lewis & Weigert, 1985). Den Hartog and De Hoogh (2009) find positive relationships of both trust in colleagues and trust in management with affective and normative commitment. In their study, trust in management also had a low, but negative, relationship with continuance commitment.
FUTURE DIRECTIONS The above review shows that research on the different bases and foci of trust is still fragmented. To date, trust research has primarily focused on the interpersonal side, with far less attention given to employees’ trust in collectives
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
173
such as organizations. More work in that area is needed to clarify whether the same drivers operate for trusting collectives as for trusting other individuals. Furthermore, to date much of the research attention on the interpersonal basis of trust has been directed towards the employee as a trustee of the supervisor, rather than for other relationships employees have in the workplace. Recent research has begun to explore the reciprocal nature of trust to reveal how actions and reactions of others impact upon, and alter, trust (e.g., Gerbasi, Cook, Jody, et al., 2008). More attention is being paid to trust in dyads, particularly looking at the impact of asymmetries between the parties (De Jong & Dirks, 2010), and the impact of being trusted for employees (e.g., Deutsch-Salamon & Robinson, 2008; Dirks & Skarlicki, 2009). There are, however, several other areas in the trust field that are in need of future research. Nascent Topics in Employee Trust Research There are some emerging issues that lie beyond these traditional lines of inquiry that are worthy of exploration. These include an analysis of the interplay of trust and organizational and societal cultures, and the nature and role of dark side of trust. Trust and organizational culture To date, very few trust researchers have examined the links between trust and organizational culture. Such links can be conceptualized in three ways: culture as an antecedent of trust, culture as an outcome of trust, and trust as a core value in an organizational culture. A number of trust researchers implicitly view strong organizational cultures, or in other words cultures where values are broadly agreed upon and intensively approved (O’Reilly, Chatman, & Caldwell, 1991), as an antecedent to trust. For example, Sitkin (1995: 188) argues that value-based trust is pertinent if “the other party’s beliefs and values are perceived as being congruent with your own.” In similar vein, Lewicki and Bunker’s (1996) identification-based trust is based on a mutual understanding such as that one party can think and act in accordance with the other party. Extrapolating from the communities of practice literature (Brown & Duguid, 1991; Wenger, 2000), one can also argue that a shared culture is an outcome of interpersonal trust. Wenger (2000), for example, argues that for a community of practice to arise members must have a strong sense of mutuality. In other words, they must trust one another and trust that all will seek to contribute to the community. Although communities are obviously not synonymous with organizational cultures, as they exist for specific exchanges and can even transcend organizational boundaries, this work does suggest that trust might help build a shared culture.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
174 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Finally, trust is often seen to be an important core value of organizational culture. For example, the knowledge management literature argues that a caring culture, which is characterized by mutual trustworthiness perceptions and trust-based actions, is responsible for the willingness of individuals to share knowledge with each other (Von Krogh, 2002). We suggest that two approaches are needed to further disentangle the relationship between trust and organizational culture. First, the conceptual difference between certain types of trust, such as identification-based trust, and organizational culture need to be sharpened; identification-based trust is likely to be a more local concept than organizational culture and therefore identity processes, such as salience and border delineation, need to be incorporated in order to understand the evolution of identification-based trust. Second, the interplay of trust and culture can best be studied in situations of change. For example, research could examine whether trust is an antecedent, a moderator, or an outcome of a new corporate culture in the wake of a merger. Trust and societal culture Although intuitively one would assume that trust is so basic to human interaction that trust processes will occur across many if not all societal cultures, exactly what makes people trust others and how such trust is enacted may still differ from one culture to another. Important questions thus include the extent to which manifestations of trust and trustworthiness appear similar within and across different cultures and whether the same drivers of trust are relevant within and across those cultures. For example, do people in societies that are strongly uncertainty avoidant trust to the same degree or on the same bases as those who are from a society that embraces risk-taking much more? Do leaders in high power distance cultures use the same ways to demonstrate trust in subordinates as those in more egalitarian societies? Although these seem important questions, trust has not often been included in crosscultural research. One way to address culture differences in research is to include culture dimensions. Hofstede’s (1980, 2001) important but also heavily criticized work proposed four such dimensions (individualism-collectivism, power distance, masculinity, and uncertainty avoidance). Later work added a fifth one (future orientation; e.g., Hofstede & Bond, 1988). In an attempt to overcome some of the psychometric problems and other critiques of Hofstede’s work (e.g., Spector, Cooper, & Sparks, 2001), later studies, such as the GLOBE study (House, Hanges, Hanges, et al., 2004), have tried to better operationalize these dimensions, in addition to several other dimensions of culture, not least performance orientation. Kirkman, Lowe, and Gibson (2006) extensively reviewed the research using Hofstede’s (1980, 2001) culture dimensions and included studies looking at the main or moderating effects of culture on different levels of analyses. Their analysis revealed very few cross-cultural studies
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
175
that explicitly take trust into account. An exception, conducted in Hong Kong, showed that power distance at the individual level moderated the relationship between procedural justice and trust in the supervisor, such that the relationship between procedural justice and trust in supervisor was stronger for those lower, rather than higher, in power distance (Lee, Pillutla, & Law, 2000). There are few large-scale empirical studies that focus more specifically on trust in different cultures and take a wide sample of people from many different cultures to test comparative hypotheses (such as the aforementioned GLOBE study in the leadership area; cf. House, Hanges, Hanges, et al., 2004). However, some related studies compare trust relevant variables between people from a few backgrounds (e.g., USA and Japan or Singapore). Ferrin and Gillespie (2009) review the available studies to date that contribute knowledge on the effects of national-societal culture on interpersonal trust development. Their findings suggest that cultural-specific determinants of trust exist including: cultural differences around power distance, risk-taking, and reciprocity, as well as several country-level macro-economic and institutional factors. In addition, they found some determinants and processes of trust that seem to be more universal, including personal and impersonal communication and strategies for repairing inter-personal trust (e.g., apology and denial). They also suggest that the inter-personal trustworthiness characteristics of ability, benevolence, and integrity are ubiquitous determinants of trust. Yet, in some countries, additional culture-specific aspects of trustworthiness seem to exist. Ferrin and Gillespie (2009) caution that their findings are preliminary, due to the very limited body of available work. Few replications of studies exist and, to date, only a limited number of countries or cultural dimensions have been examined empirically. Clearly, therefore, this is a worthwhile area ripe for further exploration. Li (2010) notes that the limited literature that does exist on cross-cultural trust has primarily focused on the comparative study of intra-cultural trust (trust between people of the same cultural background) rather than intercultural trust (trust between people of different cultural backgrounds). In the words of Gelfand, Erez, and Aycan (2007: 497): “far less attention has been paid to the dynamics of culture in intercultural encounters, or what we would refer to as the ‘cross-cultural interface’.” In other words, our knowledge of how trust is built and maintained when people from different cultures interact is even more limited than our knowledge of how trust and the drivers of trust may be different within certain cultures. Several theoretical models that relate to cross-cultural trust development have been proposed. For example, Johnson and Cullen (2002) suggest that national culture affects which bases of trust are specifically relevant (e.g., calculation, knowledge, reputation, dispositional trust). The base of trust is manifested in individuals’ trust-related behaviors (i.e., cooperation, negotiation, leadership), which in turn can signal trustworthiness in a cross-cultural relationship. This framework recognizes that two parties may operate with
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
176 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY different bases of trust and, therefore, they may value or interpret distinct and different signals of trustworthiness. In contrast, Doney, Cannon, and Mullen (1998) building on Hofstede’s (1980) dimensions, argue that cultural dimensions affect the way trustors develop trust in another party. They suggest that trust is more likely to occur when the parties share norms and values. These new developments show the importance of looking more critically at Hofstede’s original work (e.g., Kirkman, Lowe, & Gibson, 2006; Spector, Cooper, & Sparks, 2001). Summarizing, some universal as well as cross-culturally contingent elements and drivers of trust seem to exist, but we do not yet know enough about what these are. In addition, trust can have a role in interactions between people from different cultures, which forms another interesting but as yet under-researched topic. The dark side of trust The dark side of trust has received surprisingly little attention in the literature. If the “dark side of trust” is considered at all, then the focus tends to be on the problem of displaced trust. For example, McEvily, Perrone, and Zaheer (2003: 100) explain how trust “can lead the trustor astray” and “may produce systemic biases that can result in judgments that are substantially flawed and costly.” To view the downside of trust as displaced trust, however, misunderstands a fundamental aspect of the trusting process, which is that of the possibility of betrayal – misplaced trust – is one of its defining conditions (Skinner, Weibel, & Dietz, 2007). Very few authors adopt a more skeptical position on “rightly placed” trust. Gambetta (1988) offers a very lucid analysis of how, in certain circumstances, the bonds of trust may be problematic, referring to the powerful bond among Mafiosi. From his study questions arise how trust can exuberate ingroup and outgroup conflicts. Sievers (2003) and Hardy, Phillips, and Lawrence (1998) show how trust can be used as a facade to manipulate people. For example, in a case study of a plant failure, Hardy, Phillips, and Lawrence (1998) illustrate how the management team by upholding a facade of trust fooled the union into accepting unfortunate terms for their members. This study begs the question under what circumstances trust can be misused as a manipulation device and whether and when trustors are able to spot the deception.
SUMMARY AND CONCLUSIONS Employee trust is an important topic for both researchers and practitioners to consider. In our review we have focused specifically on when and why employees trust other individuals, such as a colleague or leader, as well as the organization as a whole. We have explored trust processes and highlighted the dynamic nature of trust as well as work on what is needed to repair trust
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
177
if and when it is broken. We also examined the impact of trust on different types of outcomes, such as performance, innovation and learning, motivation, attitudes, and discretionary behavior. We have highlighted available findings pertaining to a number of individual and contextual factors that drive such trust and outlined some potentially fruitful directions for future work. From this review, a number of conceptual and methodological recommendations emerge. Conceptual and Measurement Issues To date, trust research has been very productive with numerous studies scrutinizing the antecedents, processes, and outcomes of trust. Yet, despite the breadth and richness of such research, there has been a disappointing lack of coherence in the pattern of findings, particularly when it comes to the outcomes of trust. In addition, limited attention has considered the impact of organizational level on trust findings, or on the strength of different contexts’ influence on the salience of trust. In part, inconsistent results may arise from the lack of a coherent trust scale, with most researchers seemingly capturing trust as beliefs concerned with the trustees’ ability, benevolence, and integrity. In reality, however, each scale has actually measured very different aspects of these dimensions. Yet, rather than addressing these issues upfront, most research today has glossed over these differences. There is a requirement for a more nuanced set of scales to be developed which vary systematically with the context. Further development is also required in order to differentiate between the different types of trust, such as calculative-based, knowledge-based, relationalbased, and identification-based trust. Research tools need to be broadened to enable examination how organizational identification and value congruence relates to employees’ trust. Although fruitful research has explored cultural differences in trust and distrust, more attention is needed in the creation of robust culturally sensitive trust scales. The current domination of trust as a belief needs to be extended by researchers to allow further examination of the two other constituent parts of trust which Dietz and Den Hartog (2006) highlight: trust as a decision and as a set of behaviors. Until we add these distinct components to our analysis of employee trust, we have only a partial view of employee trust in organizations. This is especially significant in the area of trust processes, which are driven by trust decisions and trust-related behaviors. Thus, the measurement and integration of these two components are pivotal to understand the selfperpetuating nature of trust (or distrust). A second set of recommendations concerns the application and development of a richer variety of research methods than those previously adopted in the trust literature. As shown, trust research has been dominated by self-reported survey research and by laboratory experiments with students. Insight into the
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
178 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY nature of the relationship between the employee and their organization could be enriched through the application of longitudinal and qualitative research designs. Examination of which agent, or sets of agents, employees focus on when they formulate their perceptions of the organization and trust require more open-ended and qualitative approaches. Exploration of the triggers and dynamics of trust calls for longitudinal designs to reveal how employees’ perceptions, decisions, actions, and subsequent reactions concerning trust shift over time. Longitudinal approaches are also important to the examination of how trust is signaled or jeopardized at the very onset and early in the employer–employee relationship. It is important to see how such clues may vary across different organizational and cultural contexts. Studies of trust repair at the inter-personal level (e.g., Kim, Ferrin, Cooper, et al., 2004) have revealed the power of scenario design to tease out and test systematically key elements; similar approaches may also be of value at the organizational level. Alternatively, explorations of trust repair may be better served by methods that capture the processes and events as the change over time in real situation, through using diary studies (Searle, 2010) and event-contingent procedures (Wheeler & Reis, 1991). Other currently under-utilized but promising approaches to trust research include case study methods and ethnography. These approaches could extend insights into how events are interpreted and the development of norms that underpin individuals’ responses and their consequential impact on outcomes. In addition, these approaches may be significant in furthering study of the dark and blind side of trust. In conclusion, trust is a fragile commodity, but also one that is being relied upon increasingly within and outside organizations to enhance employees’ productivity, innovation, and citizenship. The future developments we have outlined would provide a more balanced and critical perspective on employee trust, identifying both the benefits and drawbacks. Although such insights may well enable organizations to achieve their objectives more effectively, they might also assist individual employees to understand and improve their work contexts, particularly their relationships with others.
REFERENCES Adams, J.S. (1963). Toward an understanding of inequity. Journal of Abnormal and Social Psychology, 67, 422–63. Allen, T., Freedman, D., Russell, J., Reizenstein, R., & Rentz, J. (2001). Survivor reactions to organizational downsizing: Does time ease the pain? Journal of Occupational and Organizational Psychology, 74, 145–64. Amabile, T.M. (1979). Effects of external evaluation on artistic creativity. Journal of Personality and Social Psychology, 37, 221–33. Andrews, K.M., & Delahaye, B.L. (2000). Influences on knowledge processes in organizational learning: The psychosocial filter. Journal of Management Studies, 37, 797–810.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
179
Appelbaum, E., Bailey, T., Berg, P., & Kallenberg, A. (2001). Do high performance work systems pay off? Research in Sociology of Work, 10, 85–107. Aquino, K., Tripp, T.M., & Bies, R.J. (2006). Getting even or moving on? Power, procedural justice, and types of offense as predictors of revenge, forgiveness, reconciliation, and avoidance in organizations. Journal of Applied Psychology, 91, 653–68. Argyris, C. (1952). The Impact of Budgets on People. New York. Argyris, C. (1957). The individual and organization: Some problems of mutual adjustment. Administrative Science Quarterly, 2, 1–24. Argyris, C. (1964). Integrating the Individual and the Organization. New York. Aryee, S., Budwhar, P.S., & Chen, Z.X. (2002). Trust as a mediator of the relationship between organizational justice and work outcomes: Test of a social exchange model. Journal of Organizational Behavior, 23, 267. Avolio, B.J., & Chan, A. (2008). The dawning of a new era for genuine leadership development. In G.P. Hodgkinson, & J.K. Ford (Eds), International Review of Industrial and Organizational Psychology (volume 23), (pp. 197–238) Chichester, UK: Wiley. Avolio, B.J., Gardner, W.L., Walumbwa, F.O., Luthans, F., & May, D.R. (2004). Unlocking the mask: A look at the process by which authentic leaders impact follower attitudes and behaviors. The Leadership Quarterly, 15, 801–23. Bandura, A. (1986). Social Foundations of Thought and Action. Englewood Cliffs, NJ: Prentice Hall. Banking Times (2010). RBS fined £5.6 million for sanctions breach, by Gill Montia. http://www.bankingtimes.co.uk/03082010-rbs-fined-5-6-million-forsanctions-breach/ (accessed 3 August 2010). Bass, B. (1985). Leadership and Performance Beyond Expectations. New York: Free Press. Bass, B. (1990). From transactional to transformational leadership: Learning to share the vision. Organizational Dynamics, 18, 19–31. Bass, B. (1997). Does the transactional-transformational leadership paradigm transcend organizational and national boundaries? American Psychologist, 52, 130–9. BBC (2010). RBS fined for breach of money laundering checks, by Russell Lynch. http://news.bbc.co.uk/1/hi/business/8594231.stm (accessed 4 August 2010). Bernerth, J.B., & Walker, H.J. (2009). Propensity to trust and the impact on social exchange: An empirical investigation. Journal of Leadership and Organizational Studies, 15(3), 217–26. Bies, R.J., & Moag, J.S. (1986). Interactional justice: Communication criteria of fairness. In R.J. Lewicki, B.H. Sheppard, & B.H. Bazerman (Eds), Research on Negotiations in Organizations. Greenwich CT. Bies, R.J., & Tripp, T. (1996). Beyond distrust: “Getting even” and the need for revenge. In R. Kramer, & T. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research. Thousand Oaks, CA: Sage. Bigley, G.A., & Pearce, J.L. (1998). Straining for shared meaning in organization science: Problems of trust and distrust. Academy of Management Review, 23, 405–21. Bijlsma, K., & Van de Bunt, G. (2003). Antecedents of trust in managers: A “bottom up” approach. Personnel Review, 32, 638–64. Bijlsma-Frankema, K., Sitkin, S., & Weibel, A. (2007). Reversing a vicious cycle of mutual distrust. Paper presented at the Academy of Management Annual Meeting. Black, A., & Deci, E. (2000). The effects of instructors’ autonomy support and students’ autonomous motivation on learning organic chemistry: A self-determination theory perspective. Science Education, 84, 740–56. Blau, P.M. (1963). The Dynamics of Bureaucracy: A Study of Interpersonal Relations in Two Government Agencies. Chicago, IL: University of Chicago Press. Bohnet, I., & Zeckhauser, R. (2004). Trust, risk and betrayal. Journal of Economic Behavior & Organization, 55(4), 467–84.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
180 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Boselie, P., Dietz, G., & Boon, C. (2005). Commonalities and contradictions in HRM and performance research. Human Resource Management Journal, 15, 67–94. Bottom, W., Gibson, K., Daniels, S., & Murnighan, J. (2002). When talk is not cheap: Substantive penance and expressions of intent in rebuilding cooperation. Organization Science, 13, 497–513. Bradner, E., & Mark, G. (2002). Why distance matters: Effects on cooperation, persuasion and deception. Proceedings of Computer Supported Cooperative Work, 226–35. Brann, P., & Foddy, M. (1987). Trust and the consumption of a deteriorating common resource. Journal of Conflict Resolution, 31, 615–30. Brockner, J., Siegel, P.A., Daly, J.P., Tyler, T., & Martin, C. (1997). When trust matters: The moderating effect of outcome favorability. Administrative Science Quarterly, 42, 558–83. Brown, J.S., & Duguid, P. (1991). Organizational learning and communities-ofpractice: Toward a unified view of working, learning, and innovation. Organization Science, 2, 40–57. Brown, M., Trevino, L., & Harrison, D. (2005). Ethical leadership: A social learning perspective for construct development and testing. Organizational Behavior and Human Decision Processes, 97, 117–34. Burt, R.S. (1992). Structural Holes: The Social Structure of Competition, Cambridge, MA: Harvard Business School Press. Burt, R.S., & Knez, M. (1995) Kinds of third-party effects on trust. Rationality and Society, 7, 255–92. Butler, J.K. (1991). Toward understanding and measuring conditions of trust: Evolution of a conditions of trust inventory. Journal of Management, 17(3), 643. Caldwell, R. (2003). The changing roles of personnel managers: Old ambiguities, new uncertainties. Journal of Management Studies, 40, 983–1004. Cardinal, L.B., & Sitkin, S.B. (In press). A configurational theory of control. In S.B. Sitkin, L.B. Cardinal, & K. Bijsma-Frankema (Eds), Control in Organizations: New Directions in Theory and Research. Cambridge, UK: Cambridge University Press. Carnevale, D.G. (1995). Trustworthy Government: Leadership and Management Strategies for Building Trust and High Performance. San Francisco, CA: Jossey-Bass. Challagalla, G.N., & Shervani, T.A. (1997). A measurement model of the dimensions and types of output and behavior control: An empirical test in a salesforce context. Journal of Business Research, 39, 159–72. Chiaburu, D., & Lim, A. (2008). Manager trustworthiness or interactional justice? Predicting organizational citizenship behaviors. Journal of Business Ethics, 83, 453–67. Child, J., & Mollering, G. (2003). Contextual confidence and active trust development in the Chinese business environment. Organization Science, 14, 69–80. Choudhury, E. (2008). Trust in administration: An integrative approach to optimal trust. Administration Society, 40(6), 586–620. Chughtai, A.A., & Buckley, F. (2007). The relationship between work engagement and foci of trust: A conceptual analysis. 13th Asia Pacific Management Conference, Melbourne, Australia 2007 (pp. 73–85). Clark, M.S., & Mills, J. (1993). The difference between communal and exchange relationships: What it is and is not. Personality and Social Psychology Bulletin, 19(6), 684–91. Cohen-Charash, Y., & Spector, P.E. (2001). The role of justice in organizations: A meta-analysis. Organizational Behavior and Human Decision Processes, 86, 278–321. Cohn, L., Macfarlane, S., Yanez, C., & Imai, W. (1995). Risk-perception: Differences between adolescents and adults. Health Psychology, 14(3), 217–22. Coleman, J.S. (1975). Social structure and a theory of action. In P.M. Blau (Ed.), Approaches to the Study of Social Structure. New York.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
181
Coletti, A.L., Sedatole, K.L., & Towry, K.L. (2005). The effect of control systems on trust and cooperation in collaborative environments. Accounting Review, 80, 477– 500. Collins, C., & Smith, K. (2006). Knowledge exchange and combination: The role of human resource practices in the performance of high-technology firms. Academy of Management Journal, 49, 544–60. Colquitt, J.A., Conlon, D.E., Wesson, M.J., Porter, C., & Ng, K.Y. (2001). Justice at the millennium: A meta-analytic review of 25 years of organizational justice research. Journal of Applied Psychology, 86, 425–45. Colquitt, J.A., Scott, B., & Lepine, J. (2007). Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 92, 909–27. Conway, N., & Briner, R.B. (2009). Fifty years of psychological contract research: What do we know and what are the main challenges? In G.P. Hodgkinson, & J.K. Ford (Eds), International Review of Industrial and Organizational Psychology (volume 21), (pp. 71–131). Chicester, UK: Wiley. Cook, J., & Wall, T. (1980). New work attitude measures of trust, organizational commitment and personal need non-fulfillment. Journal of Occupational Psychology, 53, 39–52. Cook, K.S., Yamagishi, T., Cheshire, C., Cooper, R., Matsuda, M., & Mashima, R. (2005). Trust building via risk taking: A cross-societal experiment. Social Psychology Quarterly, 68(2), 121–42. Costa, A.C., & Bijlsma-Frankema, K. (2007) Trust and control inter-relations: New perspectives on the trust-control nexus – towards integrated perspectives. Group and Organization Management, 32(4), 392–406. Costa, P.T., & McCrae, R.R. (1992). NEO PI-R (professional manual). Florida: Psychological Assessment Resources. Coyle-Shapiro, J. (2002). A psychological contract perspective on organizational citizenship behavior. Journal of Organizational Behavior, 23, 927–46. Cropanzano, R., Byrne, Z.S., Bobocel, D.R., & Rupp, D.E. (2001). Moral virtues, fairness heuristics, social entities, and other denizens of organizational justice. Journal of Vocational Behavior, 58, 164–209. Crozier, M. (1964). The Bureaucratic Phenomenon. Chicago, IL: University of Chicago Press. Csikszentmihalyi, M. (1991). Flow: The Psychology of Optimal Experience. New York, Harper Perennial. Cunningham, J.B., & MacGregor, J. (2000). Trust and the design of work complementary constructs in satisfaction and performance. Human Relations, 53(12), 1575–91. Currall, S.C., & Epstein, M.J. (2003). The fragility of organizational trust: Lessons from the rise and fall of Enron. Organizational Dynamics, 32, 193–206. Darley, J.M., & Fazio, R.H. (1980). Expectancy confirmation processes arising in the socal interaction sequence. American Psychologist, 35, 867–81. Das, T.K., & Teng, B. (1998). Between trust and control: Developing confidence in partner cooperation in alliances. Academy of Management Review, 23(3), 491–512. Davis, J., Schoorman, F., Mayer, R.C., & Tan, H. (2000). The trusted general manager and business unit performance: Empirical evidence of a competitive advantage. Strategic Management Journal, 21, 563–76. De Jong, B., & Dirks, K.T. (2010). Trust and control within teams: Joint implications of mean and dispersion levels. 5th EIASM Workshop on Trust Within and Between Organizations, Madrid. Deci, E.L., Connell, J.P., & Ryan, R.M. (1989). Self-determination in a work organization. Journal of Applied Psychology, 74, 580–90.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
182 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Deci, E.L., Koestner, R., & Ryan, R.M. (1999). The undermining effect is a reality after all – extrinsic rewards, task interest, and self-determination: Reply to Eisenberger, Pierce, and Cameron (1999) and Lepper, Henderlong, and Gingras (1999). Psychological Bulletin, 125, 692–700. Delery, J.E. (1998). Issues of fit in strategic human resource management: Implications for research. Human Resource Management Review, 8, 289–309. Den Hartog, D. (2003). Trusting others in organizations: Leaders, management and coworkers. In B. Nooteboom, & F. Six (Eds), The Trust Process. London: Edward Elgar. Den Hartog, D.N., & De Hoogh, A.H.B. (2009). Empowering behaviour and leader fairness and integrity: Studying perceptions of ethical leader behaviour from a levelsof-analysis perspective. European Journal of Work and Organizational Psychology, 18, 199–230. Den Hartog, D.N., House, R., Hanges, P., Ruiz-Quintanilla, S., & Dorfman, P. (1999). Culture specific and cross-culturally generalizable implicit leadership theories: Are attributes of charismatic/transformational leadership universally endorsed? Leadership The Quarterly, 10, 219–56. Deutsch, M. (1960). The effect of motivational orientation upon trust and suspicion. Human Relations, 13, 123–39. Deutsch-Salamon, S., & Robinson, S.L. (2008). Trust that binds: The impact of collective felt trust on organizational performance. Journal of Applied Psychology, 93, 593–601. Dietz, G., & Den Hartog, D.N. (2006). Measuring trust inside organisations. Personnel Review, 35(5), 557–88. Dirks, K.T. (2000). Trust in leadership and team performance: Evidence from NCAA basketball. Journal of Applied Psychology, 85, 1004–12. Dirks, K.T., & Ferrin, D.L. (2001). The role of trust in organizational settings. Organization Science, 12(4), 450–67. Dirks, K.T., & Ferrin, D.L. (2002). Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology, 87, 611–28. Dirks, K., Lewicki, R., & Zaheer, A. (2009). Repairing relationships within and between organizations: Building a conceptual foundation. Academy of Management Review, 34, 68–84. Dirks, K.T., & Skarlicki, D.P. (2009). The relationship between being perceived as trustworthy by coworkers and individual performance. Journal of Management, 35(1), 136–57. Dobbins, G., Platz, S., & Houston, J. (1993). Relationship between trust in appraisal and appraisal effectiveness: A field study. Journal of Business and Psychology, 7, 309–22. Doney, P., Cannon, J., & Mullen, M. (1998). Understanding the influence of national culture on the development of trust. Academy of Management Review, 23, 601–20. Driscoll, J. (1978). Trust and participation in organizational decision making as predictors of satisfaction. Academy of Management Journal, 21, 44–56. Dukerich, J., Kramer, R., & McLean Parks, J. (1998). The dark side of organizational identification. In D. Whetten, & P. Godfrey (Eds), Identity in Organizations: Building Theory Through Conversations. Thousand Oaks, CA: Sage. Dyer, J.H., & Nobeoka, K. (2000). Creating and managing a high performance knowledge-sharing network: The Toyota case. Strategic Management Journal, 21, 345–67. Earley, P.C. (1986). Trust, perceived importance of praise and criticism, and work performance: An examination of feedback in the United States and England. Journal of Management, 12, 457.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
183
Eberl, P. (2004). The development of trust and implications for organizational design: A game-and attribution-theoretical framework. Schmalenbach Business Review, 56, 258–73. Edelman (2009). Edelman Trust Barometer. http.//www.edelman.co.uk/fi les/ trustbarometer- 2009.pdf. Edmondson, A. (2004). Psychological safety, trust, and learning in organizations: A group-level lens. In R. Kramer & K. Cook (Eds), Trust and Distrust in Organizations: Dilemmas and Approaches (pp. 239–72). New York: Russell Sage. Edwards, R. (1979). Contested Terrain: The Transformation of the Workplace in the Twentieth Century. London, Heinemann. Eisenhardt, K.M. (1985). Control: Organizational and economic approaches. Management Science, 31, 134–49. Falk, A., & Kosfeld, M. (2006). The hidden costs of control. American Economic Review, 96, 1611–30. Ferrin, D.L., Bligh, M.C., & Kohles, J.C. (2007a). Can I trust you to trust me? A theory of trust, monitoring, and cooperation in interpersonal and intergroup relationships. Group & Organization Management, 32, 465–99. Ferrin, D.L., Bligh, M.C., & Kohles, J.C. (2008). It takes two to tango: An interdependence analysis of the spiraling of perceived trustworthiness and cooperation in interpersonal and intergroup relationships. Organizational Behavior and Human Decision Processes, 107(2), 161–78. Ferrin, D.L., Kim, P.H., Cooper, C.D., & Dirks, K.T. (2007b). Silence speaks volumes: The effectiveness of reticence in comparison to apology and denial for responding to integrity- and competence-based trust violations. Journal of Applied Psychology, 92, 893–908. Ferrin, D.L., & Gillespie, N. (2009). Cultural differences and universals in the development of trust. 22nd Annual International Association of Conflict Management Conference. Kyoto, Japan. Finkelstein, M., & Penner, L. (2004). Predicting organizational citizenship behavior: Integrating the functional and role identity approaches. Social Behavior and Personality, 32, 383–98. Fiske, A.P. (1992). The 4 elementary forms of sociality: Framework for a unified theory of social-relations. Psychological Review, 99(4), 689–723. Fleishman, E.A., & Harris, E.F. (1962). Patterns of leadership behavior related to employee grievances and turnover. Personnel Psychology, 15, 43–56. Folger, R., & Konovsky, M. (1989) Effects of procedural and distributive justice on reactions to pay raise decisions. Academy of Management Journal, 32, 115–30. Fox, A. (1974). Beyond Contract: Work, Power and Trust Relations. London: Faber and Faber. Frazier, M.L., Johnson, P.D., Gavin, M., Gooty, J., & Snow, D.B. (2010). Organizational justice, trustworthiness, and trust: A multifoci examination. Group & Organization Management, 35, 39–76. Frey, B.S. (1997) A constitution for knaves crowds out civic virtues. Economic Journal, 107, 1043–53. Frey, B.S., & Jegen, R. (2001). Motivation crowding theory: A survey of empirical evidence. Journal of Economic Surveys, 15, 589–611. Frost, J., Osterloh, M., & Weibel, A. (2010). Governing knowledge work: Transactional and transformational solutions. Organizational Dynamics, 39, 126–36. Gagn´e, M. (2003) The role of autonomy support and autonomy orientation in prosocial behavior engagement. Motivation and Emotion, 27, 199–223. Gagne, M., & Deci, E.L. (2005). Self-determination theory and work motivation. Journal of Organizational Behavior, 26, 331–62.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
184 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Gambetta, D. (1988). Can we trust trust? In D. Gambetta (Ed.), Trust: Making and Breaking Cooperative Relations (pp. 213–37). New York: Basil Blackwell. Gardner, W.L., Avolio, B.J., Luthans, F., May, D.R., & Walumbwa, F.O. (2005.). “Can you see the real me?” A selfbased model of authentic leader and follower development. The Leadership Quarterly, 16, 343–72. Gelfand, M., Erez, M., & Aycan, Z. (2007). Cross-cultural organizational behavior. Annual Review of Psychology, 58, 479–514. Gerbasi, A., Cook, K.S., Jody, C.W., & Dawn, T.R. (2008). The Effect of Perceived Trustworthiness on Affect in Negotiated and Reciprocal Exchange. Social Structure and Emotion. San Diego: Academic Press. Ghoshal, S., & Moran, P. (1996). Bad practice: A critique of the transaction cost theory. Academy of Management Review, 21, 13–47. Giddens, A. (1990). The Consequences of Modernity. Oxford. Gill, H., Boies, K., Finegan, J.E., & McNally, J. (2005). Antecedents of trust: Establishing a boundary condition for the relation between propensity to trust and intention to trust. Journal of Business and Psychology, 19, 287–302. Gillespie, N. (2003). Measuring trust in work relationships: The Behavioural Trust Inventory. Paper presented at the Academy of Management Meeting, Seattle. Gillespie, N., & Dietz, G. (2009). Trust repair after an organization-level failure. Academy of Management Review, 34, 127–45. Goffman, E. (1967). Interaction Rituals. Essays on Face-to-Face Behavior. Anchor Books. Gomez, C., & Rosen, B. (2001). The leader-member exchange as a link between managerial trust and employee empowerment. Group & Organization Management, 26, 53. Gould-Williams, J. (2003). The importance of HR practices and workplace trust in achieving superior performance: A study of public-sector organizations. International Journal of Human Resource Management, 14, 28–54. Govier, T. (1994). Is it a jungle out there: trust, distrust and the construction of socialreality. Dialogue, 33(2), 237–53. Grant, A., & Sumanth, J. (2009). Mission possible? The performance of prosocially motivated employees depends on manager trustworthiness. Journal of Applied Psychology, 94, 927–44. Grover, S.L., & Crooker, K.J. (1995). Who appreciates family-responsive humanresource policies: The impact of family-friendly policies on the organizational attachment of parents and non-parents. Personnel Psychology, 48, 271–88. Guest, D. (1998). Is the psychological contract worth taking seriously? Journal of Organizational Behavior, 19, 649–64. Guest, D. (2002). Human resource management, corporate performance and employee wellbeing: Building the worker into HRM. Journal of Industrial Relations, 44(3), 335–58. Guest, D., & Conway, N. (2002). Communicating the psychological contract: An employer perspective. Human Resource Management Journal, 12(2), 22–38. Gustafson, P. (1998). Gender differences in risk perception: Theoretical and methodological perspectives. Risk Analysis, 18(6), 805–11. Hardin, R. (1993). The street-level epistemology of trust. Politics & Society, 21(4), 505–29. Hardy, C., Phillips, N., & Lawrence, T. (1998). Distinguishing trust and power in interorganizational relations: Forms and facades of trust. In C. Lane, & R. Bachmann (Eds), Trust Within and Between Organizations. Oxford: Oxford University Press. Heider, F. (1958). The Psychology of Interpersonal Relations. New York, London: John Wiley: Chapman & Hall.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
185
Helliwell, J.F., & Wang, S. (2010). Trust and well-being. NBER Working Paper Series, 15911. Hofstede, G. (1980). Culture’s Consequences: International Differences in Work-related Values. Beverly Hills, CA: Sage. Hofstede, G. (2001). Culture’s Consequences: Comparing Values, Behaviors, Institutions, and Organizations across Nations. Sage Publications. Hofstede, G., & Bond, M.H. (1988). The Confucius connection: From cultural roots to economic growth. Organizational Dynamics, 16(4), 5–21. House, R., Hanges, P., Javidan, M., Dorfman, P., & Gupta, V. (2004). Culture, Leadership, and Organizations: The GLOBE Study of 62 Societies. Sage Publications. Hsu, M., Ju, T.L., Yen, C., & Chang, C. (2007). Knowledge sharing behavior in virtual communities: The relationship between trust, self-efficacy, and outcome expectations. International Journal of Human-Computer Studies, 65, 153–69. Huemer, L., von Krogh, G., & Roos, J. (1998). Knowledge and the concept of trust. In G. von Krogh, J. Roos, & D. Kleine (Eds), Knowing in Firms: Understanding, Managing and Measuring Knowledge (pp. 123–45). Sage. Huff, L., & Kelley, L. (2003). Levels of organizational trust in individualist versus collectivist societies: A seven-nation study. Organization Science, 14, 81–90. Huselid, M.A. (1995). The impact of human-resource management-practices on turnover, productivity, and corporate financial performance. Academy of Management Journal, 38, 635–72. Husted, K., & Michailova, S. (2002). Diagnosing and fighting knowledge-sharing hostility. Organizational Dynamics, 31, 60–73. Iles, P., Mabey, C., & Robertson, I. (1990). HRM practices and employee commitment: Possibilities, pitfalls and paradoxes. British Journal of Management, 1, 147–57. James, H.S. (2002). The trust paradox: A survey of economic inquiries into the nature of trust and trustworthiness. Journal of Economic Behavior & Organization, 47(3), 291–307. Johnson, J., & Cullen, J. (2002). Trust in cross-cultural relationships. In M.J. Gammon, & K. Newman (Eds), The Blackwell Handbook of Cross-Cultural Management (pp. 335–60). Oxford, UK: Wiley-Blackwell. Judge, T., Piccolo, R., & Ilies, R. (2004). The forgotten ones? The validity of consideration and initiating structure in leadership research. Journal of Applied Psychology, 89, 36–50. Kalshoven, K., & Den Hartog, D.N. (2009). Ethical leader behavior and leader effectiveness: The role of prototypicality and trust. International Journal of Leadership Studies, 3, 68–83. Kalshoven, K., De Hoogh, A.H.B., & Den Hartog, D.N. (In press). Ethical leadership at work questionnaire (ELW): Development and validation of a multidimensional measure. The Leadership Quarterly. Kee, H.W., & Knox, R.E. (1970). Conceptual and methodological considerations in the study of trust and suspicion. Journal of Conflict Resolution, 14, 357–66. Kelley, H.H. (1973). The processes of causal attribution. American Psychologist, 28(2), 107–28. Kernan, M.C., & Hanges, P.J. (2002). Survivor reactions to reorganization: Antecedents and consequences of procedural, interpersonal, and informational justice. Journal of Applied Psychology, 87, 916–28. Kim, P.H., Dirks, K.T., Cooper, C.D., & Ferrin, D.L. (2006). When more blame is better than less: The implications of internal vs. external attributions for the repair of trust after a competence- vs. integrity-based trust violation. Organizational Behavior and Human Decision Processes, 99, 49–65.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
186 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Kim, P.H., Ferrin, D.L., Cooper, C.D., & Dirks, K.T. (2004). Removing the shadow of suspicion: The effects of apology versus denial for repairing competence- versus integrity-based trust violations. Journal of Applied Psychology, 89, 104–18. Kim, P., Dirks, K., & Cooper, C. (2009). The repair of trust: A dynamic bilateral perspective and multilevel conceptualization. Academy of Management Review, 34, 401–22. Kim, W., & Mauborgne, R. (2003). Fair process: Managing in the knowledge economy. Harvard Business Review, 81, 127–36. Kirkman, B., Lowe, K., & Gibson, C. (2006). A quarter century of Cultures Consequences: A review of empirical research incorporating Hofstedes cultural values framework. Journal of International Business Studies, 37, 285–320. Kirsch, L.J. (2004). Deploying common systems globally: The dynamics of control. Information Systems Research, 15(4), 374–95. Konovsky, M.A., & Cropanzano, R. (1991). Perceived fairness of employee drugtesting as a predictor of employee attitudes and job-performance. Journal of Applied Psychology, 76, 698–707. Konovsky, M.A., & Pugh, D.S. (1994). Citizenship behavior and social exchange. Academy of Management Journal, 37, 656–69. Korsgaard, M.A., Brodt, S.E., & Whitener, E.M. (2002). Trust in the face of conflict: The role of managerial trustworthy behavior and organizational context. Journal of Applied Psychology, 87, 312–9. Kosfeld, M., Heinrichs, M., Zak, P., Fischbacher, U., & Fehr, E. (2005). Oxytocin increases trust in humans. Nature, 435, 673–6. Kramer, R.M. (2006). Trust as situated cognition: An ecological perspective on trust decisions. In R. Bachmann & A. Zaheer (Eds), Handbook of Trust Research (pp. 68–84). Chichester, UK: Edward Elgar. Kramer, R.M. (1996). Divergent realities and convergent disappointments in the hierarchic relation: Trust and the intuitive auditor at work. In R. Kramer, & T. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research (pp. 216–45). Thousand Oaks, CA: Sage. Kramer, R.M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50, 569–98. Leana, C.R. (1986). Predictors and consequences of delegation. Academy of Management Journal, 29, 754–74. Lee, C., Pillutla, M., & Law, K. (2000). Power-distance, gender and organizational justice. Journal of Management, 26, 685. Lester, S. & Brower, H. (2003). In the eyes of the beholder: The relationship between subordinates’ felt trustworthiness and their work attitudes and behaviors. Journal of Leadership & Organizational Studies, 10, 17–33. Lewicki, R.J., & Bunker, B.B. (1995). Trust in relationships: A model of trust development and decline. In B.B. Bunker, & J.Z. Rubin (Eds), Conflict, Cooperation, and Justice. San Francisco, CA: Jossey Bass. Lewicki, R.J., & Bunker, B. (1996). Developing and maintaining trust in working relationships. In R. Kramer, & T. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research. Thousand Oaks, CA: Sage. Lewicki, R.J., Wiethoff, C., & Tomlinson, E.C. (2005). What is the role of trust in organizational justice? In J. Greenberg, & J.A. Colquitt (Eds), Handbook of Organizational Justice. Mahwah, NJ: Erlbaum. Lewicki, R.J., McAllister, D.J., & Bies, R.J. (1998). Trust and distrust: New relationships and realities. Academy of Management Review, 23, 438–58. Lewicki, R.J., Tomlinson, E.C., & Gillespie, N. (2006). Models of interpersonal trust development: Theoretical approaches, empirical evidence, and future directions. Journal of Management, 32(6), 991–1022.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
187
Lewis, J.D., & Weigert, A. (1985). Trust as a social reality. Social Forces, 63, 967–85. Li, P. (2010). Intercultural trust and trust building: The contexts and strategies of adaptive learning in acculturation. FINT-EIASM Trust Workshop, Madrid, Spain. Lindsley, D.H., Brass, D.J., & Thomas, J.B. (1995). Efficacy-performance spirals: A multilevel perspective. Academy of Management Review, 20, 645–78. Long, C., & Sitkin, S. (2006). Trust in the balance: How managers integrate trustbuilding and task control. In R. Bachmann, & A. Zaheer (Eds), Handbook of Trust Research (p. 87). Chichester, UK: Edward Elgar. Luhmann, N. (1988). Familiarity, confidence, trust: Problems and alternatives. In D. Gambetta (Ed.), Trust: Making and Breaking Cooperative Relations (pp. 94–107). New York: Basil Blackwell. Mahon, J.F. (2002). Corporate reputation: Research agenda using strategy and stakeholder literature. Business & Society, 41(4), 415. March, J.G., & Olsen, J.P. (1975). Uncertainty of past: Organizational learning under ambiguity. European Journal of Political Research, 3, 147–71. Martinko, M., Douglas, S., & Harvey, P. (2006). Attribution theory in industrial and organizational psychology: A review. International Review of Industrial and Organizational Psychology, 21, 127. Masuch, M. (1985) Vicious circles in organizations. Administrative Science Quarterly, 30, 14–33. Mayer, R.C., & Davis, J.H. (1999). The effect of the performance appraisal system on trust for management: A field quasi-experiment. Journal of Applied Psychology, 84, 123–36. Mayer, R.C., Davis, J.H., & Schoorman, F.D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–34. Mayer, R.C., & Gavin, M. (2005). Trust in management and performance: Who minds the shop while employees watch the boss? Academy of Management Journal, 48, 874–88. McAllister, D.J. (1995) Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. McDermott, R. (1999). Learning across teams. Knowledge Management Review, 8, 32–6. McEvily, B., Perrone, V., & Zaheer, A. (2003). Trust as an organizing principle. Organization Science, INFORMS. Institute for Operations Research, 14, 91–103. McEvily, B., & Tortoriello, M. (2005). Measuring trust in organizational research: Review and recommendations. Journal of Trust Research, 32, 991–1022. McGregor, D. (1960). The Human Side of Enterprise. New York, McGraw Hill. McKnight, D.H., Cummings, L.L., & Chervany, N.L. (1998). Initial trust formation in new organizational relationships. Academy of Management Review, 23, 473–91. McNeish, J. & Mann, I.J.S. (2010). Knowledge sharing and trust in organizations. IUP Journal of Knowledge Management, 8, 18–38. Merton, R.K. (2010). The self-fulfilling prophecy. Antioch Review, 68, 173–90. Meyer, J.P., Bobocel, D.R., & Allen, N.J. (1991). Development of organizational commitment during the first year of employment: A longitudinal-study of preentry and postentry influences. Journal of Management, 17, 717–33. Meyer, J.P., & Herscovitch, L. (2001). Commitment in the workplace: Toward a general model. Human Resource Management Review, 11(3), 299–326. Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital and the organizational advantage. Academy of Management Review, 23, 242–66. Nakayachi, K., & Watabe, M. (2005). Restoring trustworthiness after adverse events: The signaling effects of voluntary “Hostage Posting” on trust. Organizational Behavior and Human Decision Processes, 97, 1–17. Newell, S., & Swan, J. (2000). Trust and inter-organizational networking. Human Relations, 53, 1287.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
188 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Nooteboom, B., & Six, F. (2003). The Trust Process in Organizations. Edward Elgar. O’Reilly II, C.A., Chatman, J., & Caldwell, D.F. (1991). People and organizational culture: A profile comparison approach to assessing person-organization fit. Academy of Management Journal, 34, 487–516. Osterloh, M., & Frey, B.S. (2000). Motivation, knowledge transfer, and organizational forms. Organization Science, 11, 538–50. Parks, C., Henager, R., & Scamahorn, S. (1996). Trust and reactions to messages of intent in social dilemmas. Journal of Conflict Resolution, 40, 134–51. Pearce, J.L., Branyiczki, I., & Bigley, G.A. (2000). Insufficient bureaucracy: Trust and commitment in particularistic organizations. Organization Science, 11, 148–62. Perrone, V., Zaheer, A., & McEvily, B. (2003). Free to be trusted? Organizational constraints on trust in boundary spanners. Organization Science, 14(4), 422–39. Perry-Smith, J.E., & Blum, T.C. (2000). Work-family human resource bundles and perceived organizational performance. Academy of Management Journal, 43, 1107–17. Pillai, R., Schriesheim, C., & Williams, E. (1999). Fairness perceptions and trust as mediators for transformational and transactional leadership. Journal of Management, 25, 897–933. Podsakoff, P.M., Mackenzie, S.B., Moorman, R.H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers’ trust in leader, satisfaction, and organizational citizenship behaviors. The Leadership Quarterly, 1, 107–42. Podsakoff, P.M., Mackenzie, S.B., Paine, J.B., & Bachrach, D.G. (2000). Organizational citizenship behaviors: A criticial review of the theoretical and empirical literature and suggestions for future research. Journal of Management, 26, 513–63. Reb, J., Goldman, B.M., Kray, L.J., & Cropanzano, R. (2006). Different wrongs, different remedies? Reactions to organizational remedies after procedural and interactional injustice. Personnel Psychology, 59(1), 31–64. Ren, H., & Gray, B. (2009). Repairing relationship conflict: How violation types and culture influence the effectiveness of restoration rituals. Academy of Management Review, 34, 105–26. Rhee, M., & Valdez, M.E. (2009). Contextual factors surrounding reputation damage with potential implications for reputation repair. Academy of Management Review, 34(1), 146–68. Rigby, C.S, Deci, E., Patrick, B., & Ryan, R. (1992). Beyond the intrinsic–extrinsic dichotomy: Self-determination in motivation and learning. Motivation and Emotion, 16, 165–85. Robinson, S. (1996). Trust and breach of the psychological contract. Administrative Science Quarterly, 41, 574–99. Robinson, S., & Rousseau, D. (1994). Violating the psychological contract: Not the exception but the norm. Journal of Organizational Behavior, 15, 245–59. Rotter, J.B. (1967). A new scale for the measurment of interpersonal trust. Journal of Personality, 35, 651–65. Rotter, J.B. (1980). Interpersonal trust, trustworthiness, and gullibilty. American Psychologist, 35, 1–7. Rousseau, D.M. (1989). Psychological and implied contracts in organizations. Employee Responsibilities and Rights Journal, 2(2), 121–39. Rousseau, D.M. (1995). Psychological Contracts in Organizations: Understanding Written and Unwritten Agreements. Thousand Oaks, CA: Sage Publications. Rousseau, D.M., & McLean Parks, J. (1993). The contracts of individuals and organizations. Research in Organizational Behavior, 15, 1–13. Rousseau, D.M., Sitkin, S.B., Burt, R.S., & Camerer, C.F. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
189
Rupp, D.E., & Cropanzano, R. (2002). The mediating effects of social exchange relationships in predicting workplace outcomes from multifoci organizational justice. Organizational Behavior and Human Decision Processes, 89, 925–46. Ryan, R.M., & Deci, E.L. (2000). Self-determination theory and the facilitating of intrinsic motivation. American Psychologist, 55, 68–78. Sako, M. (1992). Prices, Quality and Trust: Inter-firm Relations in Britain and Japan. Cambridge: Cambridge University Press. Saunders, M.N.K., & Thornhill, A. (2003). Organisational justice, trust and the management of change: An exploration. Personnel Review, 32, 360–75. Schoorman, F., Mayer, R.C., & Davis, J. (2007). An integrative model of organizational trust: Past, present, and future. Academy of Management Review, 32(2), 344–54. ¨ Searle, R. (2010). Diary methods in Trust research. In F. Lyon, G. Mollering, M. Sanders & T. Hatzakis (Eds), Handbook of Research Methods on Trust. Cheltenham: Edward Elgar. Searle, R., & Billsberry, J. (In press). The development and destruction of organizational trust during recruitment and selection. In R. Searle, & D. Skinner (Eds), Trust and HRM. Edward Elgar. Searle, R., & Skinner, D. (Eds). (In press). Trust and HRM. Chichester: Edward Elgar. Searle, R., Den Hartog, D.N., Weibel, A., Gillespie, N., Six, F., Hatzakis, T., & Skinner, D. (In press). Trust in the employer: The role of high involvement work practices and procedural justice in European organizations. International Journal of Human Resource Management. Serva, M.A., Fuller, M.A., & Mayer, R.C. (2005). The reciprocal nature of trust: A longitudinal study of interacting teams. Journal of Organizational Behavior, 26, 625–48. Shapiro, S.P. (1987). The social control of impersonal trust. American Journal of Sociology, 93, 623–58. Sheppard, B.H., & Sherman, D.M. (1998). The grammars of trust: A model and general implications. Academy of Management Review, 23(3), 422–37. Shore, L., Bommer, W., Rao, A., & Seo, J. (2009). Social and economic exchange in the employee–organization relationship: the moderating role of reciprocation wariness. Journal of Managerial Psychology, 24(8), 701–21. Sievers, B. (2003). Fool’d with hope, men favour the deceit,’ or, can we trust in trust? In R. Westwood, & St. Clegg (Eds), Debating Organization: Point-Counterpoint in Organization Studies (pp. 356–367). Oxford: Blackwell. Simons, T. (2002). Behavioral integrity: The perceived alignment between managers’ words and deeds as a research focus. Organization Science, 13(1), 18–35. Sitkin, S.B. (1995). On the positive effect of legalization on trust. Research on Negotiation in Organizations, 5, 185–217. Sitkin, S.B., Cardinal, L., & Bijlsma-Frankema, K. (2009). Control in Organizations: New Directions in Theory and Research. Cambridge University Press: Cambridge, UK. Sitkin, S.B., & Roth, N.L. (1993). Explaining the limited effectiveness of legalistic “remedies” for trust/distrust. Organization Science, 4(3), 367–92. Sitkin, S.B., & Stickel, D. (1996). The road to hell: The dynamics of distrust in an era of quality. In R.M. Kramer & T.R. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research (pp. 196–215). Thousand Oaks, CA: Sage. Skinner, D., Saunders, M.N.K, & Duckett, H. (2004). Policies, promises and trust: Improving working lives in the National Health Service. International Journal of Public Sector Management, 17, 558–70. Skinner, D., Weibel, A., & Dietz, G. (2007). The gift of trust a poisoned chalice: 6 critical scenarios. Paper presented at the 4th EAISM Workshop on Trust Within and Between the Organisation, Amsterdam, The Netherlands.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
190 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Snow, C.C., Fjeldstad, Ø.D., Lettl, C., & Miles, R.E. (2010). Introduction to the special issue. Organizational Dynamics, 39(2), 91–2. Song, L.J., Tsui, A.S., & Law, K.S. (2009). Unpacking employee responses to organizational exchange mechanisms: The role of social and economic exchange perceptions. Journal of Management, 35(1), 56–93. Spector, P.E., Cooper, C.L., & Sparks, K. (2001). An international study of the psychometric properties of the hofstede values survey module 1994: A comparison of individual and country/province level results. Applied Psychology An International Review, 50, 269–81. Spreitzer, G., & Mishra, A.K. (1997). Survivor Responses to Downsizing: The Mitigating Effects of Trust and Empowerment. Southern California Studies Center. University of Southern California. Spreitzer, G., & Mishra, A.K. (1999). Giving up control without losing control. Group and Organisation Management, 24(2), 155–87. Stack, L. (1978). Trust. In H. London, & J. Exner (Eds), Dimensions of Personality. New York, NY: John Wiley & Sons. Stogdill, R. (1950). Leadership, membership and organization. Psychological Bulletin, 47, 1–14. Strickland, L. (1958). Surveillance and trust. Journal of Personality, 26(2), 200–15. Sweeney, P.J. (2010). Do soldiers re-evaluate their trust in their leaders prior to combat operations? Military Psychology, 22, 70–88. Tan, H.H., & Lim, A.K.H. (2009). Trust in coworkers and trust in organizations. Journal of Psychology, 143, 45–66. Tan, H.H., & Tan, C.S.F. (2000). Toward the differentiation of trust in supervisor and trust in organization. Genetic Social and General Psychology Monographs, 126, 241–60. The Guardian (2009). RBS ex-chief enjoying £650,000 pension, by Jill Treanor and Patrick Wintour. http://www.guardian.co.uk/business/2009/feb/26/royal-bankof-scotland (accessed 26 February 2009). Thompson, P.J., & Warhurst, C. (1998). Workplaces of the Future. Basingstoke: Macmillan Business. Tomlinson, E.C., & Mayer, R.C. (2009). The role of causal attribution dimensions in trust repair. Academy of Management Review, 34, 85–104. Tripp, T.M., Bies, R.J., & Aquino, K. (2002). Poetic justice or petty jealousy? The aesthetics of revenge. Organizational Behavior and Human Decision Processes, 89, 966–84. Tripp, T., Bies, R.J. & Aquino, K. (2007). A vigilante model of justice: Revenge, reconciliation, forgiveness, and avoidance. Social Justice Research, 20, 10–34. Tschannen-Moran, M., & Hoy, W.K. (2000). A multidisciplinary analysis of the nature, meaning, and measurement of trust. Review of Educational Research, 70(4), 547–93. Tyler, T.R., & Blader, S.L. (2000). Cooperation in Groups: Procedural Justice, Social Identity, and Behavioral Engagement. Philadelphia: Psychology Press. Tzafrir, S.S. (2005). The relationship between trust, HRM practices and firm performance. International Journal of Human Resource Management, 16(9), 1600–22. Vandenberg, R.J., Richardson, H.A., & Eastman, L.J. (1999). The impact of high involvement work processes on organizational effectiveness: A second-order latent variable approach. Group & Organization Management, 24, 300–39. Von Krogh, G. (1998). Care in knowledge creation. California Management Review, 40, 133–53. Von Krogh, G. (2002). The communal resource and information systems. Journal of Strategic Information Systems, 11, 85–107. Walumbwa, F.O., Avolio, B.J., Gardner, W.L., Wernsing, T.S., & Peterson, S.J. (2008). Authentic leadership: Development and validation of a theory-based measure. Journal of Management, 34, 89–126.
P1: OTA/XYZ P2: ABC c05 JWST046-Hodgkinson
January 21, 2011
12:59
Printer Name: Yet to Come
EMPLOYEE TRUST IN ORGANIZATIONAL CONTEXTS
191
Wayne, S.J., Shore, L.M., & Liden, R.C. (1997). Perceived organizational support and leader-member exchange: A social exchange perspective. Academy of Management Journal, 40, 82—111. Weaver, G. (1999). Compliance and values oriented ethics programs: Influences on employees’ attitudes and behavior. Business Ethics Quarterly, 9, 315–35. Weibel, A. (2007). Formal control and trustworthiness: Shall the twain never meet? Group Organization Management, 32(4), 500–17. Weibel, A., Rost, K., & Osterloh, M. (2010). Pay for performance in the public sector: Benefits and (hidden) costs. Journal of Public Administration Research and Theory, 20(2), 387–412. Weiner, B. (1986). An Attributional Model of Motivation and Emotion. New York: Springer-Verlag. Wenger, E. (2000). Communities of practice and social learning systems. Organization, 7, 225–46. Wheeler, L., & Reis, H.T. (1991). Self-recording of everyday life events: Origins, types, and uses. Journal of Personality, 59(3), 339–54. Whitener, E.M. (1997). The impact of human resources activities on employees trust. Human Resource Management Review, 7, 389–404. Whitener, E.M. (2001). Do “high commitment” human resource practices affect employee commitment? A cross-level analysis using hierarchical linear modeling. Journal of Management, 27(5), 515–35. Whitener, E.M., Brodt, S.E., Korsgaard, M.A., & Werner, J.M. (1998). Managers as initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23, 513–30. Whitley, R. (1987). Taking firms seriously as economic actors: Towards a sociology of firm behaviour. Organization Studies, 8(2), 125–47. Wright, P.M., & Nishii, L.H. (In press). Strategic HRM and organizational behavior: Integrating multiple levels of analysis. In D.E. Guest, J. Paauwe, & P.M. Wright (Eds), Human Resource Management and Performance: Progress and Prospects. Oxford, UK: Blackwell Publishing. Yakovleva, M., Reilly, R.R., & Werko, R. (2010). Why do we trust? Moving beyond individual to dyadic perceptions. Journal of Applied Psychology, 95, 79–91. Yamagishi, T. (2003). Cross-societal experimentation on trust: A comparison of the United States and Japan. Trust and Reciprocity, 6, 352–70. Yang, J.X., & Mossholder, K.W. (2010) Examining the effects of trust in leaders: A bases and foci approach. The Leadership Quarterly, 21, 50–63. Zaheer, A., McEvily, B., & Perrone, V. (1998). Does trust matter? Exploring the effects of interogranizational and interpersonal trust on performance. Organization Science, 9, 141–59. Zand, D.E. (1972). Trust and managerial problem solving. Adminstrative Science Quarterly, 17, 229–39. Zucker, L.G. (1986). Production of trust: Institutional sources of economic structure, 1840–1920. Research in Organizational Behavior, 8, 53–111.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
Chapter 6 THE PHYSICAL ENVIRONMENT OF THE OFFICE: CONTEMPORARY AND EMERGING ISSUES Matthew C. Davis, Desmond J. Leach, and Chris W. Clegg Leeds University Business School, University of Leeds, Leeds, LS2 9JT, UK INTRODUCTION An organization’s workspace, the physical environment an organization provides for its employees to carry out their work activities, constitutes the second largest financial overhead (after human resources) for most organizations (McCoy, 2005). Of the workspace provided, most employees in developed countries work in some form of office environment (Duffy, 1997) and studies of this practice have found that it has a powerful role in shaping a diverse range of psychological and behavioral outcomes, including individual work motivation (e.g., Oldham & Brass, 1979), job satisfaction (e.g., Veitch, Charles, Farley, et al., 2007), and patterns of interactions (e.g., Boyce, 1974; Ives & Ferdinands, 1974; Sundstrom & Sundstrom, 1986). Furthermore, the impact of offices upon their occupants’ personal productivity has been estimated to be somewhere in the region of 20% (e.g., Leaman & Bordass, 2005). Within the organizational literature, offices have been typically described as either traditional (sometimes referred to as enclosed or cellular offices) or open-plan. Traditional offices tend to house one or two individuals in private rooms, enclosed by walls, often containing most of the amenities required for their job (Danielsson & Bodin, 2008). Open-plan offices are characterized by a lack of interior walls, tend to be larger and contain greater numbers of workers, with individual workstations arranged within the office in groups (Brennan, Chugh, & Kline, 2002; Brookes & Kaplan, 1972). Workspace design, however, is currently under organizational scrutiny due to the changing nature of work. It is evident that many organizations are re-evaluating their facilities to ensure
International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
194 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY their workspace meets the needs of an increasingly diverse and demanding workforce (see e.g., Laing, 2006). Architects have noted a definite shift in terms of how employees, especially knowledge-based workers, spend their time, the kinds of task they engage in and, crucially, where they choose to work (Duffy, 2000). As Gillen commented: “Work environments are in a state of transition from something familiar and predictable to something not yet defined, multilocational, virtual and physical” (Gillen, 2006: 62). In response, organizations are increasingly investing in innovative offices, upgrading the open-plan office to support more nomadic, group-based, flexible, or remote working styles. However, office redesign is often based upon managers’ own interpretations and experiences of employee work patterns, largely without specific research or professional input (e.g., Laing, 2006). Optimizing existing offices (embarking on office redesign) manifestly involves change for the individual workers concerned. Alterations to factors such as the physical layout or configuration of space, and the provision of office facilities and services, can have significant effects on how individuals or teams go about their work (e.g., Laing, Duffy, Jaunzens, et al., 1998). However, despite the extensive change management literature (e.g., By, 2005; Kanter, Stein, & Jick, 1992; Luecke, 2003; Pettigrew, Woodman, & Cameron, 2001; Weick, 1979), there is currently limited guidance on how the process of office design and implementation can be successfully managed. Developing an appreciation of managing such processes is important if we wish to avoid new offices, or the changes in working practices that they necessitate and/or foster, being rejected by disaffected workers or undermined by counterproductive work behaviors (e.g., Chapman, Sheehy, Heywood, et al., 1995; Vischer, 2005). To help drive a fresh approach to the study of workplaces, and to aid managers’ decision making, this chapter collates and synthesizes, from a disparate range of sources, the findings of research that has investigated workers’ reactions to, and interactions with, their workspace. Given the prevalence of open-plan offices, we first appraise the value of such work environments and describe outcome-related contingencies. In so doing, we differ from previous reviews that have bounded or compartmentalized the literature by physical feature or design choice, for example, they have examined the effects of the density of a workspace separately from the openness of an office’s design (e.g., Baron, 1994; Elsbach & Pratt, 2007; Oldham, Cummings, & Zhou, 1995; Sundstrom & Sundstrom, 1986). Second, we review ways in which open-plan offices are evolving to suit the modern organization and what the implications might be for individuals and organizations. Third, we discuss the need to manage the process of change that office design and optimization involves. We examine some of the approaches that have been applied to date and also reflect on the similarity to wider organizational change principles. The chapter concludes by identifying how industrial and organizational (I/O) psychology research can contribute to optimal office design by extending current theory and utilizing fresh methodological techniques.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
195
RISE OF THE OPEN-PLAN OFFICE The office has emerged as the stereotypical place of work for the post-industrial age (e.g., Becker, 1981), with over 70% of workers occupying a form of openplan office at the turn of the century (e.g., Brill, Weidemann, & BOSTI Associates, 2001; Vischer, 1996). In this section we review the benefits and risks of open-plan working, consider the trade-offs involved in pursuing an open-plan strategy, and highlight individual and contextual factors affecting open-plan outcomes. In order to set the scene and provide appropriate context, we begin by revisiting the origins of research into the physical work environment and chart the rise of the open-plan office. Historical Overview The physical environment was a major topic of interest for early I/O psychologists (circa 1910 onwards), with attention focusing predominantly on the effects of ambient conditions (e.g., lighting, temperature, ventilation) on workers’ productivity (e.g., Morgan, 1916; Vernon, 1919, 1921). This approach is still reflected in the more recent ergonomic and environmental psychology literatures (Baron, 1994; Becker, 1981; Brennan, Chugh, & Kline, 2002; Oldham, Cummings, & Zhou, 1995; Sundstrom & Sundstrom, 1986). Notable relationships were established, for example between excessive noise and workers’ health and productivity (Baron, 1994). However, the publishing of the Hawthorne experiments (Roethlisberger & Dickson, 1939) marked a watershed in organizational research, with this long-running field study publicly failing to establish a link between changes to the physical environment and worker productivity. The lack of success in establishing environment–behavior links in the Hawthorne experiments coincided with a general decline in interest in the physical environment which would last until the 1960s (Oldham, Cummings, & Zhou, 1995). I/O psychologists may have conducted little research into the physical environment during the 1940s–1960s, but the topic was not wholly neglected and pockets of research activity by other disciplines did prevail. For example, social psychologists and architectural schools were researching the interaction of individuals with the built environment (albeit with limited attention to workplaces), demonstrating how the manipulation of the physical environment could produce profound differences in the way that people interact with one another. For example, the spatial configuration of furniture was found to influence the amount and nature of conversation between individuals (Osmond, 1959; Sommer, 1959), and the location of people within a building helped determine with whom they interacted and formed friendships (Festinger, Schachter, & Back, 1950). The widespread introduction of open-plan and b¨urolandschaft (landscaped) offices in North America in the 1960s and 1970s (e.g., Brookes & Kaplan,
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
196 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY 1972; Hundert & Greenfield, 1969; Zeitlin, 1969), saw I/O psychologists and organizational scholars begin once again to become interested in the relationship between workers and their physical workspace (for an excellent review of the development of office environments see Duffy, 1997). The effects that changes to established office design may have upon office occupants became a common concern and the issue was taken up by journalists (e.g., Business Week, 1978) and scholarly researchers (Brookes & Kaplan, 1972; Oldham & Brass, 1979). Proponents of the open-office predicted that it would produce better inter- and intra-team communication (Brookes & Kaplan, 1972; Lee & Brand, 2005; Pile, 1976). Such claims helped persuade scores of corporations to experiment with the demolition of interior office walls and so began the rapid rise of open-plan offices. The open-plan design soon became a vehicle for organizations to reduce their fixed overheads (e.g., Duffy, 1997; Vischer, 2005) and to increase the density of employees housed in previously enclosed spaces. Gradually, design features, such as the inclusion of plants and angled desk placements, were marginalized. At the same time, distances between neighboring desks were reduced and circulation space sacrificed for “efficiency” gains (Laing, 2006). In turn, concern over effectiveness triggered a new wave of research into the effects of introducing open-plan working (Brennan, Chugh, & Kline, 2002; Oldham, Cummings, & Zhou, 1995). These concerns are still influential within I/O psychology and management research, with a continuing emphasis upon the examination of key aspects of open-plan configuration, for example the density of workers housed within the office, the proximity of co-workers to one another, and the openness of the office (e.g., De Croon, Sluiter, Kuijer, et al., 2005).
Benefits of Open-Plan Offices The open-plan office has become the dominant choice when considering workspace strategies (e.g., Brill, Weidemann, & BOSTI Associates, 2001; Vischer, 1996), primarily for economic reasons (Brookes & Kaplan, 1972; Duffy, 1997; Laing, 2006). Fewer interior walls (and enclosed offices) permit larger floor plans to be achieved, which allow greater numbers of employees to be accommodated (e.g., Marquardt, Veitch, & Charles, 2002; Vischer, 2005). Increasing the density of workers housed within an office space through openplan configurations has consequently become an important method through which organizations attempt to reduce overheads (e.g., Duffy, 2000; Veitch, Charles, Farley, et al., 2007; Vischer, 2005). Higher office densities allow substantial savings to be made in either rental, land, or build costs, with lower services (e.g., heating and ventilation) and security charges (e.g., Duffy, 2000; Zeitlin, 1969). Reflecting these savings, the latest figures show a 40% increase in average UK office density since 1997 (from 16.6 m2 per person to 11.8 m2 today; British Council of Offices, 2009).
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
197
Cost savings can also be realized through an increase in flexibility. It is far easier to move furniture around in a large open-plan office than within enclosed offices. This flexibility reduces the costs of future reorganizations, with desks readily reorganized as individual and organizational requirements change, for example as project teams change or new technology is required. Individuals and teams can also be organized around work-flows and departmental groupings, enabling rationalizations such as the centralized storage of group files and work materials (e.g., Foland, Rowlen, & Watson, 1995). In addition to financial benefits, another driver of the rapid adoption of open-plan offices has been the proposition that they aid inter- and intra-team communication (Brookes & Kaplan, 1972). For example, advocates of the social relations approach have proposed that the physical environment is able to affect the frequency and nature of the interactions and communication that its inhabitants conduct (Festinger, Schachter, & Back, 1950; Oldham & Brass, 1979; Zalesny & Farace, 1987). It has been suggested that offices that facilitate greater communication and interaction (e.g., those that place individuals close to one another and remove physical barriers to communication, as open-plan offices frequently do) allow individuals to share task-relevant information, promote feedback, and create friendship opportunities (Oldham & Brass, 1979), leading in turn to increased inter-personal relations, reduced conflict, increased job satisfaction and motivation (Zalesny & Farace, 1987). Indeed, studies have found that more open workspace generates greater group sociability (e.g., Brookes & Kaplan, 1972) and an increase in interaction has been typically observed (e.g., Boyce, 1974; Hundert & Greenfield, 1969; Ives & Ferdinands, 1974; Sundstrom & Sundstrom, 1986). Furthermore, open-plan configurations have been found to affect the pattern of interaction, with less time spent in formal meetings and an increase in informal communication (e.g., more conversations held around desks) observed following its introduction (Brennan, Chugh, & Kline, 2002). Changes to an organization’s workspace can also act as powerful symbolism, with the physical environment communicating information about the organization and its values (e.g., Davis, 1984), effectively supporting or undermining the desired culture and working practices (e.g., Allen & Henn, 2007; Becker & Steele, 1995; Higgins & McAllaster, 2004; McElroy & Morrow, 2010; Turner & Myerson, 1998). For example, design has been used to connect employees to organizational missions and functions, symbolically reflecting and promoting the organization and its working culture. In the case of BMW’s Central Building, for example, the physical flow of cars extends throughout the building, from the shop-floor through the design, technical, and corporate areas, thereby connecting (both physically and symbolically) staff from all functions within the plant to the company’s core business of making cars (Gannon, 2006). Open-plan offices have been proposed as a means to initiate and support more open and collaborative working practices, to integrate business
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
198 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY functions, and to reflect a lack of hierarchy (e.g., Brennan, Chugh, & Kline, 2002; Brookes & Kaplan, 1972). McElroy and Morrow (2010) have recently reported a post-intervention study, incorporating a treatment and control group. They found that office refurbishment (involving the combined use of brighter d´ecor, new furniture, greater openness, and higher workspace density) yielded positive changes in employee perceptions of organizational culture, whereas no such changes were observed in respect of the control group. Employees in the refurbished, more open office reported their organizational culture as being more innovative, less formal, providing more professional control, and fostering greater collaboration than their counterparts in the nonrefurbished control office. In addition, occupants in the refurbished office were found to report greater co-worker satisfaction and affective organizational commitment. Findings in respect of workspace perceptions showed that although employees in the refurbished office were more positive regarding the layout of their office, they were significantly more dissatisfied with the amount of personal space and degree of distraction that accompanied the refurbishment. The study’s design precludes the examination of the contribution of each individual aspect of the refurbishment, with only the effects of the combined intervention observable. Despite the confounding nature of the intervention, these findings support the proposition that a new workspace can aid the adoption of changes to working practices and culture, with physical features imbuing meaning and serving to reinforce nascent change (Higgins & McAllaster, 2004). Exemplifying this line of reasoning, Hall and Ford (1998) described the design of a new factory for Keltec which included the adoption of an open-plan office and manufacturing space to aid communication and improve quality processes. Following the redesign of the plant, which incorporated the removal of many of the physical barriers separating white collar and production teams, the staff demonstrated greater empathy and there was greater understanding between teams, together with speedier communications and resolution of problems. The removal of physical barriers was seen as symbolic of the desired cultural change within the factory and led to greater integration between design and manufacturing. Like the aforementioned McElroy and Morrow (2010) field study, this case study illustrates the potential for open-plan offices not only to cut overheads or affect the frequency of interpersonal interaction, but also to act as a catalyst for wider cultural change within an organization (for further discussion of the symbolism of design see Davis, 1984; Vilnai-Yavetz, Rafaeli, & Yaacov, 2005).
Risks of Open-plan Offices The previous section highlighted the financial benefits that open-plan offices can deliver through savings on facilities and their associated overheads. Indeed, many organizations still regard the design of their office space as a
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
199
largely technical issue, best left to facilities managers and furniture designers (Duffy, 2000). However, we suggest that the design of physical workspaces poses considerable risks (as well as an opportunity for gain) for organizations in financial, organizational, and human terms. At present, the limited attention paid to the interaction between workspace and individuals by businesses (Duffy, 2000) and, indeed, by organization theorists (Becker, 1981), makes the design and implementation of new or refurbished work environments a relatively unmanaged risk. There is a need for managers and researchers alike to consider the risks that housing employees in an open-plan office may pose and to evaluate whether the predominant open-plan format (Vischer, 1996) adequately satisfies user and organizational needs. Although some of the findings we are about to discuss concern environmental factors not solely related to open-plan offices, they are often associated with the implementation of open-plan working and as such are relevant considerations for designers, managers, and staff. For example, reduced architectural privacy (through the lack of walls or significant screens) and increased density in open-plan offices can increase the frequency of uncontrolled interactions (e.g., conversations initiated by particular individuals which other workers in close proximity have little or no opportunity to avoid). Although increased communicative spontaneity is one of the fundamental outcomes that open-plan configurations seek to promote (cf. Brookes & Kaplan, 1972), open-plan offices risk negatively affecting cognitive processes and task performance and/or contributing to stress (e.g., Baron, 1994; Cohen, 1980; Evans, Johansson, & Carrere, 1994; Oldham, Cummings, & Zhou, 1995; Paulus, Annis, Seta et al., 1976; Stokols, Smith, & Prostor, 1975; Sundstrom, Town, Rice, et al., 1994). One major risk of open-plan offices is the greater opportunity for cognitive overload or over-stimulation to occur. Cognitive theory indicates that negative outcomes will occur (e.g., withdrawal from the workplace, reduced environmental satisfaction, decremented task performance) when individuals are subject to excessive social interactions or distraction, which cause them to become overloaded (e.g., Cohen, 1980) or perceptually over-stimulated (Desor, 1972; Paulus, 1980). The proposition is that distractions in the environment can increase cognitive effort, adding to the demands that work may place upon employees, and once an individual’s finite information processing capacity is exceeded, organizations run the risk that task performance and attention will diminish (Baron, 1994). Increased distraction or interruption (e.g., Brookes & Kaplan, 1972; Hedge, 1982; O’Neill, 1994; Sundstrom, Herbert, & Brown, 1982; Sundstrom & Sundstrom, 1986; Sutton & Rafaeli, 1987), together with other risks, such as reduced levels of concentration (e.g., Oldham & Brass, 1979; Oldham & Rotchford, 1983) and lower levels of motivation (Oldham & Brass, 1979), have been consistently associated with high density, open-plan offices with relatively few physical screens between staff. Evidence regarding an organizational consequence of such reactions is provided by Craig’s (2010) survey of 38 000 knowledge workers’ use of predominantly open-plan office
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
200 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY space which found that one of the biggest losses of productive time during the day stemmed from interruptions by colleagues. A further risk is the exposure of workers to a lack of psychological privacy (e.g., Brookes & Kaplan, 1972; Hedge, 1982; Kupritz, 1998; O’Neill, 1994; Oldham, 1988; Oldham & Rotchford, 1983; Sundstrom, Herbert, & Brown, 1982; Sundstrom & Sundstrom, 1986; Zalesny & Farace, 1987), which may result in inhibited overt behaviors, for example personal or confidential discussions and work-related feedback have been found to decrease under openplan or higher density conditions (e.g., Oldham & Brass, 1979; Oldham & Rotchford, 1983). Psychological privacy concerns the amount of control individuals perceive they have over regulating their social contact with others, not least the degree to which they feel visually and/or acoustically exposed (e.g., Altman, 1975; Sundstrom, Burt, & Kamp, 1980). The organizational consequences of reduced psychological privacy, such as inhibited confidential discussions and feedback, will likely vary in relation to an employee’s job role and level, in addition to the tasks in which they are engaged. Environmental satisfaction, usually taken as the degree to which an individual is satisfied with their immediate workspace or area, has been frequently measured in some form in studies involving the physical environment (e.g., Brennan, Chugh, & Kline, 2002; May, Oldham, & Rathert, 2005; O’Neill, 1994; Oldham, 1988; Oldham, Kulik, & Stepina, 1991; Sundstrom, Burt, & Kamp, 1980; Sundstrom, Town, Rice, et al., 1994; Sutton & Rafaeli, 1987). Open-plan workspaces (e.g., Brennan, Chugh, & Kline, 2002) and those offices with raised density or increased proximity of co-workers (e.g., May, Oldham, & Rathert,, 2005; O;Neill, 1994; Oldham, 1988; Oldham, Kulik, & Stepina, 1991; Sundstrom, Burt, & Kamp, 1980) have been related to reduced levels of environmental satisfaction. Given that environmental satisfaction has been found to be positively related to job satisfaction (e.g., Veitch, Charles, Farley, et al., 2007), and in turn to organizational commitment and turnover intent (Carlopio, 1996), clearly another risk that needs to be managed when introducing open-plan working is the potential concomitant decrease in job or work satisfaction (e.g., Oldham & Brass, 1979; Zalesny & Farace, 1987). Indeed, satisfaction with the physical environment is included explicitly as a component of some measures of job satisfaction (e.g., Warr, Cook, & Wall, 1979). Yet another risk that needs to be managed in open-plan workspace is noise. Noise, defined as unwanted sound (Baron, 1994), has often been reported as the greatest issue of dissatisfaction that staff raise when questioned about their open-plan work environments (e.g., Sutton & Rafaeli, 1987). Indeed, Leaman and Bordass (2005) describe noise as the issue that workers would most like to be able to control. The reduction in walls, screens, and acoustical materials, in addition to increased numbers and groups of employees occupying a single space, can give rise to greater noise than would be experienced in single or low occupancy offices. In general, laboratory studies have found relationships
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
201
between increased background noise and detrimental task performance (e.g., Glass & Singer, 1972; Rashid & Zimring, 2008; Smith-Jackson & Klein, 2009). For example, Perham, Banbury, and Jones (2007) found serial recall of digits to be significantly reduced when participants were played background office noise. However, the evidence linking noise and real world job performance is more variable (e.g., Evans & Johnson, 2000; Sundstrom, Herbert, & Brown 1982; Sundstrom, Town, Rice, et al., 1994). The risks associated with open-plan offices illustrate the need for workspace to be considered beyond traditional technical matters. The organizational risk that office design or redesign presents requires a structured response, both to identifying such risks and in evaluating the extent of the threat that they may pose – in essence appropriate risk assessment needs to be developed. Once such environmental risks have been identified mitigation strategies and techniques aimed at limiting or eradicating the effects may be employed. Later in this chapter we briefly revisit the issue of mitigation, reflect upon the trade-offs between the risks and benefits of open-plan working, and explore the potential for the evolving office to satisfy competing user and organizational needs. Individual and Contextual Factors Affecting Open-plan Offices Within the management and I/O psychology literatures, researchers have attempted to investigate whether employee reactions to their workspace, openplan in particular, is uniform (whether negative or positive). A number of studies have attempted to assess the effects that job level and complexity might have on workers’ interactions with their environments (e.g., Brennan, Chugh, & Kline, 2002; Carlopio & Gardner, 1992; Ferguson & Weisman, 1986; Hedge, 1982; Konar, Sundstrom, Brady, et al., 1982; O’Neill, 1994; Oldham, Kulik, & Stepina, 1991; Sundstrom, Burt, & Kamp, 1980; Sundstrom, Herbert, & Brown, 1982; Zalesny & Farace, 1987). With regard to job level, Carlopio and Gardner (1992) found that managers were more satisfied in enclosed offices than their clerical colleagues. The latter preferred more open arrangements. Sundstrom Herbert, and Brown (1982) found that managers who relocated from enclosed to more open workspace reported larger reductions in their privacy than other staff members who experienced similar changes to their workspace. In partial support of these findings, O’Neill (1994) found a weak but significant relationship between job level and environmental satisfaction. Although job level has not been found to be significant in all studies (e.g., Ferguson & Weisman, 1986; Oldham et al., 1991), overall results support the assertion that managers and supervisors respond more negatively to environments that reduce their privacy. The mixed results in this area may partly be explained by the differing operationalization of job level, with some studies simply classifying respondents as managerial or not (e.g., O’Neill, 1994), others using aspects such as job type and number of supervisees (e.g., Ferguson & Weisman, 1986). Charles and Veitch (2002) have noted that, in the main, the
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
202 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY literature points to groups of workers being differentially affected by variations in workspace density, with those individuals in lower level jobs being less affected. Sundstrom, Town, Rice, et al. (1994) have suggested that this is likely to be due to managers requiring greater confidentiality to perform aspects of their role. Alternatively or in addition, a symbolic interpretation would posit that managers and other higher level staff may experience negative reactions, not simply because of the functional inadequacies of an open-plan office, but also because of the loss of status and differentiation that uniform or smaller openplan workstations confer (for further discussions of this issue see Davis, 1984). The effects of task complexity on interactions with office space have also been investigated. Block and Stokes (1989) demonstrated that individuals performed better on a complex task in a room on their own, while a simple repetitive task was performed better in the presence of others. Furthermore, studies have found that specific skills can influence the relationship between job complexity and reactions to the physical environment. For example, stimulus screening skills – how well an individual is able to screen out unimportant, unwanted aspects of their environment (Mehrabian, 1977) – have been found to interact with job complexity, with stronger screeners reporting more favorable outcomes than weak screeners in more open or distracting conditions (e.g., Fried, 1990; Oldham, Kulik, & Stepina, 1991). However, overall the literature is inconsistent: some field studies not having found significant relationships between task-complexity and the work environment (e.g., Sundstrom, Burt, & Kamp, 1980). In addition to examining job level and task complexity, researchers have employed a range of theoretical approaches to assess how individuals perceive or react to their environments. Such approaches include cognitive theories, for example information overload (Cohen, 1980) and overstimulation (e.g., Desor, 1972; Paulus, 1980); social interference theory (e.g., Baum & Paulus, 1987; Oldham, Cummings, & Zhou, 1995); and stress-based models (e.g., Paciuk, 1990). In general, cognitive approaches have suggested that workers who are not cognitively challenged by their work have greater capacity to accommodate unexpected social interactions or distractions (e.g., Baron, 1994). A Trade-offs Perspective Previous reviewers (e.g., Elsbach & Pratt, 2007) have noted that the design of the physical environment involves trade-offs in the management of competing tensions between its different aspects. The evidence surrounding the benefits and risks of adopting an open-plan workspace strategy illustrates the need to ensure that potential negatives, such as increased distraction, noise, and reduced privacy (e.g., Brookes & Kaplan, 1972; Hedge, 1982; Leaman & Bordass, 2005; O’Neill, 1994; Sundstrom, Herbert, & Brown, 1982; Sundstrom & Sundstrom, 1986), do not outweigh the financial and behavioral positives that might be delivered (e.g., Duffy, 2000; Hundert & Greenfield,
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
203
1969; Ives & Ferdinands, 1974; Zeitlin, 1969). However, mixed findings (Boyce, 1974; Brennan, Chugh, & Kline, 2002; Brookes & Kaplan, 1972; Hedge, 1982; Oldham, 1988; Oldham & Brass, 1979; Zalesny & Farace, 1987) illustrate the difficulty in attempting to draw clear-cut conclusions in regard to when an open-plan office is most appropriate for an organization, or which aspects of such a design pose the greatest potential risk (e.g., higher density levels, lower level screens between or around workstations). Although there are substantial risks to implementing an open-plan concept, there is the potential to minimize these effects. For example, techniques such as pumping in white noise (low-level unstructured noise from across the audible sound spectrum) or piped music, or the use of noise dampening materials, may be used to mask intermittent office noise (e.g., human speech or telephones ringing) (e.g., Vischer, 1989), although their efficacy is not confirmed (Navai & Veitch, 2003). Furthermore, Brennan, Chugh, and Kline (2002) have suggested that the use of agreed protocols may provide a technique with which to minimize the effects of disturbing unpredictable noise, such as co-worker conversations. In their evaluation of an office relocation, they commented that the increase in desk-side impromptu meetings, which accompanied the introduction of open-plan working, might have been avoidable if clear protocols had been agreed to regulate where such activities took place. The use of such behavioral protocols may be an alternative approach to reducing auditory interruptions, without resorting to costly technical or reconfiguration techniques. Designers need to be aware that employees may not react uniformly to open-plan offices (Sundstrom, Herbert, & Brown, 1982), as the tasks and roles that staff perform influence the extent to which the design poses a risk. Furthermore, differences in the configuration of open-plan space, such as the spatial density of employees, may make an office less suitable for some types of employee (Charles & Veitch, 2002). Consequently, housing large, diverse groups of workers within a uniform open-plan office may be counterproductive for organizations. A more nuanced view is required, one that recognizes that open-plan inherently involves trade-offs. These trade-offs may in part be negotiated by varying the configuration of open-plan within an office, for example, providing different forms of open-plan space for differing employees, striking a balance between competing needs. In summary, the flexibility of space that open-plan offices provide (e.g., Marquardt, Veitch, & Charles, 2002) may need to be adapted and fine-tuned to suit the needs of diverse sets of employees.
EVOLUTION OF OPEN-PLAN OFFICES Open-plan offices may have become the workspace solution of the twentieth century but the office continues to change and evolve (Laing, Duffy, Jaunzens, et al., 1998), posing fresh challenges to I/O psychologists’ understanding of
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
204 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY workers’ interactions with their environments. Open-plan is evolving in that the format is being adapted and modified to engineer spaces that better reflect modern workers and the modern business landscape. In this section we discuss the driving forces behind such changes, the form that these new offices are taking, and what we currently understand about the effects of their design. The Drivers of Change The design and operation of workspace has always been driven by a number of often competing interests, such as: 1. The cost of building, maintaining and servicing the space; 2. Providing for the comfort and security of occupants; 3. Accommodating new technologies (e.g., the emergence of personal computers); 4. Supporting working styles and processes; 5. Upholding organizational structure and corporate image; 6. Aiding recruitment (through providing an attractive place to work); and 7. Location (see e.g., Allen & Henn, 2007; Becker, 1981; Becker & Steele, 1995; Duffy, 1997; Duffy, McMahan, & Pringle, 1999; Laing, 2006; Sundstrom & Sundstrom, 1986; Vischer, 2005 for further discussion). The work environment both reflects and accommodates the changing economic circumstances and the nature of work itself and so is prone to adaptation as business needs progress. Just as new technology has shaped and influenced the nature of offices in the past (e.g., the typewriter produced large typing pools, the personal computer altered the nature of tasks performed at a desk), it is once again revolutionizing the way we work and the space requirements that this entails. The advent of increasingly affordable laptop computers means that workers are no longer bound to a single desk to operate the technology; computers can be readily moved around an office or multiple locations. Indeed, battery power and wireless network connections mean that traditional desks are not a prerequisite for work at all – coffee tables, touch-down spots, or even just an individual’s knee can be sufficient.1 Video conferencing, remote network access, and reroutable telephone lines allow workers to work with colleagues and teams from around the globe (Felstead, Jewson, & Walters, 2005; Laing, 2006). Colocation is no longer a necessity for work groups and teams may operate in 1
Increased portability offers flexibility to workers regarding where they can physically work and allows them to maximize their working time. However, care needs to be exercised to ensure that working away from a desk does not compromise safe working. Ergonomic and health and safety considerations make permanent or extensive use of laptop computers or similar technologies in such conditions undesirable.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
205
temporally disparate patterns (Bell & Kozlowski, 2002), enabling interaction with colleagues in other time zones. As with the rise of open-plan working, the adoption of such technologies is partly attributable to the organizational costsaving that can be realized through the use of technologically enabled practices such as tele-working and home-working (e.g., Chapman, Sheehy, Heywood, et al., 1995; Felstead, Jewson, & Walters, 2005; Ng, in press) which allow both transport and accommodation costs of employees to be reduced. In addition to technological advances and cost reduction, the changing nature of work is an important driver of current office evolution (Laing, 2006). Key to this evolution is the continued growth of knowledge working, both as a percentage of the economy and of the labor force (Davenport, 2005). Knowledge work can be described as involving the application of “theoretical and analytical knowledge,” exemplified by individuals involved in areas such as product development or consultancy work (Parker, Wall, & Cordery, 2001). Knowledge work is often contingent upon the collaborative efforts of multiple individuals. Previously, open-plan offices enabled organizations to house workers in spaces that promoted inter- and intra-team information sharing and interaction, by locating individuals proximally to one another and removing physical walls and obstructions (e.g., Brookes & Kaplan, 1972; Hundert & Greenfield, 1969; Ives & Ferdinands, 1974). Whilst useful in supporting knowledge working, such an approach remains a relatively blunt tool, as it fails to acknowledge the variety of tasks that modern knowledge workers may be involved in, the distributed nature of their interactions, and the shifting temporal nature of their roles and tasks. Our own preliminary analysis of data gathered from the post-occupancy evaluation of a new research and development facility supports the view that staff utilize different workspaces dependent upon the task with which they are engaged. For example, we have found that within the new facility, 70% of the facility’s staff spend at least 40% of their work time in spaces other than their individual workstation (predominantly in formal or semi-formal meeting spaces) (Davis, Leach, & Clegg, 2010). Differences in the nature of tasks individual knowledge workers engage in have been noted by Becker and Sims (2001), who discuss evidence regarding how the time spent on solo tasks and more collaborative activities can vary widely between individuals of similar job titles. Indeed, Robinson (2010) analyzed how design engineers spend their time and established that individuals averaged over 55% of their work time engaged in information behaviors (including answering colleagues’ questions and conversing socially), with around 31% of time spent on solo technical activities. Furthermore, Craig’s (2010) study of task and space use, involving over 38 000 knowledge workers, found that on average they spend at least 40% of their time engaged in interactive or collaborative tasks. Collectively, these findings illustrate that knowledge workers frequently undertake a range of tasks, that these tasks may be undertaken in different work spaces, and that the combinations of tasks and spaces are likely to vary between individual workers.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
206 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY The changing nature of work and workspace is causing fundamental shifts in how organizations approach their space planning and manage their staff (Laing, 2006: 33). Architects and designers are being asked to deliver workspaces that are able to accommodate the competing demands of fluctuating occupancy levels, to enable employees to participate in a greater range of work tasks, and to facilitate collaboration across work groups and departments – and to do so within budgets that are more constrained than ever. The Form of the Evolving Office Alternatives to the established open-plan design and traditional enclosed offices are becoming more commonplace in practice (Gillen, 2006). One approach to accommodate the competing demands described previously is to design offices based primarily upon the patterns of work of its occupants and their respective needs for collaboration. Such designs often incorporate social hearts (or hubs) and “streets” that enable planned and unplanned encounters to take place. These offices also provide spaces that offer different functionality that all workers can access as and when required (e.g., team spaces, reading rooms, computer hubs, formal meeting rooms, and caf´e areas). Financial and space savings can be realized through reducing the provision of strictly “individual” workspaces, with the emphasis upon providing mixes of space that are appropriate to groups of workers (see e.g., Allen & Henn, 2007; Becker & Steele, 1995; Gillen, 2006). Other approaches such as utilizing hot-desking (where desks are available to any worker as and when required) or hoteling (where unassigned desks are reserved by workers for a given period) within established open-plan workspaces are also being employed. This can allow organizations to reduce the total number of desks (and concomitant office space) as they no longer have to provide or assign desks to each individual. These practices can be particularly useful where workers frequently work at client offices or spend a large amount of time traveling or in meetings. Such practices reflect the reality that office occupation rates are unlikely to be 100%, and in organizations that involve activities such as large amounts of traveling by sales staff or consultants, then this rate may be substantially lower (Markland, 1995). To support more mobile or transient working patterns, non-traditional satellite offices or neighborhood work centers have been adopted to allow workers (either from the same or from a number of different organizations) to use office space based upon their location (Cascio, 2000). Non-traditional satellite offices tend to be sited in convenient locations and draw workers from across an organization based upon proximity rather than organizational structure. Neighborhood work centers serve a similar function, allowing workers to use offices closer to where they live or need to be; in these cases, however, the offices are shared by a number of organizations, allowing access to a greater
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
207
number of locations than a single organization could provide (Fritz, Higa, & Narasimhan, 1995). Workers are able to “hotel” at the office that is most convenient to them at the time, rather than being restricted to where their particular department is located or their company’s nearest sole occupancy office. Social and informal meeting spaces are also taking on enhanced roles in the evolving office. Becker and Steele (1995) observe that it is necessary for organizations to provide areas that allow workers to meet informally if intraand inter-team collaboration is to flourish. This goes beyond simply removing office walls and partitions, or seating colleagues closer together; rather, the focus is upon designing a variety of spaces that can help to foster the types of interactions desired, in addition to allowing space for more individualistic tasks. Case studies exploring the provision of social space within contemporary office redesigns have consistently found that it helps to foster informal meetings and wider interactions (Becker & Steele, 1995). Furthermore, flexible workspace and easy access to meeting rooms have been related to higher job satisfaction and group cohesiveness (Lee & Brand, 2005). Allen and Henn (2007) argue that it is important for the physical space to be configured to facilitate the communication and work patterns required by the job. This may mean providing what Becker and Steele (1995: 78) term “activity magnet areas,” such as caf´e areas where individuals may eat their lunch, have a drink, hold informal meetings with colleagues, or use for quiet reading. McCoy (2005) notes that providing a mix of different meeting spaces close to teams can help increase impromptu meetings and serendipitous interactions (e.g., Peponis, Bafna, Bajaj, et al., 2007), thereby encouraging team communication and collaboration. Providing adequate space for impromptu meetings to occur within the office may help to maximize the potential of open-plan working (e.g., increased visibility and communication) while limiting negative effects on those working on solitary tasks (i.e., by moving impromptu meetings away from co-workers’ desks). In a similar vein, Duffy (1997) has suggested that modern offices should offer workers a variety of differing types of workspace, dependent upon the characteristics of their job and work styles. These characteristics include the degree of autonomy that the job entails, the level of interaction required between colleagues, the duration of the work that they engage in, and the amount of office-based time (occupancy level). Duffy (1997) articulated a schema comprising four differing workspace solutions that are best suited to supporting distinct types of workers and working patterns, based upon dimensions of autonomy and interaction (the hive, cell, den, and club) (see Figure 6.1 for an illustration of this schema). According to Duffy (1997), increasing the fit between the design of the workspace and the demands of the work will lead to more effective and satisfied employees (see also Laing, Duffy, Jaunzens, et al., 1998). More generally, this approach of satisfying needs and demands
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
208 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY High
Knowledge Working, e.g. Advertising
Group Working e.g. Media
DEN
CLUB
Team Space – Meeting and Work Spaces
Multiple Task Space – Diverse, Manipulable Spaces
Interaction Concentrated Working, e.g. Lawyers
Solo-Working, e.g. Tele-Sales
Low
HIVE
CELL
Individual Workstations – Simple Open-Plan
Cellular Offices or Highly Screened Open-Plan Autonomy
High
Figure 6.1 Schematic illustrating Duffy’s (1997) distinction between differing office designs and their support for working practices. (Figure based upon concepts developed by DEGW, Frank Duffy and Andrew Laing, published in Laing, A., Duffy, F., Jaunzens, D., & Willis, S. (1998). New Environments for Working: The Redesign of Offices and Environmental Systems for New Ways of Working. London: Construction Research Communications Ltd., page 23, reproduced by kind permission of DEGW.)
is incorporated under the umbrella of psychological needs-based approaches to workspace design (see also Vischer, 1989). Such approaches have been found to be applicable in a range of organizational contexts, with working patterns and use of space largely explained by the particular classification system adopted (for additional representative examples see Allen, Bell, Graham, et al., 2004; Laing, 2006; Laing, Duffy, Jaunzens, et al., 1998). Turner and Myerson (1998) suggest, from their experience of both research and the design of new workspaces, that “it is the rich and varied setting of the ‘Club’ which best illustrates the way the new office is going, with its high levels of both autonomy and interaction” (1998: 73). Duffy’s (1997) schematic captures the way in which contemporary offices are becoming ever more diverse, ranging from the traditional enclosed single occupancy offices and high density open-plan forms, through offices containing large amounts of team space and meeting areas but which offer little individual desk space, to those which have large amounts of all of these spaces and more (e.g., reflective space, libraries, and caf´es).
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
209
Effects of Evolving Offices Contemporary office designers are increasingly seeking to provide a mix of workspaces within largely open-plan offices which provide for workers’ diverse needs and reflect their increasingly flexible work patterns (see e.g., Laing, 2006). For instance, offices that incorporate a mix of differing workspaces (e.g., individual workspaces, quiet rooms, team-spaces, meeting rooms) to facilitate different styles of working and types of tasks have been successfully implemented by the architectural consulting group, DEGW, in a number of UK public sector refurbishment and redesign projects (Allen, Bell, Graham, et al., 2004). These projects have demonstrated that it is possible to design multiple workspaces, often within a broadly open-plan style office, which facilitate different levels of interaction, forms of working, and technology use. For example, a refurbishment of the UK’s HM Treasury offices involved the introduction of a large number of informal meeting areas, partly to increase the amount of team-working space. This project was used to help support collaborative working and to ensure that the individual areas were sufficiently quiet to enable cognitively demanding work to be undertaken (i.e., space that is quiet enough for individuals not to require separate “quiet booths”). Within the UK Department for Trade and Industry, a flexible workspace concept was introduced utilizing modern IT (e.g., wifi, laptops, telephone systems that can reroute numbers to any desk) to allow hot-desking within open-team space. In addition, “touch down” spots (places with network connections around the facility to allow workers to use laptops without requiring a traditional workstation), project areas, quiet spaces, and a caf´e were introduced to support flexible working around the building. Hot-desking and the inclusion of other work areas allowed the designers to reduce the individual desk space from 1 : 1 to 8 : 10, freeing space for a higher proportion of task relevant space. Contemporary offices that involve a reduction in individual workspace (either to enable space rationalization or to allow the inclusion of other activity areas) or changes to working practices (e.g., compulsory remote working to allow a reduction in the number of desks) have not been introduced without controversy. Offices where employees do not have their own desk or personal space have been criticized for failing to provide adequate personal control or territory for individual workers (e.g., Danielsson & Bodin, 2008), which in turn can lead to counterproductive work behaviors (for a comprehensive review of literature concerning territoriality see Brown, Lawrence, & Robinson, 2005). Danielsson and Bodin (2008), however, have found somewhat conflicting evidence. They surveyed occupants of a number of different types of office: cell office (tradition single enclosed room/workspace); shared room (two to three people sharing a room); small, medium, or large open-plan offices; “flex-office” (no individual workstations but comprising a variety of spaces to support different types of working); and the “combi-office” (employees spend more than 20% of their time in workspaces other than their own,
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
210 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY e.g., team-based space). Their findings indicate that workers are as satisfied in a flex-office as in a shared room or cell office, and more satisfied than in open-plan or combi offices. These results, although only based on a relatively restricted sample, suggest that in the right circumstances, flexible workspaces can offer both individuals and organizations a good solution to managing diverse work needs. As noted by several authors, there is very limited evidence with which to evaluate the effects that workspace concepts such as tele-working, desk sharing, or hoteling might have on individual or organizational outcomes (De Croon, Sluiter, Kuijer, et al., 2005; Ng, in press; Vos & van der Voordt, 2001). There is a paucity of published work that describes the outcomes and contingencies for workers housed in these new workspaces, or for those who tele-work from home frequently. De Croon, Sluiter, Kuijer, et al. (2005), however, note that the limited evidence available suggests that desk sharing (or hot-desking) may improve communication between workers; although Vischer (2005) has highlighted potential dangers of implementing such radical shifts in workspace use as it can be accompanied by rejection of the new working practices that accompany such designs. A recent study by Millward, Haslam, and Postmes (2007) of workers who had been randomly assigned fixed desks or hot-desking found relatively neutral reactions. For instance, they found that workers assigned to hot-desking were not alienated by the change, although they did place a higher value on electronic communication than their assigned desk counterparts. Once again, organizational cost saving is suggested as a driving force behind the rapid promotion and adoption of tele-working and home-working (e.g., Cascio, 2000; Felstead, Jewson, & Walters, 2005; Ng, in press). Encouraging employees to work at home, or from client sites or coffee shops, allows organizations to shift some of the costs of providing workspace onto other parties or the employees themselves. In return the employee may be able to take greater control over choosing the work area that they feel most comfortable in, and in managing their work–life balance. Indeed, one recent review has suggested that home-working may provide a number of benefits to employees, including well-being, and job and life satisfaction (Redman, Snape, & Ashurst, 2009), although empirical analysis examining the individual experience of such arrangements is limited. In summary, contemporary offices are evolving from the established openplan format to become more diverse, less desk-bound, and more adaptive in form. Organizations are redesigning their existing open-plan office space to optimize it for contemporary working practices. This change is driven in large part by the advances in mobile and communications technology (e.g., Duffy, 1997; Felstead, Jewson, & Walters, 2005; Laing, 2006) and a desire for further cost reduction (e.g., Duffy, 2000), as well as the increasing prevalence of knowledge working (e.g., Davenport, 2005) and the diverse range of tasks that employees engage in (e.g., Becker & Sims, 2001). Optimized open-plan or more flexible office spaces often utilize techniques such as hot-desking, or
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
211
home-working, to allow space either to be saved or freed up to be used in different ways (e.g., Allen, Bell, Graham, et al., 2004). The prevalence of more sophisticated open-plan and flexible workspace is likely to accelerate as organizations continue to redesign existing office space and to invest in new buildings that reflect ongoing technological advances and increasingly complex work and work patterns. In order to provide advice and insights that can inform the design and management of such environments, sustained research attention in this area is required, mindful of the fact that the introduction of new workspaces and the redesign of existing ones in ways that affect an individual’s territory, work practices, or experienced control may produce negative reactions (Danielsson & Bodin, 2008).
MANAGING THE PROCESS OF CHANGE This section reviews theory and research pertaining to the management of the process of change that accompanies the design of new workspaces or the redesign of existing ones. While acknowledging that there is a substantial literature that concerns organizational change in general (e.g., Burnes, 2004; By, 2005; Clegg & Walsh, 2004), we focus upon theory and case studies that have been applied specifically to the domain of contemporary office environments. We discuss the idea that new or redesigned workspaces can involve significant changes for employees, the similarities of the process of workspace design to organizational change, the role of user involvement in changing physical workspace, and the application of socio-technical principles. New or Redesigned Workspace Involves Change Whether a firm embarks upon a modest refurbishment of an existing openplan office or seeks to introduce a highly contemporary workspace, for example incorporating aspects of flexible space and tele-working, the activity of design and eventual occupation will almost certainly usher in changes, both for individual workers and for the organization as a whole. The design of a new office (or redesign of an existing one) often involves changes in spatial configurations, facilities, or technologies that can significantly alter the way in which individuals and teams go about their work (e.g., Laing, Duffy, Jaunzens, et al., 1998). This is aside from the altered sensory experience that features of a welldesigned office, such as improved lighting or ergonomic furniture, may deliver. More specifically, the adoption of open-plan working can have major effects on employees’ work experiences, most likely originating from differences in the frequency and nature of interactions (e.g., Ives & Ferdinands, 1974), visual and auditory distraction (e.g., Sundstrom & Sundstrom, 1986), and the location of other teams and colleagues (see also McElroy & Morrow, 2010). Indeed, even modest redesigns to existing open-plan offices, for example introducing break-out areas, may significantly affect work experiences for better or worse. For instance, a greater level of background noise for individuals
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
212 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY located near the break-out areas might have a detrimental effect on performance. Furthermore, introducing a radical new office concept, for example including street layouts, collaborative rooms, and reduced individual workspace, may require workers to embrace new working practices, including a more informal approach to meetings (e.g., Brennan, Chugh, & Kline, 2002) and hot-desking (e.g., Duffy, 1997). All of these changes to the physical environment, therefore, require careful design, facilitation, and implementation if the result is to reflect and meet the needs of individual employees (Becker, 1981). Similarity to Organizational Change Management The design and implementation of a new office concept, or the redesign of an existing one, can be considered as a form of discontinuous organizational change as it introduces a one-time change to the group affected (Luecke, 2003). However, active management of the design process leading up to the introduction of a new office environment and support following its introduction can transform the process into a less discrete change. Indeed, a new office can initiate and support changes to working practices (e.g., enhancing collaboration) and culture (e.g., Turner & Myerson, 1998), transforming such interventions into incremental forms of organizational change. Badly, managed, such interventions will breed resistance and resentment, as with any poorly orchestrated organizational change process. Despite the substantial literature concerning change management within the management and I/O psychology domains (e.g., Brown & Eisenhardt, 1997; Burnes, 1996; By, 2005; Clegg & Walsh, 2004; Holman, Axtell, Clegg, et al., 2000; Kanter, Stein, & Jick, 1992; Kotter, 1996; Luecke, 2003; Pettigrew, 1985; Pettigrew, Woodman, & Cameron, 2001; Van de Ven, & Poole, 1995; Weick, 1979; Woodman, 1989), there is currently only a very limited acknowledgement of the potential for workspace to support or initiate change, whether intended or not (e.g., Lawler & Worley, 2006; McElroy & Morrow, 2010). Architectural and design-led studies exploring this issue have found that engaging end-users in, and allowing them a degree of control over, the design process is beneficial both to the design of new workspaces and to aiding employee acceptance of changes to working practices (e.g., Blundell-Jones, Petrescu, & Till, 2005; Turner & Myerson, 1998). Studies examining the effects of enduser involvement in the design of information systems and work processes show similar positive findings (e.g., Mumford, 1983). Oldham, Cummings, and Zhou (1995) have previously alluded to the potential positive effects of worker participation in the design of their own workspace. Studies of employee control over more specific features of their workspace (in the form of environmental control or physical adjustability) have generally found such opportunities to be related to increased job satisfaction, performance, communication, privacy, and satisfaction with the environment (e.g., Huang, Robertson, & Chang, 2004; Lee & Brand, 2005, 2010; O’Neill, 1994). Architectural research exploring the effects of building design in healthcare settings suggests that the
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
213
provision of control over the environment to patients is associated with tangible individual benefits including improved treatment completion times, reduced medication levels, and enhanced well-being (e.g., Lawson & Phiri, 2003). More broadly, Vischer (2005) proposed seven principles specifically for the management of effective workspace change, which emphasize how the design process may be used to empower stakeholders to challenge the status quo, to re-evaluate work processes and structures, and to use the process to surface and overcome potential resistance. Underpinning Vischer’s principles is a focus upon user participation and the bi-directional sharing of information and suggestions. Vischer’s (2005) approach and the wider architectural practice commending user participation and engagement (e.g., Blundell-Jones, Petrescu, & Till, 2005) share similarities with much of the change management literature, in which employee involvement is actively encouraged as part of a change management strategy (e.g., Armenakis & Bedeian, 1999; Clegg & Walsh, 2004; Kanter, Stein, & Jick, 1992; Mumford, 1983; Woodman, 1989). Supporting this principle, user involvement has been demonstrated as a key factor in determining the success of more general organizational change programs (Holman, Axtell, Clegg, et al., 2000). We suggest that the design of a new work facility encompasses similar issues to change programs in general, and to technology-led innovations in particular, due to the tendency for “experts,” such as IT professionals, to “design a system, and then push it at its end users” (Clegg & Shepherd, 2007: 215). In this context, the equivalent process is one whereby facilities managers or designers specify and design a new office space without due involvement of the workgroups to be accommodated. This is in direct opposition to what has been described as “pull-based user-owned change” (Clegg & Walsh, 2004: 235), whereby end-users pull the project through to successful completion by taking ownership of, and having input into, the design and implementation process, ensuring that it meets their needs. The involvement of employees provides a means to ensure that the work environment not only better reflects their requirements, but also allows them to take ownership over the process. Furthermore, acceptance of changes to workspace is important if new flexible concepts are being introduced that affect other aspects of work processes (e.g., introducing home or tele-working) (e.g., Baruch, 2001; Chapman, Sheehy, Heywood, et al., 1995; Daniels, Lamond, & Standen, 2001). Successful User Involvement in Workspace Design A number of studies within the I/O psychology and management literatures have examined the effects of changes in physical office design or configuration (e.g., Brookes & Kaplan, 1972; Oldham, 1988; Oldham & Brass, 1979; Zalesny & Farace, 1987) on employee reactions; however, there has been limited examination specifically of the process of change (McElroy & Morrow, 2010) and of user involvement in particular. Case studies from environmental
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
214 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY psychology and architectural spheres have demonstrated how the process of user participation in design can be used to successfully manage organizational change (e.g., Allen, Bell, Graham, et al., 2004). Furthermore, related approaches that incorporate user involvement (e.g., Socio-Technical Systems Design) support this contention (e.g., Mumford, 1983). To highlight the techniques adopted and the potential benefits that user involvement may deliver, we describe two case studies. The first (Foland, Rowlen, & Watson, 1995) concerns the introduction of open-plan working, whereas the second (reported in Box 6.1) describes our own reflections on the redesign of an existing open-plan office. We present these case studies as exemplars of the work being conducted in this field and to spur further investigation in the area.
Box 6.1 Redesign of an Existing Open-Plan Office Over the past 2 years, we have been involved in the redesign and evaluation of a number of large open-plan offices within the UK operation of a global aerospace and defense engineering company. During this time we have worked closely with a number of stakeholders and staff who have been involved in the redesign of their offices and/or who have been affected by changes that have been introduced. Our experience has demonstrated that when the staff are actively involved in the design process, either helping to make decisions regarding aspects of the design or providing real input regarding the needs they have for the workspace, the staff not only report that they are more satisfied with the quality of their workspace, but also that the space more accurately reflects and accommodates their functional needs. In one set of refurbishments, two managers led a process that sought to engage with members of a 180 strong department to define their functional requirements. Representatives fed information and ideas forward and back to the managers and corporate facilities team. In addition to this, the proposed office plans were distributed to staff and physical mock-ups were constructed using the proposed furniture. Managers used the feedback from these activities to determine the aspects of the environment that were most important in enabling staff to perform their work tasks, in addition to establishing what they were not prepared to compromise on. Crucially, the managers demonstrated leadership and flexibility in negotiating with their facilities colleagues. They were consequently able to work within the company’s office standards to deliver increased desk space for a subset of the engineers (who required greater layout space for their work), together with a greater number of informal break-out areas to allow more spontaneous small group meetings (to relieve pressure on meeting
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
215
rooms). Importantly, the refurbished office achieved the overall corporate aim of an increase in density. The involvement of staff in the process was instrumental in helping to allay initial employee and union concerns over the headline density increases that the refurbishments initially appeared to embody. The case study demonstrates how involvement can help define core workspace requirements whilst acting within agreed organizational standards. Foland, Rowlen, and Watson (1995) describe a project in which facilities managers at Amoco Oil & Gas embarked on a program to rationalize their workspace costs and to embed team-based working, moving from enclosed to more open-plan workspaces. In a pilot study, the facilities department worked closely with the leader of a specific work team to facilitate a highly participatory approach to the redesign of their office space. The process capitalized on the team’s knowledge and expertise of their working practices, with staff involved in design decisions, for example furniture styles, seating arrangements, and use of workspace. The redesign became a process driven by the team’s understanding of their work processes and needs. The emphasis was on how they could work more efficiently and how the new workspace could then be designed to support these changes in working practices. The authors noted that the process itself helped the department improve conflict resolution between team members and foster a greater understanding of group needs, as well as aiding the integration of interns and temporary workers within the teams taking part. The resulting new office, accompanied by the new ways of working it enabled and supported, produced a 25% decrease in project cycle times, 75% decrease in formal meeting time, increased team learning, increased problem solving, and led to higher quality products (Foland, Rowlen, & Watson, 1995: 683). However, when the organization attempted to roll out the new office concepts across other work groups, they encountered resistance from workers, largely due to the top-down implementation and absence of a participatory approach (Vischer, 2005). These outcomes show striking similarities to the wider change management literature (e.g., Clegg & Walsh, 2004) and earlier classic work on socio-technical design in office environments (Mumford, 1983). As it had worked well in one situation, management believed that the office concept could be simply replicated across the wider organization; they failed to appreciate the role that participatory design had played in crafting the most appropriate environment for that particular team and in helping the team to accept the resulting changes in work practices. Applying Socio-Technical Principles A related approach that is applicable to the design and management of workspace change, previously touched upon during our discussion of user
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
216 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY involvement, is socio-technical systems thinking (e.g., Cherns, 1976, 1987; Clegg, 2000; Mumford, 1983; Trist & Bamforth, 1951; van Eijnatten, 1997). Socio-technical systems thinking argues that an organization is a complex system made up of a number of inter-related parts, including the individual staff, the work processes, the technologies, and so forth. The approach grew out of a series of studies conducted at the Tavistock Institute of Human Relations, London, in the 1950s and 1960s (van Eijnatten, 1997). Trist and Bamforth (1951) published seminal work based upon their observations of the “longwall” coal mining method, following the introduction of large-scale machinery. The coal mining methods demonstrated the importance of autonomy, multi-skilling, and self-supervision and the need for behavioral issues to be considered during technological design and implementation. Socio-technical thinking continued to evolve and Cherns (1976) enunciated nine core principles of socio-technical design, later extended to 10 (Cherns, 1987). The approach has been refined further, with Mumford setting out the “Ethics” approach to the design of new information systems from the late 1970s onwards (e.g., Mumford, 1983, 1995; Mumford & Weir, 1979). More recently, Clegg (2000) elaborated and extended Chern’s (1987) principles to apply to modern IT design (for a comprehensive description and timeline of the development of socio-technical systems theory, from its inception to modern advancements, see van Eijnatten, 1997). The application of socio-technical theory has predominantly focused upon the industrial sector and the introduction of new technologies (e.g., Advanced Manufacturing Technologies and office-based technologies) (Clegg, 2000), with limited attention having been paid directly to the design of the physical work environment. Previously, Mumford (1983) applied socio-technical principles to the design of information systems. Mumford’s approach involves large amounts of user participation in the design and configuration of new information systems and seeks to use technology to help improve the work experience and organizational effectiveness of the system as a whole. For example, user involvement in the design and implementation of a new word processing system was used by Mumford (1983) to find ways of meeting both user and organizational needs, increasing the acceptance of the system and its associated changes for all concerned. Despite the success of applications of socio-technical theories, I/O psychologists have rarely applied the ideas and principles to the design of the physical environment. Authors from across disciplines have, however, suggested that the physical work environment should be considered as part of the overall organizational system (e.g., Allen & Henn, 2007; Becker & Steele, 1995; Blyth & Worthington, 2001; Ferguson & Weisman, 1986; Haynes, 2007; Lawson, 2004; Preiser, 1994; Trist & Bamforth, 1951; Turner & Myerson, 1998). We argue that in practice socio-technical systems theory should be broadened to consider the whole work system, being applied more comprehensively to the design of the physical environment alongside the design of new processes,
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
Goals/Visions/ Values
217
People
Buildings/ Infrastructure
Processes/ Procedures
Culture
Technology
Figure 6.2 Socio-technical system, illustrating the inter-related nature of an organizational system. (Source: Challenger, Clegg, & Robinson, 2010: 74. Reproduced by permission of the Stationary Office.)
job roles, and technologies (i.e., extending the scope of the work system under investigation). Furthermore, this new application domain provides excellent opportunities for us to explore how current socio-technical design principles (e.g., Clegg, 2000) may be extended to take account of the specific challenges and contingencies that workspace design involves. A systems approach is applicable to workspace design as it encourages conflicts or detrimental effects to be identified as decisions are made, minimizing the likelihood of one part of the system, or set of drivers, forcing unintended change upon the others (see Figure 6.2 for diagrammatic representation of the inter-related nature of a work system). Socio-technical theory acknowledges that design involves compromise, and this can be viewed as part of the process that establishes a balance between the competing elements of the work system (Clegg & Shepherd, 2007; Hendrick, 1997; Nadin, Waterson, & Parker, 2001). Indeed, as others have noted previously (e.g., Allen & Henn, 2007; Elsbach & Pratt, 2007; Sundstrom, Town, Rice, et al., 1994), work environments involve trade-offs between what is most appropriate or desirable for the staff and other stakeholders involved and what is necessary or possible within organizational and technical constraints. A socio-technical approach to design can be viewed as one way of enabling and promoting open and systematic consideration of these competing demands, to help find new ways of working and working practices that may meet the joint needs of the various stakeholders and the organization (Ridgway, Cerulli, Davis, et al., 2008). A socio-technical approach to the design of the physical work environment would encourage the integration of disciplinary knowledge and expertise, for example bringing together architects, engineers, psychologists, technology specialists, with users and stakeholders. To illustrate how the principles can be applied in practice, we present a recently completed case study that has investigated a socio-technical approach to workspace design (Box 6.2).
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
218 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
Box 6.2 A Socio-technical Approach to Workspace Design Ridgway, Cerulli, Davis, et al. (2008) describe the application of this systems approach throughout the design of a new R&D facility. The design process was organized in a series of stages and included, in particular: early work (prior to the architectural brief) on the goals, mission, and vision of the new facility; development of a good understanding on the kinds of work and projects that would be undertaken, including the technologies that would be used; an understanding of the kinds of staff and numbers that would be employed; the definition of the working culture that the building was trying to promote and support; the design of the layouts of the office and shopfloor areas; the selection of d´ecor and furnishings; the design of key social spaces, including meeting rooms, a social hub, and the dining and reception areas; and the overall design from sustainability and energy-use perspectives. The approach included: extensive user and stakeholder involvement (using a range of techniques); multidisciplinary design meetings (consisting of architects, facilities managers, other professionals and academics); and post-occupancy evaluations. A key element of this process was the initial engagement and facilitation activities to define the brief for tendering architects, essentially setting the direction for the whole design process using scenario planning techniques (Clegg, Cooch, Hornby, et al., 1996). These preliminary activities included workshops with stakeholders and staff to identify the organizational vision, structure, and working practices for the factory. During the stakeholder event, break-out groups discussed key questions relating to the factory: What is our vision of the new factory? What excites us about this new factory? What are the key operational decisions we need to make before we start building? During the scenario planning workshop, stakeholders were encouraged to examine different scenarios for the new facility in terms of its main processes, staff, and outputs. Overall, this socio-technical approach not only identified previously unknown requirements for the R&D facility, which would not have been highlighted without the involvement of frontline staff, but also ensured that design aspects of particular importance to stakeholders and staff were not engineered out to reduce costs (e.g., the social heart and flexible break-out areas) (Ridgway, Cerulli, Davis, et al., 2008). The involvement of the staff provided insights into the functions that the workspace would need to provide and confirmed that a generic space would not be adequate to support the varied nature of the engineers’ roles. It was especially apparent that meeting space was a high priority and the level of space provided for this would need to be far higher than was anticipated prior to consultation (based on traditional assumptions as to the nature of the
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
219
engineers’ jobs), with a mixture of both formal and informal meeting spaces being supplied. Post-occupancy interviews have demonstrated that although the user involvement did not always result in employees feeling that they had had a meaningful impact on the end design (potentially due to budgetary constraints limiting some design features), they reported that the process had helped them to understand the change that was imminent and to feel included in the design process. Ultimately, the combination of techniques used to understand the human and organizational needs for the new workspace have resulted in a building that provides a mix of office and engineering space, reflecting the diverse tasks that the staff are involved in (McGourlay, Ridgway, Davis, et al., 2009).
In summary, the design and implementation of new offices alter how individuals and teams go about and experience their work (e.g., Laing, Duffy, Jaunzens, et al., 1998; McElroy & Morrow, 2010) and can act as an enabler for wider cultural change (e.g., Turner & Myerson, 1998). The organizational change management literature (e.g., Brown & Eisenhardt, 1997; Burnes, 1996; Kanter, Stein, & Jick, 1992; Kotter, 1996; Luecke, 2003; Mumford, 1983; Pettigrew, 1985; Pettigrew, Woodman, & Cameron, 2001) argues that for such organizational changes to be successful, they need to be managed effectively. To date, however, there has been limited application of existing organizational change theory to this domain (McElroy & Morrow, 2010). Nevertheless, architectural and environmental psychology principles (e.g., Blundell-Jones, Petrescu, & Till, 2005; Vischer, 2005) have emphasized the importance of user involvement and information sharing during the design and implementation of new offices and buildings, as did earlier work informed by socio-technical systems thinking (e.g., Mumford, 1983). Although these principles are similar to the central tenets of general change management theories (e.g., Kanter, Stein, & Jick, 1992), we suggest that the traditional technical nature of office design (being typically led by architects, engineers, or facilities managers) makes it especially comparable to IT-led change programs. A socio-technical approach (e.g., Clegg, 2000; Mumford, 1983) provides a framework which is well suited to the specific problem of managing workspace change, as its emphasis is upon not only user involvement and ownership, but also on finding ways of managing and coping with the competing interests and needs of various stakeholders. Approaches that maximize the involvement of staff and other stakeholders, focus upon the functional and human needs of the office occupants, and are open and transparent, appear more likely to result in successful workspace design than do traditional expert-led push-based approaches to design and change. We return to this issue below.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
220 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
OPPORTUNITIES AND FUTURE DIRECTIONS Within this chapter we have taken a broad approach to the design of office environments, from the benefits and pitfalls of open-plan offices, through the continuing optimization of the office, to issues concerning the management of change. These three areas present distinct opportunities for I/O psychology scholars aiming to contribute to better office design and implementation. In this final section, we outline the opportunities for future research into office design, offer suggestions for theory development, and consider practical and methodological issues. New Opportunities The prevalence and continuing evolution of office working (e.g., Brill, Weidemann, & BOSTI Associates, 2001; Duffy, 1997; Vischer, 1996) points to the potential impact that I/O psychology researchers and practitioners can achieve through offering advice regarding the design and implementation of physical environments. Although a significant body of work on the effects of the introduction of office concepts, such as new IT systems (Clegg, 2000; Mumford, 1983), open-plan offices and adjustments in spatial features (e.g., Brennan, Chugh, & Kline, 2002; Brookes & Kaplan, 1972; May, Oldham, & Rathert, 2005; Oldham, 1988; Sundstrom, Herbert, & Brown, 1982; Sundstrom & Sundstrom, 1986; Sutton & Rafaeli, 1987) has been already amassed, there is now an opportunity for an acceleration of studies that look to guide designers’ and stakeholders’ decision making in selecting and optimizing office design. Given that the prevailing business mindset on office design is that it represents in large part a technical issue (Duffy, 2000), behavioral research is now required to provide users, managers, practitioners, and designers with meaningful data that can be used to help undertake system design, including weighing up the various trade-offs that need to be negotiated (cf. Elsbach & Pratt, 2007). This will involve the generation of further, nuanced research, and the presentation of analyses regarding the contextual, individual, and organizational contingencies that may affect the efficacy of office designs, especially their layout or spatial features. There remains a need for advice and insight concerning the effectiveness of differing types of offices for various groups of staff, with an emphasis upon the nature of the tasks performed and the organizational structures within which they operate. There is an opportunity not only to reflect the changing nature of the office in future research, but also to influence the form that these redesigns take and to promote consideration of the effects on individuals, organizational cultures, and processes. Innovative offices and workplaces are often being designed and optimized without the support of professional architects or designers (Laing, 2006) and this represents a real danger for our discipline, too, as new developments pass by without our effective engagement and impact. Just as in the
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
221
period when open-plan working became prominent, we are at risk of failing to evaluate the impact of evolving office forms or to help guide their design to maximize the positive effects on both staff and organizations. There is a need for further work that explores the interaction between evolving office design, new technology, and changing work patterns. The literature would benefit in particular from research examining the effects of new working practices that may accompany redesigned or highly flexible open-plan office space, such as hot-desking, home or tele-working (e.g., Baruch, 2001; Chapman, Sheehy, Heywood, et al., 1995; Daniels, Lamond, & Standen, 2001; De Croon, Sluiter, Kuijer, et al., 2005; Ng, in press; Vos & van der Voordt, 2001). To date, there has only been limited examination of how the introduction of new or redesigned offices may be successfully managed. As others have recently noted (e.g., McElroy & Morrow, 2010), research that recognizes the potential for workspace to support or initiate change in general is very much in its infancy, with as yet limited mainstream consideration. Research in this area, thus far, has been driven largely by case studies and programs of work that have arisen more often from the architectural or environmental psychology disciplines (e.g., Allen, Bell, Graham, et al., 2004; Turner & Myerson, 1998) than from the traditional I/O psychology literatures. There is clearly a need for more empirical exploration in relation to the management of new or redesigned offices, in order to validate present case study findings, in addition to testing associated propositions more extensively. A further timely extension relating to the design of the physical office environment concerns research to support the design, implementation, and operation of sustainable buildings. The activities of private and public sector organizations generate a significant proportion of world carbon emissions, waste generation, and water usage (Davis & Challenger, 2009). The build and operation of work facilities is an important contributor to an organization’s environmental impact, and there is an increasing awareness of the role that new technologies and improved design may have in improving building performance (e.g., Natsu, 2008; Yudelson, 2009). However, technology or innovative design on its own is unlikely to be able to bring the required environmental gains – gaining an understanding of staff behaviors and needs is also massively important. Wener and Carmalt (2006: 158) have noted that “Some of the oft-cited ecological benefits of green buildings are dependent on the ability to correctly predict user behavior.” Appreciating how individuals respond to different work environments and conditions will be critical in ensuring that new technology or design features are used appropriately, so as to avoid counterproductive behaviors. For example, failing to provide adequate storage facilities for staff may lead to shelving being added after the building is built, obstructing efficient ventilation systems and necessitating less efficient “work-arounds” (e.g., opening external windows and doors) (for further discussion see Wener & Carmalt, 2006). The configuration of offices and other workspaces can affect staff uptake of sustainable activities, for example by making sustainable behaviors more convenient and reducing perceived
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
222 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY behavioral barriers. The location of recycling receptacles is a good illustration of this principle in practice, with the placement of recycling bins having been found to influence recycling rates in academic buildings (Ludwig, Gray, & Rowell, 1998). Currently, there are only limited indicative studies that can help guide designers and organizations in using design to support more sustainable behaviors or improve the efficiency of ecologically inspired work buildings. Exploring and understanding the linkages between design and sustainable behaviors thus represents a major opportunity and priority for future research. Theory Development, and Extensions The literature on workspace design and its impact can be characterized by an absence of a unifying theoretical approach. Theories and frameworks have been drawn from social relations, cognitive psychology, systems thinking, symbolic, and physiological standpoints to investigate relationships between workers and their physical environment (e.g., Altman, 1975; Baum & Paulus, 1987; Becker, 1981; Carnevale, 1992; Cohen, 1980; Cummings, 1978; Davis, 1984; De Croon, Sluiter, J., Kuijer, et al., 2005; Desor, 1972; Duffy, 1997; Elsbach & Pratt, 2007; Ferguson & Weisman, 1986; Festinger, Schachter, & Back, 1950; Geen & Gange, 1977; Oldham, Cummings, & Zhou, 1995; Paciuk, 1990; Paulus, 1980; Schuler, 1980; Steele, 1973; Stokols, Smith, & Prostor, 1975; Sundstrom, Burt, & Kamp, 1980; Sutton & Rafaeli, 1987; Vischer, 1989, 2007). However, as discussed by several previous reviewers (e.g., Baron, 1994; Elsbach & Pratt, 2007; Oldham, Cummings, & Zhou, 1995), none of these approaches has received overwhelming empirical support. Although use of a diverse range of theoretical stances has enabled a broad view to be taken of the topic, it has also meant that there has been a lack of consistency in terms of outcome evaluation (i.e., a range of outcomes have been measured), making it difficult to assess theoretical efficacy and consistency. In effect, the variety of approaches has meant that research attention has been spread relatively thinly. The field requires greater direct empirical testing of competing theories to allow informed and incremental theorization to progress (Oldham & Brass, 1979; Zalesny & Farace, 1987). Previous authors have also noted that it is unlikely that there will be a single mechanism explaining the interaction of workers and their workspace (e.g., Elsbach & Pratt, 2007). The complexity of the physical office and its constituent parts may partly explain this, but we propose that greater effort is required to integrate successful aspects of these competing theories. While we do not necessarily argue for a single meta-theory, for such an exercise would in all probability yield a cumbersome outcome, integration within congruent theoretical approaches would be welcome (cf. Hodgkinson & Healey, 2008; Locke & Latham, 2004). For example, the ability to exert control over one’s environment is explicit within social interference theory (e.g., Baum & Paulus, 1987; Oldham, Cummings, & Zhou, 1995) and the environmental
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
223
comfort model (Vischer, 1989), in addition to being implicit in cognitive theories, such as overload (e.g., Cohen, 1980). Although direct testing of control as a mechanism involved in the interaction of individuals with their environment is still in its infancy (e.g., Huang, Robertson, & Chang, 2004; Lee & Brand, 2005, 2010; O’Neill, 1994), this is an area to be capitalized upon. Indeed, the importance of being able to move and act with freedom and control has been suggested as being intimately related not only to individuals’ well-being, but also to their creativity at work (Csikszentmihalyi, 2003). Becker (1991) argues that an ability to adjust the workspace may be significant in influencing how individuals feel about and behave in all aspects of their work life. Our review has demonstrated that knowledge workers often engage in a variety of tasks during the course of the day (e.g., Becker & Sims, 2001; Craig, 2010) and that the space individuals utilize can vary on a daily, weekly, or monthly basis (e.g., Laing, 2006; Ridgway, Cerulli, Davis, et al., 2008). Unfortunately, to date there has been limited theoretical acknowledgement that worker demands and interaction with workspaces are dynamic (but for a notable exception see Duffy, 1997). Clearly, therefore, this issue warrants greater attention. Such an approach would be in line with the progression occurring within other established areas of organizational theory, not least job design, which have sought to incorporate the dynamic nature of work practices into contemporary models (e.g., Clegg & Spencer, 2007); indeed, activities such as job crafting require temporality to be dealt with explicitly (e.g., Wrzesniewski & Dutton, 2001). It is clear there are opportunities to link areas of theory-building and expertise that are currently treated as separate and distinct domains. Thus, extending the argument above about the job design and job crafting, to date there have been few attempts either theoretically or empirically to examine the extent to which physical spaces and environments shape and influence job designs and the opportunities for job crafting. Hence, although it is clear that physical layouts and proximity to other staff influence patterns of social interaction (Oldham & Brass, 1979; Zalesny & Farace, 1987) and thereby shape the social and relational aspects of work (see Grant & Parker, 2009; Kilduff & Brass, 2010), we need to explore further the constraints that workspaces place on job design and, looking at it in the opposite direction, the ways in which people may craft their jobs to shape and change their environments. Finally, we have made use earlier of a socio-technical systems framework to inform the design of a new building (see page 218 (box 6.2)). We believe this has real merit and potential, both theoretically and as a practical approach. It is clear that the underlying principles of socio-technical design were developed and articulated primarily with a focus on the links between new technologies and the social systems around them (see e.g., Cherns, 1976, 1987; Clegg, 2000; Mumford, 1983; Trist & Bamforth, 1951). But this approach cannot, in our view, remain static. To the best of our knowledge, these principles and ways
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
224 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY of thinking have rarely been used to support the design, implementation, and evaluation of new buildings, new workspaces, and the issues that arise there from, including sustainability and so-called “green issues.” One major theoretical challenge for people working in this area is to apply existing socio-technical principles to this new domain of application and to use these experiences to update and improve the principles. This is entirely consistent with an action research philosophy (e.g., Cassell & Johnson, 2006; Susman & Evered, 1978). Practical and Methodological Considerations for Researchers A number of practical and methodological suggestions can be made to aid researchers in designing studies that are better able to exploit and examine the opportunities and challenges of this field. Analysis of tipping points The literature is rife with examples of where compromises need to be made in the design of offices, for instance between providing a workspace that is open and one that provides too many distractions. We believe that there is an opportunity to explore these trade-offs through looking for tipping points that occur within these relationships. The issue of potential tipping points is not something that has received noticeable attention amongst field studies in the literature. However, identifying specific points of inflexion at which aspects of the physical environment (e.g., the proximity of co-workers, the amount of available meeting space) are likely to produce greater detrimental effects than benefits would be of real value. In addition to advancing understanding of the relative effects of such workspace factors, more meaningful advice and guidance could be offered to designers, managers, and staff who have to resolve competing demands in this area. Evidence from specific areas of the workspace literature, however, indicates that an appreciation of tipping points will require systematic analysis. For example, multiple factors (e.g., job complexity, screening ability, gender, and tenure) have been found to affect reactions to density (Epstein & Karlin, 1975; Fried, Slowik, Ben-David, et al., 2001; Oldham, Kulik, & Stepina, 1991). Understanding the complex nature of tipping points will be a challenge for future research but such inquiry should yield information of both practical and theoretical importance. Adopting quasi-experimental approaches Observations of changes to the physical environment provide researchers with an ideal opportunity to utilize quasi-experimental methodology. Quasiexperiments are similar to traditional experiments in that they involve the study of a change in an independent variable (e.g., the removal of partition walls); however, they occur in field settings, and do not require the experimenter to either directly control the manipulation of the independent variable nor to
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
225
randomly assign participants to treatment groups (for an extensive description and discussion of quasi-experimental methodology see Grant & Wall, 2009). This means that interventions such as the introduction of open-plan working can be studied opportunistically, that is without the researcher necessarily having to control how or to whom it is introduced (for an example of a classic open-plan office quasi-experiment see Oldham & Brass, 1979). To date the use of quasi-experiments has been one of the great strengths of the literature on the design of workspaces, as the technique provides the opportunity to achieve high levels of external validity and strengthen causal inferences (Cook & Campbell, 1979). Indeed, as discussed by Grant and Wall (2009), the Hawthorne experiments can be considered one of the earliest exemplars of the quasi-experimental method in use in this particular context. Quasi-experimental designs have been successfully employed in a number of studies in this area. For instance, Oldham (1988) surveyed three open-plan offices of the same company to examine the effects of change. Occupants of the first moved to a new office which incorporated partitions whereas those of the second moved to a new, lower density office. The third office acted as a nonequivalent control (i.e., where no change occurred). Surveys were administered prior to the office moves and again after occupancy. The quasi-experimental design allowed comparisons to be made between times one and two for all three groups. The findings showed that both the introduction of lower density open-plan workspace and the use of partitions were accompanied by increased perceptions of privacy and environmental satisfaction, together with reduced crowding in office occupants, in comparison with the control group. Workers in the lower density open-plan office also reported increased work satisfaction. An inference of these finding is that the presence of physical screens or a lower density of workers within an open office configuration reduces excessive stimulation from the surrounding environment. Temporal/real-time data collection Research has demonstrated that the nature of tasks and the space that workers utilize to fulfill them vary over time and between individuals (e.g., Becker & Sims, 2001; Craig, 2010). Capturing the temporality of such interactions, and the potentially changing experience, requires techniques that are more sophisticated than those generally employed in the domain of workspace evaluation and employee–environment interaction (e.g., cross-sectional surveys or questionnaires administered months apart). Two related techniques, the Experience Sampling Method (ESM) and Work Sampling Method (WSM) are examples of tools that may suit such purposes (e.g., Ayoko, Ashkanasy, & Jehn, 2010). ESM captures within-person, temporal experiences within natural settings, which is achieved through asking participants to provide information regarding their subjective experience on multiple occasions (often at frequent points each day over a period of time) (Totterdell, 2006). WSM is similar and requires
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
226 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY participants to identify and record the tasks they are involved in at any given point in a similar fashion (e.g., Robinson, 2010). Although diaries and online surveys have often been used to collect data of this kind in the past, Personal Digital Assistants (PDAs) are being recognized as providing advantages to collecting data in this regard. PDAs allow efficiency of data processing, fast input of responses, and portability (Robinson, 2010; Totterdell, 2006). These techniques can be extended to the study of the physical workspace (Ayoko Ashkanasy, & Jehn, 2010), allowing researchers to capture what tasks employees are engaged in, where they are performing them, and the related psychological experience. The collection of such rich real-time data can help inform how knowledge workers use office space in practice and guide the development of new theory and more sophisticated techniques for the optimization of existing office space. Incorporating physiological data Research concerning the evaluation and effects of open-plan offices within field settings has been dominated by perceptual and self-report measurements, with the inherent dangers of common method bias (e.g., Podsakoff, MacKenzie, Lee, et al., 2003; Spector, 1992). The collection of physiological data would allow objective insights to be gained into the effects that an office change, for example, the introduction of more workers, might elicit in individuals (Elsbach & Pratt, 2007). Ayoko, Ashkanasy, and Jehn (2010) suggest electrocardiograph (ECG) and blood pressure monitoring as techniques that researchers might utilize to assess physiological reactions to working in open-plan space. We contend that serum cortisol (a prominent stress hormone) sampling would also yield valuable information with which to appraise such reactions. Collecting data of this kind would enable a more direct integration of findings with related literatures (e.g., occupational stress), and would also provide another source of “hard” data for designers and other stakeholders (cf. Ganster, Fox, & Dwyer, 2001). Moving beyond basic productivity/business outcomes Design and redesign of working space require compromises and trade-offs (Elsbach & Pratt, 2007; Ridgway, Cerulli, Davis, et al., 2008). The above review has shown that the basis upon which to make these decisions is currently weighted toward technical or operational considerations, with data readily available regarding financial implications of pursuing different office strategies (e.g., the financial savings of reducing an office floor plan or minimizing build costs is easily calculable). However, when considering the costs of such changes on human behavior and reactions to redesign, objective evaluations are much harder to calculate due to a paucity of measurement of explicit organizational outcomes in current research. Although self-report evaluations (e.g., individual
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
227
productivity) are typically available (e.g., Leaman & Bordass, 2005), future studies that utilize measurements of time use (e.g., Craig, 2010) or higher level organizational outcomes such as project completion times (Foland, Rowlen, & Watson, 1995) would provide designers and practitioners with more robust data on which to determine the effects of office design on individuals and organizations. Overall, the provision of bottom line indicators would enable I/O psychology researchers to offer a credible argument in favor of design choices that may not be the most financially attractive in the short run, but which deliver longer term human and organizational benefits. Enhancing the precision of our measures through greater cross-disciplinary collaboration A lack of standardization of definition and operationalization, both within the behavioral literature and in relation to standards and practices used in other disciplines (e.g., architecture and facilities management), hampers comparison across studies, thereby limiting generalizability. There is a need for researchers to adopt more closely defined constructs when considering office space, in addition to being aware of measurements and norms commonly used by other disciplines. To illustrate this problem, we can consider studies that have specifically explored office density. Although Net Indoor Area (NIA) is an industry standard for measuring the density of employees in a given office space (being the total internal area of an office building, excluding unusable areas such as stairways, corridors, or entrance halls, divided by the number of occupants), two different conceptualizations – setting density or workspace density – have been generally employed by I/O psychology scholars (Oldham, Cummings, & Zhou, 1995). Furthermore, there have been differences in the measurement of the office space used in the workspace density calculations. For instance, Sutton and Rafaeli (1987) used the dimensions of the whole office to calculate the square footage, while researchers have excluded areas covered by furniture from this calculation (e.g., May, Oldham, & Rathert, 2005). At a broader level, offices are inherently difficult to classify due to the sheer differences in building types, structures, nature of the physical services, and furniture systems, together with the variance that organizational structures and cultures bring to bear on office design. The task of classifying such concepts is undoubtedly more difficult for I/O psychology researchers than for those from more design-led professions and disciplines, whose expertise lie in understanding such physical forms (Veitch, Charles, Farley, et al., 2007). Although it is probably unrealistic to expect researchers to adopt a single classification for office types, future research that seeks to understand differences between traditional enclosed space, open-plan office concepts, and new flexible offices, would benefit from paying reference to the distinctions made by Duffy (1997), Brennan, Chugh, and Kline (2002) and Danielsson and Bodin (2008). These classification systems distinguish between variations
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
228 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY in open-plan concepts; however, Danielsson and Bodin (2008) use a more comprehensive categorization which allows future office concepts to be more precisely defined and studied. As illustrated earlier, their typology incorporates architectural thinking to classify seven office types: cell office, shared room office, small open-plan office, medium-sized open-plan office, large open-plan office, flex office, and combi office. A standardized approach to recognizing, recording, and reporting differing types of office designs will enable researchers to make more stable judgments between and within competing concepts, reducing some of the current inconsistencies. An example of such inconsistency concerns the application of the term open-plan. The use of relatively loose criteria (Brennan, Chugh, & Kline, 2002; Danielsson & Bodin, 2008; Ferguson & Weisman, 1986; Oldham, Cummings, & Zhou, 1995) has resulted in noisy data, with some offices defined as traditional enclosed offices containing sections of open-plan (e.g., Brookes & Kaplan, 1972; Zalesny & Farace, 1987). One way of enabling and encouraging the adoption of more sophisticated and useful typologies will be for I/O psychologists to work together in projects with designers and architects – as with other domains, there is much to be gained from inter-disciplinary working. It is also clear that architects and other designers may have much to gain by working with I/O psychologists. One of the authors, for example, is heavily engaged with a leading global architectural practice which is actively developing what it calls a “people-centered approach to design.” The method integrates the complexities of the organization, people, processes, and technology with the construction and architectural aspects of design by taking a systems view to generate performance and sustainability benefits. The approach includes a flexible framework and a toolkit to support each stage of design. We believe that theory-based practical methods and toolkits developed through such people-centered multidisciplinary working will ultimately provide a real way forward for improving building design.
CONCLUSIONS Our review has shown how the physical environment of the office has developed over the past decades, with the open-plan office becoming and remaining the most popular office design (Brill, Weidemann, & BOSTI Associates, 2001). As information technologies continue to advance, with the growing proportion of knowledge workers within the economy showing no sign of abating (e.g., Davenport, 2005), we can be confident that the evolution of office working is set to continue, throwing up an ever-increasing range of environments in which individuals and groups will work. As in many areas of organizational evolution, there is a real danger that our professional capabilities and offerings will lag behind practice. As I/O psychologists we have a professional duty to understand the complex interactions between employees, their ways of working, and the environments within which they work. We also have a responsibility to try to influence the design of these inter-dependent systems and this will make heavy
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
229
demands of our empirical and theoretical work and of our capability to make it available to the stakeholders involved. But we do believe the opportunities are enormous and we have tried to identify some of the specific ways in which we believe this potential might be realized. Not least amongst these are the needs for more joined-up and systemic approaches to theory building, the development of theory-based practical approaches and toolkits, and the need for multidisciplinary work.
REFERENCES Allen, T., Bell, A., Graham, R., Hardy, B., & Swaffer, F. (2004). Working Without Walls: An Insight into the Transforming Government Workplace. DEGW. Allen, T., & Henn, G.W. (2007). The Organization and Architecture of Innovation: Managing the Flow of Technology. Oxford: Elsevier. Altman, I. (1975). The Environment and Social Behavior: Privacy, Personal Space, Territory, and Crowding. Monterey, CA: Brookes. Armenakis, A.A., & Bedeian, A.G. (1999). Organizational change: A review of theory and research in the 1990s. Journal of Management, 25(3), 293–315. Ayoko, O.B., Ashkanasy, N.M., & Jehn, K.A. (2010). Issues and methods in studying territoriality and the physical workspace. Paper presented at the Annual Meeting of the Academy of Management, Montr´eal, Canada. Baron, R.A. (1994). The physical environment of work settings: Effects on task performance, interpersonal relations, and job satisfaction. In B.M. Straw & L.L. Cummings (Eds), Research in Organizational Behavior (volume 16), (pp. 1–46). Greenwich, CT: JAI Press. Baruch, Y. (2001). The status of research on teleworking and an agenda for future research. International Journal of Management Reviews, 3(2), 113–29. Baum, A., & Paulus, P.B. (1987). Crowding. In D. Stokols & I. Altman (Eds), Handbook of Environmental Psychology (pp. 533–70). New York: Wiley. Becker, F.D. (1981). Workspace: Creating Environments in Organizations. New York: Praeger Publishers. Becker, F.D. (1991). Workplace planning, design, and management. In E.H. Zube & G.T. Moore (Eds), Advances in Environment, Behavior and Design (volume 3), (pp. 115–51). London: Plenum Press. Becker, F.D., & Sims, W. (2001). Offices That Work: Balancing Communication, Flexibility and Cost: Cornell University. Becker, F.D., & Steele, F. (1995). Workplace by Design: Mapping the High-Performance Workscape. San Francisco, CA: Jossey-Bass. Bell, B.F., & Kozlowski, S.W.J. (2002). A typology of virtual teams: Implications for effective leadership. Group and Organization Management, 27, 14–49. Block, L.K., & Stokes, G.S. (1989). Performance and satisfaction in private versus nonprivate work settings. Environment and Behavior, 21, 277–92. Blundell-Jones, P., Petrescu, D., & Till, J. (2005). Archeticture and Participation. London, UK: Spon Press. Blyth, A., & Worthington, J. (2001). Managing the Brief for Better Design. London: Taylor & Francis Group. Boyce, P.R. (1974). User’s assessments of a landscaped office. Journal of Architectural Research, 3, 44–62. Brennan, A., Chugh, J.S., & Kline, T. (2002). Traditional versus open office design: A longitudinal study. Environment and Behavior, 34(3), 279–99.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
230 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Brill, M., Weidemann, S., & BOSTI Associates. (2001). Disproving Widespread Myths about Workplace Design. Jasper, IN: Kimball International. British Council of Offices (2009). Sharp increase in office density reveals today’s changing working environment. http://www.bco.org.uk/news/detail.cfm?rid=118 (accessed 3 November 2009). Brookes, M.J., & Kaplan, A. (1972). The office environment: Space planning and affective behavior. Human Factors and Ergonomics in Manufacturing, 14(5), 373–91. Brown, G., Lawrence, T.B., & Robinson, S.L. (2005). Territoriality in organizations. Academy of Management Review, 30(3), 577–94. Brown, S.L., & Eisenhardt, K.M. (1997). The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42, 1–34. Burnes, B. (1996). No such thing as . . . A “one best way” to manage organizational change. Management Decision, 34(10), 11–8. Burnes, B. (2004). Managing Change: A Strategic Approach to Organizational Dynamics. Harlow, UK: FT/Prentice Hall. Business Week (1978) The trouble with open offices. Business Week (7 August 1978), 88, 84–5. By, R.T. (2005). Organisational change management: A critical review. Journal of Change Management, 5(4), 369–80. Carlopio, J.R. (1996). Construct validity of a physical work environment satisfaction questionnaire. Journal of Occupational Health Psychology, 1(3), 330–44. Carlopio, J.R., & Gardner, D. (1992). Direct and interactive effects of the physical work environment on attitudes. Environment and Behavior, 24(5), 579–601. Carnevale, D.G. (1992). Physical settings of work: A theory of the effects of environmental form. Public Productivity & Management Review, 15(4), 423–36. Cascio, W.F. (2000). Managing a virtual workplace. Academy of Management Executive, 14(3), 81–90. Cassell, C., & Johnson, P. (2006). Action research: Explaining the diversity. Human Relations, 59(6), 783–814. Challenger, R., Clegg, C.W., & Robinson, M.A. (2010). Understanding Crowd Behaviours (volume 1): Practical Guidance and Lessons Identified. London: The Stationary Office. Chapman, A.J., Sheehy, N.P., Heywood, S., Dooley, B., & Collins, S.C. (1995). The organizational implications of teleworking. In C.L. Cooper & I.T. Robertson (Eds), International Review of Industrial and Organizational Psychology (volume 10), (pp. 229–48). Chichester, UK: John Wiley & Sons Ltd. Charles, K.E., & Veitch, J. (2002). Environmental Satisfaction in Open-Plan Environments: 2. Effects of Workstation Size, Partition Height and Windows. Institute for Research in Construction, National Research Council Canada. Cherns, A. (1976). The principles of sociotechnical design. Human Relations, 29(8), 783–92. Cherns, A. (1987). Principles of sociotechnical design revisited. Human Relations, 40(3), 153–61. Clegg, C.W. (2000). Sociotechnical principles for system design. Applied Ergonomics, 31(5), 463–77. Clegg, C.W., Cooch, P., Hornby, P., Maclaren, R., Robson, J.I., Symon, G.J., & Carey, N. (1996). Tools to incorporate some psychological and organizational issues during the development of computer-based systems. Ergonomics, 39(3), 482–511. Clegg, C.W., & Shepherd, C. (2007). The biggest computer programme in the world ever!: Time for a change in mindset? Journal of Information Technology, 22, 212–21.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
231
Clegg, C.W., & Spencer, C. (2007). A circular and dynamic model of the process of job design. Journal of Occupational and Organizational Psychology, 80, 321–339. Clegg, C.W., & Walsh, S. (2004). Change management: Time for a change! European Journal of Work and Organizational Psychology, 13(2), 217–39. Cohen, S. (1980). After effects of stress on human performance and social behavior: A review of research and theory. Psychological Bulletin, 88, 82–108. Cook, T.D., & Campbell, D.T. (1979). Quasi-experimentation: Design and Analysis Issues for Field Settings. Chicago, IL: Rand McNally. Craig, D. (2010). The workplace’s impact on time use and time loss. Paper presented at the Annual Meeting of the Academy of Management, Montr´eal, Canada. Csikszentmihalyi, M. (2003). Good Business: Leadership, Flow, and the Making of Meaning. New York: Penguin Group. Cummings, T.G. (1978). Self-regulating work groups: A socio-technical synthesis. Academy of Management Review, 3(3), 625–34. Daniels, K., Lamond, D., & Standen, P. (2001). Teleworking: Frameworks for organizational research. Journal of Management Studies, 38(8), 1151–85. Danielsson, C.B., & Bodin, L. (2008). Office type in relation to health, well-being, and job satisfaction among employees. Environment and Behavior, 40(5), 636–68. Davenport, T.H. (2005). Thinking for a Living: How to Get Better Performance and Results from Knowledge Workers. Boston, MA: Harvard Business School Press. Davis, M.C., & Challenger, R. (2009). Climate change: Warming to the task. Psychologist, 22(2), 112–4. Davis, M.C., Leach, D.J., & Clegg, C.W. (2010). The physical environment: An evolving topic. Paper presented at the Annual Meeting of the Academy of Management, Montreal. Davis, T.R.V. (1984). The influence of the physical environment in offices. Academy of Management Review, 9(2), 271–83. De Croon, E., Sluiter, J., Kuijer, P.P., & Frings-Dresen, M. (2005). The effect of office concepts on worker health and performance: A systematic review of the literature. Ergonomics, 48(2), 119–34. Desor, J.A. (1972). Toward a psychological theory of crowding. Journal of Personality and Social Psychology, 21, 79–83. Duffy, F. (1997). The New Office. London: Conran Octopus. Duffy, F. (2000). Design and facilities management in a time of change. Facilities, 18, 371–75. Duffy, F., McMahan, J., & Pringle, J. (1999). Offices. In D. Adler (Ed.), Metric Handbook: Planning and Design Data (2nd edn). Oxford: Architectural Press. Elsbach, K.D., & Pratt, M.G. (2007). The physical environment in organizations. Academy of Management Annals, 1, 181–224. Epstein, Y.M., & Karlin, R.A. (1975). Effects of acute experimental crowding. Journal of Applied Social Psychology, 5, 34–53. Evans, G.W., & Johnson, D. (2000). Stress and open-office noise. Journal of Applied Psychology, 85(5), 779–83. Evans, G.W., Johansson, G., & Carrere, S. (1994). Psychosocial factors and the physical environment: Inter-relations in the workplace. In C.L. Cooper, & I.T. Robertson (Eds), International Review of Industrial and Organizational Psychology. New York: John Wiley. Felstead, A., Jewson, N., & Walters, S. (2005). The shifting locations of work: New statistical evidence on the spaces and places of employment. Work, Employment and Society, 19(2), 415–31. Ferguson, G.S., & Weisman, G.D. (1986). Alternative approaches to the assessment of employee satisfaction with the environment. In J.D. Wineman (Ed.), Behavioral Issues in Office Design (pp. 85–108). New York: Van Nostrand Reinhold Company.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
232 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Festinger, L., Schachter, S., & Back, K. (1950). Social Pressure in Informal Groups. Stanford, CA: Stanford University Press. Foland, S.S., Rowlen, S., & Watson, S. (1995). Team space and empowerment: A formula for success. Paper presented at the World Workplace ‘95, IFMA Annual Conference, Miami, FL. Fried, Y. (1990). Workspace characteristics, behavioral interferences, and screening ability as joint predictors of employee reactions: An examination of the intensification approach. Journal of Organizational Behavior, 11(4), 267–80. Fried, Y., Slowik, L.H., Ben-David, H.A., & Tiegs, R.B. (2001). Exploring the relationship between workspace density and employee attitudinal reactions: An integrative model. Journal of Occupational and Organizational Psychology, 74, 359–72. Fritz, M.E.W., Higa, K., & Narasimhan, S. (1995). Toward a telework taxonomy and test for suitability: A synthesis of the literature. Group Decision and Negotiation, 4, 311–34. Gannon, T. (2006). Zaha Hadid: BMW Central Building, Leipzig, Germany. New York: Princeton Architectural Press. Ganster, D.C., Fox, M.L., & Dwyer, D.J. (2001). Explaining employees’ health care costs: A prospective examination of stressful job demands, personal control, and physiological reactivity. Journal of Applied Psychology, 86, 954–64. Geen, R.G., & Gange, J.J. (1977). Drive theory of social facilitation: Twelve years of theory and research. Psychological Bulletin 84(6), 1267–88. Gillen, N.M. (2006). The future workplace, opportunities, realities and myths: A practical approach to creating meaningful environments. In J. Worthington (Ed.), Reinventing the Workplace ( 2nd edn), (pp. 61–78). Oxford: Architectural Press. Glass, D.C., & Singer, J.E. (1972). Urban Stress: Experiments on Noise and Social Stressors. New York: Academic Press. Grant, A.M., & Parker, S.K. (2009). Redesigning work design theories: The rise of relational and proactive perspectives. Academy of Management Annals, 3, 317–75. Grant, A.M., & Wall, T.D. (2009). The neglected science and art of quasiexperimentation: Why-to, when-to, and how-to advice for organizational researchers. Organizational Research Methods, 12, 653–86. Hall, D.J., & Ford, T.Q. (1998). A quality approach to factory design. Industrial Management and Data Systems, 6, 241–5. Haynes, B.P. (2007). Office productivity: A theoretical framework. Journal of Corporate Real Estate, 9(2), 97–110. Hedge, A. (1982). The open-plan office: A systematic investigation of employee reactions to their work environment. Environment and Behavior, 14(5), 519–42. Hendrick, H. (1997). Organizational design and macroergonomics. In G. Salvendy (Ed.), Handbook of Human Factors and Ergonomics (pp. 594–637). New York: John Wiley & Sons. Higgins, J.M., & McAllaster, C. (2004). If you want strategic change, don’t forget to change your cultural artifacts. Journal of Change Management, 4, 63–74. Hodgkinson, G.P., & Healey, M.P. (2008). Cognition in organizations. Annual Review of Psychology, 59, 387–417. Holman, D., Axtell, C., Clegg, C., Pepper, K., Waterson, P., Cantista, I., et al. (2000). Change and innovation in modern manufacturing practices: An expert panel survey of UK Companies. Human Factors and Ergonomics in Manufacturing, 10(2), 121–37. Huang, Y.H., Robertson, M.M., & Chang, K.I. (2004). The role of environmental control on environmental satisfaction, communication, and psychological stress: effects of office ergonomics training. Environment and Behavior, 36(5), 617–37.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
233
Hundert, A.T., & Greenfield, N. (1969). Physical space and organizational behavior: A study of office landscape. Paper presented at the 77th Annual Convention of the American Psychological Association, Washington, D.C. Ives, R.S., & Ferdinands, R. (1974). Working in a landscaped office. Personnel Practice Bulletin, 30, 126–41. Kanter, R.M., Stein, B.A., & Jick, T.D. (1992). The Challenge of Organizational Change. New York, NY: Free Press. Kilduff, M., & Brass, D.J. (2010). Job design: A social network perspective Journal of Organizational Behavior, 31, 309–18. Konar, E., Sundstrom, E., Brady, C., Mandel, D., & Rice, R. (1982). Status demarcation in the office. Environment and Behavior, 14, 561–80. Kotter, J.P. (1996). Leading Change. Boston, MA: Harvard Business School Press. Kupritz, V.W. (1998). Privacy in the work place: The impact of building design. Journal of Environmental Psychology, 18(4), 341–56. Laing, A. (2006). New patterns of work: The design of the office. In J. Worthington (Ed.), Reinventing the Workplace ( 2nd edn). Oxford: Architectural Press. Laing, A., Duffy, F., Jaunzens, D., & Willis, S. (1998). New Environments for Working: The Redesign of Offices and Environmental Systems for New Ways of Working (pp. 29–49). London: Construction Research Communications Ltd. Lawler, E.E.I., & Worley, C.G. (2006). Built to Change: How to Achieve Sustained Organizational Effectiveness San Francisco, CA: Jossey-Bass. Lawson, B. (2004). What Designers Know. Oxford: Architectural Press. Lawson, B.R., & Phiri, M. (2003). The Architectural Healthcare Environment and its Effects on Patient Health Outcomes. London: The Stationery Office. Leaman, A., & Bordass, W. (2005). Productivity in Buildings: The Killer Variables. The Usable Buildings Trust. Lee, S.Y., & Brand, J.L. (2005). Effects of control over office workspace on perceptions of the work environment and work outcomes. Journal of Environmental Psychology, 25(3), 323–33. Lee, S.Y., & Brand, J.L. (2010). Can personal control over the physical environment ease distractions in office workplaces? Ergonomics, 53(3), 324–35. Locke, E.A., & Latham, G.P. (2004). What should we do about motivation theory? Six recommendations for the twenty-first century. Academy of Management Review, 29(3), 388–403. Ludwig, T.D., Gray, T.W., & Rowell, A. (1998). Increasing recycling in academic buildings: A systematic replication. Journal of Applied Behavior Analysis, 31(4), 683–6. Luecke, R. (2003). Managing Change and Transition. Boston, MA: Harvard Business School Press. Markland, M. (1995). The future of the office building. Facilities, 13(3), 15–22. Marquardt, C.J.G., Veitch, J.A., & Charles, K.E. (2002). Environmental Satisfaction with Open-Plan Office Furniture Design and Layout. Institute for Research in Construction, National Research Council Canada. May, D.R., Oldham, G.R., & Rathert, C. (2005). Employee affective and behavioral reactions to the spatial density of physical work environments. Human Resource Management, 44(1), 21–33. McCoy, J.M. (2005). Linking the physical work environment to creative context. Journal of Creative Behavior, 39(3), 169–91. McElroy, J.C., & Morrow, P.C. (2010). Employee reactions to office redesign: A naturally occurring quasi-field experiment in a multi-generational setting. Human Relations, 63(5), 609–39. McGourlay, J., Ridgway, K., Davis, M., Challenger, R., & Clegg, C. (2009). Designing the factory of the future. Paper presented at the Design for Human Performance, a one-day conference organized by Arup, London, UK.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
234 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Mehrabian, A. (1977). A questionnaire measure of individual differences in stimulus screening and associated differences in arousability. Environmental Psychology and Nonverbal Behavior, 1, 89–103. Millward, L.J., Haslam, S.A., & Postmes, T. (2007). Putting employees in their place: The impact of hot desking on organizational and team identification. Organization Science, 18(4), 547–59. Morgan, J.J. (1916). The overcoming of distraction and other resistances. Archives of Psychology, 28, 191–208. Mumford, E. (1983). Designing Secretaries: The Participative Design of a Word Processing System. Manchester, UK: Manchester Business School. Mumford, E. (1995). Effective Systems Design and Requirements Analysis: The ETHICS Approach. Basingstoke, UK: Macmillan. Mumford, E., & Weir, M. (1979). Computer Systems Work Design: The Ethics Method. London: Associated Business Press. Nadin, S.J., Waterson, P.E., & Parker, S.K. (2001). Participation in job redesign: An evaluation of the use of a sociotechnical tool and its impact. Human Factors and Ergonomics in Manufacturing, 11(1), 53–69. Natsu, P. (2008). US Companies embrace green technology [electronic version]. eJournal USA, 13, 4–6. Navai, M., & Veitch, J.A. (2003). Acoustic Satisfaction in Open-Plan Offices: Review and Recommendations. Institute for Research in Construction, National Research Council Canada. Ng, C.F. (in press). Teleworker’s home office: An extension of corporate office? Facilities. O’Neill, M.J. (1994). Work space adjustability, storage, and enclosure as predictors of employee reactions and performance. Environment and Behavior, 26(4), 504–26. Oldham, G.R. (1988). Effects of changes in workspace partitions and spatial density on employee reactions: A quasi-experiment. Journal of Applied Psychology, 73(2), 253–8. Oldham, G.R., & Brass, D.J. (1979). Employee reactions to an open-plan office: A naturally occurring quasi-experiment. Administrative Science Quarterly, 24(2), 267–84. Oldham, G.R., Cummings, A., & Zhou, J. (1995). The spatial conguration of organizations. In G. Ferris (Ed.), Research in Personnel and Human Resources Management (volume 13), (pp. 1–37). Greenwich, USA: JAI Press. Oldham, G.R., Kulik, C.T., & Stepina, L.P. (1991). Physical environments and employee reactions: Effects of stimulus-screening skills and job complexity. Academy of Management Journal, 34(4), 929–38. Oldham, G.R., & Rotchford, N.L. (1983). Relationships between office characteristics and employee reactions: A study of the physical environment. Administrative Science Quarterly, 28(4), 542–56. Osmond, H. (1959). The relationship between architect and psychiatrist. In C. Goshen (Ed.), Psychiatric Architecture (pp. 16–20). Washington, DC: American Psychiatric Association. Paciuk, M. (1990). The role of personal control of the environmental control on office worker perceptions of air quality. Paper presented at the EDRA Annual Conference, Edmond, OK. Parker, S.K., Wall, T.D., & Cordery, J.L. (2001). Future work design research and practice: Towards an elaborated model of work design. Journal of Occupational and Organizational Psychology, 74, 413–40. Paulus, P.B. (1980). Crowding. In P.B. Paulus (Ed.), Psychology of Group Influence (pp. 245–89). Hillsdale, NJ: Erlbaum. Paulus, P.B., Annis, A.B., Seta, J.J., Schkade, J.K. & Matthews, R.W. (1976). Density does affect task performance. Journal of Personality and Social Psychology, 34, 248–53.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
235
Peponis, J., Bafna, S., Bajaj, R., Bromberg, J., Congdon, C., Rashid, M., et al. (2007). Designing space to support knowledge work. Environment and Behavior, 39(6), 815–40. Perham, N., Banbury, S., & Jones, D.M. (2007). Do realistic reverberation levels reduce auditory distraction? Applied Cognitive Psychology, 21(7), 839–47. Pettigrew, A.M. (1985). The Awakening Giant: Continuity and Change in Imperial Chemical Industries. Oxford, UK: Blackwell. Pettigrew, A.M., Woodman, R.W., & Cameron, K.S. (2001). Studying organizational change and development: Challenges for future research. Academy of Management Journal, 44(4), 697–713. Pile, J. (1976). Interiors: Third Book of Offices. Whitney Library of Design, NY: WatsonGuptill. Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., & Podsakoff, N.P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. Preiser, W.F.E. (1994). Built environment evaluation: Conceptual basis, benefits and uses. Journal of Architectural and Planning Research, 11(2), 91–107. Rashid, M., & Zimring, C. (2008). A review of the empirical literature on the relationships between indoor environment and stress in health care and office settings: Problems and prospects of sharing evidence. Environment and Behavior, 40(2), 151–90. Redman, T., Snape, E., & Ashurst, C. (2009). Location, location, location: Does place of work really matter? British Journal of Management, 20(1), 171–81. Ridgway, K., Cerulli, C., Davis, M., Challenger, R., Wiseall, S., Hill, P., et al. (2008). Designing the factory of the future. Paper presented at the International Conference on Applied Human Factors and Ergonomics (AHFEI), Las Vegas, NV. Robinson, M. (2010). An empirical analysis of engineers’ information behaviors. Journal of the American Society for Information Science and Technology, 61(4), 640–58. Roethlisberger, F.J., & Dickson, W.J. (1939). Management and the Worker: An Account of a Research Program Conducted by the Western Electric Company, Hawthorne Works, Chicago. Cambridge, MA: Harvard University Press. Schuler, R.S. (1980). Definition and conceptualization of stress in organizations. Organizational Behavior and Human Performance, 25, 184–215. Smith-Jackson, T.L., & Klein, K.W. (2009). Open-plan offices: Task performance and mental workload. Journal of Environmental Psychology, 29(2), 279–89. Sommer, R. (1959). Studies in personal space. Sociometry, 22, 246–60. Spector, P.E. (1992). A consideration of the validity and meaning of self-report measures of job conditions. In C.L. Cooper & I.T. Robertson (Eds), International Review of Industrial and Organizational Psychology (volume 7), (pp. 123–51). Chichester, UK: John Wiley & Sons Ltd. Steele, F.I. (1973). Physical Settings and Organizational Development. Reading, MA: Addison Wesley. Stokols, D., Smith, T.E., & Prostor, J.J. (1975). Partitioning and perceived crowding in a public space. American Behavioral Scientist, 18, 792–814. Sundstrom, E., Burt, R.E., & Kamp, D. (1980). Privacy at work: Architectural correlates of job satisfaction and job performance. Academy of Management Journal, 23(1), 101–17. Sundstrom, E., Herbert, R.K., & Brown, D.W. (1982). Privacy and communication in an open-plan office: A case study. Environment and Behavior, 14(3), 379–92. Sundstrom, E., & Sundstrom, M.G. (1986). Work Places: The Psychology of the Physical Environment in Offices and Factories. New York, NY: Cambridge University Press.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
236 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Sundstrom, E., Town, J.P., Rice, R.W., & Osborn, D.P. (1994). Office noise, satisfaction, and performance. Environment and Behavior, 26(2), 195–222. Susman, G.I., & Evered, R.D. (1978). An assessment of the scientific merits of action research. Administrative Science Quarterly, 23(4), 582–603. Sutton, R.I., & Rafaeli, A. (1987). Characteristics of work stations as potential occupational stressors. Academy of Management Journal, 30, 260–76. Totterdell, P.A. (2006). Longitudinal research/experience sampling technique. In S. G. Rogelberg (Ed.), Encyclopedia of Industrial/Organizational Psychology. Thousand Oaks: Sage. Trist, E.L., & Bamforth, K.W. (1951). Some social and psychological consequences of the longwall method of coal-getting: An examination of the psychological situation and defences of a work group in relation to the social structure and technological content of the work system. Human Relations, 4(1), 3–38. Turner, G., & Myerson, J. (1998). New Workspace New Culture: Office Design as a Catalyst for Change. Aldershot: Gower Publishing. van Eijnatten, F.M. (1997). Development in socio-technical systems design (STSD). In P.J.D. Drenth, H. Thierry, & C.J. de Wolff (Eds), Handbook of Work and Organizational Psychology. (volume 4) Organizational Psychology (pp. 61–88). Sussex, UK: Lawrence. Van de Ven, A.H., & Poole, M.S. (1995). Explaining development and change in organizations. Academy of Management Review, 20, 510–40. Veitch, J.A., Charles, K.E., Farley, K.M.J., & Newsham, G.R. (2007). A model of satisfaction with open-plan office conditions: Cope field findings. Journal of Environmental Psychology, 27(3), 177–89. Vernon, H.M. (1919). The Influence of Hours of Work and of Ventilation on Output in Tinplate Manufacture. London: HMSO. Vernon, H.M. (1921). Industrial Fatigue and Efficiency. London: Routledge. Vilnai-Yavetz, I., Rafaeli, A., & Yaacov, C.S. (2005). Instrumentality, aesthetics, and symbolism of office design. Environment and Behavior, 37(4), 533–51. Vischer, J.C. (1989). Environmental Quality in Offices. New York: Van Nostrand Reinhold. Vischer, J.C. (1996). Workspace Strategies: Environment as a Tool for Work. New York: Chapman and Hall. Vischer, J.C. (2005). Space Meets Status: Designing Workplace Performance. Oxon: Routledge. Vischer, J.C. (2007). The effects of the physical environment on job performance: Towards a theoretical model of workplace stress. Stress and Health, 23, 175–84. Vos, P., & van der Voordt, T. (2001). Tomorrow’s offices through today’s eyes: Effects of innovation in the working environment. Journal of Corporate Real Estate, 4(1), 48–65. Warr, P.B., Cook, J., & Wall, T.D. (1979). Scales for the measurement of some work attitudes and aspects of psychological well-being. Journal of Occupational Psychology, 52(2), 129–48. Weick, K.E. (1979). The Social Psychology of Organizing (2nd edn). Reading, MA: Addison-Wesley. Wener, R., & Carmalt, H. (2006). Environmental psychology and sustainability in high-rise structures. Technology in Society, 28, 157–67. Woodman, R.W. (1989). Organizational change and development: New arenas for inquiry and action. Journal of Management, 15, 205–28. Wrzesniewski, A., & Dutton, J. (2001). Crafting a job: Employees as active crafters of their work. Academy of Management Review, 26(2), 179–201. Yudelson, J. (2009). Green Building Trends: Europe. Washington, DC: Island Press.
P1: OTA/XYZ P2: ABC c06 JWST046-Hodgkinson
January 21, 2011
13:2
Printer Name: Yet to Come
THE PHYSICAL ENVIRONMENT OF THE OFFICE
237
Zalesny, M.D., & Farace, R.V. (1987). Traditional versus open offices: A comparison of sociotechnical, social relations, and symbolic meaning perspectives. Academy of Management Journal, 30(2), 240–59. Zalesny, M.D., Farace, R.V., & Kurchner-Hawkins, R. (1985). Determinants of employee work perceptions and attitudes: Perceived work environment and organizational level. Environment and Behavior, 17(5), 567–92. Zeitlin, L.R. (1969). A Comparison of Employee Attitudes Towards the Conventional and the Landscaped Office. Organization and Procedures Department, Port of New York Authority.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
Chapter 7 DECEPTION AND APPLICANT FAKING: PUTTING THE PIECES TOGETHER Brian H. Kim Occidental College, Los Angeles, CA 90041, USA From mischaracterizations of past work experiences and educational credentials (Fathi, 2008; Lee, 2007; Lewin, 2007; USA Today, 2008) to cheating on selection tests (O’Connell, 2009; Steinberg, 2002; Xiaofeng & Lie, 2006), the media continually highlights people’s attempts to secure employment by distorting the truth about their qualifications and attributes. Moreover, such cases represent the more outrageous attempts that were “caught” after information could not be verified. For the assessment of knowledge, skills, and abilities (KSAs) during personnel selection, the general prevalence and severity of job applicant faking remains the subject of much debate. Nonetheless, general survey results (Gurchiek, 2008) and direct investigations of actual selection processes reveal that a nontrivial number of applicants fake responses (e.g., Ash, 1987; Donovan, Dwight, & Hurtz, 2003; Griffith, Chmielowski, & Yoshita, 2007; Heron, 1956; Mosel & Cozan, 1952). Manipulating the truth seems to characterize a basic human tendency, rather than an insidious act performed by peculiar individuals (Cronbach, 1950; Ford, 2006; Galas´ınksi, 2000; Goffman, 1959; Levin & Zickar, 2002; Miller & Stiff, 1993; Whitley, 1998), and many societies support the notion that job applicants can and should present themselves in a “favorable light.” Thus, some have suggested that faking can be ignored, based on the untenable assumption that all applicants fake, and fake equally. Others regard faking as an attribute of “substance” that helps people perform their jobs (Kashy & DePaulo, 1996; McCrae & Costa, 1983). Nevertheless, organizations have a clear cause for concern when employees are hired on false premises. Not only might those people lack important KSAs necessary for performing the job, but the denial of jobs to qualified, honest applicants would typically be regarded as unfair (Berry & Sackett, 2009; Lopes & Fletcher, 2004; Morgeson, Campion, Dipboye, et al., 2007; Stokes & Toth, 1996). Thus, selection researchers and International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
240 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY practitioners have compelling reasons to investigate how applicants misrepresent themselves to employers in high-stakes assessments. Industrial and organizational psychology research on faking has primarily addressed these issues: motivational factors/deterrents, detection methods, the impact of faking on criterion-related validities of “content tests” (i.e., tests measuring constructs intended to predict performance outcomes), and statistical corrections for recovering true scores with indices of faking magnitude. Although studies have clarified factors that motivate and enable/hinder response distortion, explanations of the constructs underlying faking remain limited (McFarland & Ryan, 2006; Mueller-Hanson, Heggestadt, & Thornton, 2006). Slow developments in theory may be partly attributable to the tendency for related, but distinct phenomena to be grouped, such as self-deception, unconscious denial, impression management, lying, response styles, and more (Griffith & McDaniel, 2006; Rogers, 1997). From a psychometric perspective, many phenomena introduce systematic error into measurements of applicants’ true scores (Cronbach, 1946, 1950; Dicken, 1959; Jackson & Messick, 1958), but only some justify accusations about applicant “faking” (Morgeson, Campion, Dipboye, et al., 2007; Tett, Anderson, Ho, et al., 2006). This chapter represents an attempt to “clean” the construct domain (as phrased by Organ, 1997) and establish groundwork for a more unified theory of faking. The scope of this chapter does not permit complete coverage of the literature on job applicant faking (henceforth labeled “faking”), so readers are referred to overviews provided by a discussion series initiated by Morgeson, Campion, Dipboye, et al. (2007), an edited book by Griffith and Peterson (2006), and recent articles (e.g., Dilchert, Ones, Viswesvaran, et al., 2006; Levashina & Campion, 2006; Rothstein & Goffin, 2006). Consistent with justifications established in prior discussions (see Griffith & McDaniel, 2006; Levashina & Campion, 2006; Morgeson, Campion, Dipboye, et al., 2007; Murphy, 1993; Snell, Sydell, & Lueke, 1999), I ground faking within theories of deception. This synthesis reveals key gaps in current faking models, but also suggests that multidisciplinary research on deception can help to address those shortcomings.
DECEPTION AND JOB APPLICANT FAKING Based on generally accepted propositions (e.g., Bok, 1978; DePaulo, Kashy, Kirkendol, et al., 1996; Ekman, 1985; Galas´ınski, 2000; Levashina & Campion, 2006; Miller, 1983; Wilson, 1997), deception can be defined as “the intentional process of communicating an inaccurate representation of some person or thing.” Within the context of selection, this definition distinguishes certain response biases (Cronbach, 1946), unconscious denial processes (Sackheim & Gur, 1978; Zerbe & Paulhus, 1987), a lack of self-awareness, tendencies to misinterpret test items, and other reasons for erroneous statements
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
241
from the deliberate attempts to undermine measurements that characterize faking (Morgeson Campion, Dipboye, et al., 2007). The requirement for intentionality also implies that a deceiver must only perceive communicated information to be false or sufficiently unrepresentative of the truth (Frankfurt, 2005; Scott & Jehn, 2003; Whaley, 1982), which precludes the need to establish a philosophical standard for absolute “truth.” I use three formal models of faking to establish a clear basis for conceptualizing this construct as a psychological process of deception. Snell, Sydell, and Lueke (1999) proposed an initial set of antecedent constructs that determine whether a faking attempt will be “successful.” In their model, faking ability mediates the influence of disposition, experience, and situational constraints on faking success, while faking motivation mediates the influence of demographics, disposition, and “perceptual” factors (i.e., social-cognitive judgments and evaluations of potential rewards) on faking success. McFarland and Ryan (2000, 2006) organized similar constructs in a temporal sequence and developed more detailed explanations of faking outcomes. In their model, motivational constructs initiate an intention to fake that interacts with faking ability to produce performance (operationalized as shifts in test scores from honest levels). Early tests of these two models, separately and in combination, have provided some initial support (McFarland & Ryan, 2000; 2006; MuellerHanson, Heggestad, Thornton, et al., 2006; Raymark & Tafero, 2009), and researchers have continued to refine and extend the models (e.g., Levashina & Campion, 2006). The third model, of workplace “dishonesty” (Murphy, 1993: 145), can be used to extend the aforementioned theories in two important ways. Murphy’s model organizes a longer sequence of events to differentiate distal determinants (competition, societal attitudes, and legal penalties) from more proximal determinants (fear of failure, group norms, detection methods). In turn, both sets of determinants drive perceptual processes (the need to be dishonest, acceptability of dishonesty, and risks associated with being caught) and, eventually, the performance of dishonest behaviors. Providing another contribution, this model connects faking performed during selection with other forms of dishonesty performed on the job, suggesting that deception can influence broader organizational outcomes than just selection decisions. Despite their strengths, these models essentially fail to describe what fakers actually do. Without specifying the decisions and actions that produce distorted responses, operational definitions of faking cannot be fully valid or be used to distinguish faking from its outcomes (e.g., composite test scores, hiring decisions, job performance, and organizational effectiveness). However, deception theories and isolated propositions from industrial and organizational psychology provide insights about the cognitive and behavioral components of faking. The broader literature also implies that faking models should account for the influence of deception targets (i.e., hiring personnel in this case). Eventual faking success is determined not only by the quality of deception performed
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
242 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY by applicants, but by an organization’s willingness to deter faking, detect and manage it (e.g., by applying score corrections or removing applicants). Integrating Deception Theories with Faking Models A coherent psychological theory of deception has seemed elusive (Hyman, 1989), but work across disciplines has been encouraging. For instance, most theorists recognize two broad dimensions of deception, labeled here as “commissive” and “omissive” (e.g., Buller & Burgoon, 1994; Ekman & Friesen, 1969; Galas´ınski, 2000; Handel, 1982; Hopper & Bell, 1984; Inbau, Reid, Buckley, et al., 2004; LaFreni`ere, 1988; Lee, 2004; Miller & Stiff, 1993; Scott & Jehn, 2003; Turner, Edgley, & Olmstead, 1975). Commissive deception introduces false information into messages (e.g., test responses), whereas omissive deception suppresses true information from being expressed. By adding or subtracting information, a deceiver can transform a message to be unrepresentative of the truth; both types of deception can be used during the same interaction. For selection assessments with structured response formats (e.g., multiple-choice tests and situational interviews), applicants have little discretion about withholding or concealing certain types of information, so commissive faking may be more salient. By contrast, most “lie detection” methods rely on failures of omission, or “leakage” (e.g., physiological indicators, irregular facial expressions) presumably caused by lying and guilt (Ekman & Friesen, 1969; Ekman, Friesen, & O’Sullivan, 1988; Zuckerman, DePaulo, & Rosenthal, 1981). Faking models are generally consistent with process models of deception. Theories of deceptive warfare (e.g., Daniel & Herbig, 1982; Handel, 1982; Whaley, 1982) and academic dishonesty (e.g., Davis, Grover, Becker, et al., ˜ 1992; Kisamore, Stone, & Jawhar, 2007; McCabe, Butterfield, & Trevino, 2006; McCabe & Trevino, 1997; Murdock & Anderman 2006; Whitley, 1998; Wowra, 2007) also specify motivation (i.e., rewards, punishments, disposition, personal values, and peer influence), abilities to deceive, and situational factors as determinants of behavior. Education studies further reveal that the motivation to cheat on exams is affected by perceptions of the likelihood that one will be caught and the severity of any punishments (McCabe, Butterfield, & ˜ 2006; McCabe & Trevino, ˜ 1993; Vandehey, Diekhoff, & LaBeff, Trevino, 2007; Volpe, Davidson, & Bell, 2008). Furthermore, peer influence tends to overpower other motivational factors (Jensen, Arnett, Feldman, et al., 2002; ˜ 2001; Storch & Storch, 2002; Williams & McCabe, Butterfield, & Trevino, Jaonsik, 2007), as when social groups spread favorable attitudes toward cheating and foster perceptions that cheating is normative. The broader literature contains propositions about deception behaviors, or “tactics,” and even includes some formal taxonomies/typologies (Davis, Grover, Becker, et al., 1992; Godson & Wirtz, 2002; Handel, 1989; McCabe, 2005; Newstead, Franklyn-Stokes, & Armstead, 1996; Whaley, 1982;
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
243
Whiten & Byrne, 1988). Yet, the crafting and deliverance of deceptive messages could involve almost any behavior, so one must specify behaviors in context. One should also distinguish behaviors that must be performed to enact deception – goal-directed decisions and behaviors that alter communications – from cooccurring behaviors (Weiss & Feldman, 2006) and leakage. Most leakage can be regarded as behavioral byproducts of performing deception that people want to minimize and that targets want to detect. Faking researchers, in particular, must also differentiate “impression management (IM)” behaviors (Leary & Kowalski, 1990; Zerbe & Paulhus, 1987) from “impression management tactics.” IM tactics consist of acts of persuasion and influence, some of which can be performed in an honest manner (Bolino & Turnley, 2003; Jones & Pittman, 1982; Levashina & Campion, 2006). Lopes and Fletcher (2004) regarded just one of six IM tactics as “inherently deceptive,” and Weiss and Feldman (2006) found only the “self-enhancement” IM tactic to correlate substantially with admissions of lies told during an interview. Levashina and Campion (2007) proposed a taxonomy of faking behaviors using propositions from research, popular press, and interviews with job applicants. Data supported a four-factor structure, which is compatible with the higher order distinction between commissive and omissive forms. Slight image creation involves exaggerating statements and altering job-irrelevant behaviors/values to seem job-relevant. Extensive image creation includes altering the content and structure of a message to fabricate or “borrow” (i.e., copy others’ responses) positive qualities. Image protection entails omitting, masking, and “distancing” (minimizing connections to) negative qualities. The fourth factor refers to false affirmations of another person’s statements (i.e., ingratiation through dissimulation). Although untested, this taxonomy is complemented by communications research showing that messages can be distorted along various dimensions such as quality, quantity, and relevance (Buller, Burgoon, Buslig, et al., 1996; Jacobs, Dawson, & Brashers, 1996; McCornack, 1997; Yeung, Levine, & Nishiyama, 1999). Because the number of deception behaviors is certain to be intractable and because no one behavior adequately explains complex response patterns, deception can be viewed as a set of strategies. Deception strategies explain how goals guide decisions to perform a coordinated set of behaviors producing a message the target will accept as authentic (Ekman, 1985; Lewicki, 1983; Shulsky, 2002). This chapter reviews a range of faking strategies produced with written modes of communication typical of selection assessments. Some aspects of faking during face-to-face assessments (e.g., employment interviews) are also addressed. Finally, faking models should incorporate the extensive research on lie/deception detection. Communication norms require people to accept most messages as truthful (Galas´ınski, 2000), resulting in a natural “truth” or “honesty” bias (Bond & DePaulo, 2006; Levine, Park, & McCornack, 1999; Taylor, Gittes, O’Neal, et al., 1994). As is the case for individuals (DePaulo & Pfeifer,
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
244 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY 1986; Ekman & O’Sullivan, 1991; Hilton, Fein, & Miller, 1993; Murphy, 1993), organizations that fail to exercise a reasonable degree of suspicion will be unlikely to monitor, detect, and manage faking through formal or informal policies, thereby increasing applicants’ chances of faking successfully. Other organizations may actively deter, detect, and penalize fakers, but with flawed detection methods. In either case, how organizations react (or fail to react) influence whether an attempt to fake will be successful. Figure 7.1 depicts an integrated model of faking. First, motivational factors lead some applicants to develop a discrete intention to fake. Next, an applicant’s capability of faking and contextual factors affect whether certain faking strategies will distort responses effectively, where test responses represent the immediate faking outcome. Distorted responses, as a set, then influence overall test scores and selection decisions, unless organizations intervene and attempt to detect and punish faking. For simplicity, Figure 7.1 depicts only basic construct linkages, but elaborations and refinements may support more complex patterns of relationships that capture the dynamic and often cyclical nature of deception (Buller & Burgoon, 1994). Throughout this chapter, the components in the figure are described and research needs are identified.
THE PROCESS OF FAKING Faking Motivation and Deterrent-based Countermeasures A potential job offer provides the primary motivation for all applicants to perform well on selection assessments. The motivation to fake as a means of performing well, however, depends on additional factors. Applicants’ perceptions of their true scores for the KSAs being assessed determine the need to fake (Leary & Kowalski, 1990; McFarland & Ryan, 2000; Peterson & Griffith, 2006; Tett & Christiansen, 2007), just as a lack of knowledge motivates students to cheat (Davis, Grover, Becker, et al., 1992; Schab, 1991). Also, the least qualified applicants have little to risk because answering honestly and being punished for faking often represent the same outcome, a lost job opportunity. Conversely, top applicants should benefit least from faking (Law, Mobley, & Wong, 2002) and experience the most risk, as faking could shift their scores into an unrealistic and suspicion-arousing range. In reality, applicants probably do not consider their true scores for every instance of faking, but Kuncel and Tellegen (2009) found that some people anchored faked test responses at their true score levels, which reduced the amount of score inflation achieved. According to expectancy theories (e.g., Vroom, 1964), applicants should be motivated by rewards only when they perceive a high probability of obtaining those rewards. Thus, applicants who believe they are incapable of faking well may not be motivated to fake, even with large incentives. Also, reward expectancies in top-down selection settings depend not only on the likelihood
•Perception of normative attitudes & behaviors toward faking
•Personal attitudes toward faking
•Personality
•Values
Figure 7.1
Faking Behavior (communication)
•Inflated Scores
•False Responses
Distortion
Model of faking process. KSA, knowledge, skills, and abilities.
•Test constraints (item presentation, response format, etc.)
•Faking / Cheating Aids
•Faking-relevant skills
•Knowledge of Test & Job
Faking Capability
•Omissive
•Comissive
Faking Strategy
Performance Outcomes
Selection Decisions
Target’s Reaction to Detected Faking
Faking Detected
13:3
Applicant Characteristics
Intention to Fake
•Faking / Cheating Aids
•Test Format & Setting
Contextual Factors
•Suspicion
January 21, 2011
•Need to compete
•Discrepancy between honest and desired KSA levels
Perceived Need to Fake to Obtain Job
Threat of Punishment (severity & likelihood)
Target Characteristics
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson Printer Name: Yet to Come
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
246 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY that an applicant will be perceived as qualified or desirable, but whether that applicant will be perceived as better than competing candidates. Perceptions of a need to compete should drive motivation, particularly as the quality of applicants increases and the selection ratio decreases (Robie, 2006). Figure 7.1 reflects these notions as key motivational determinants leading to faking intentions (Beck & Ajzen, 1991; McFarland & Ryan, 2006). Research also indicates warning people not to fake will reduce faking motivation (Doll, 1971; Kluger & Colella, 1993; Vasilopoulos, Cucina, & McElreath, 2005). In a small-scale meta-analysis, Dwight and Donovan (2003) showed that warnings had a significant, albeit small effect. In a follow-up study, they demonstrated that warnings were most effective when stating both the detection method (to increase punishment expectancy; Rothstein & Goffin, 2006) and ensuing punishment. However, past studies may have underestimated the impact of warnings because test scores are determined by many factors. Warnings represent deterrents intended to reduce faking motivation and intentions, so those constructs should be assessed rather than test scores. Also, no study has compared different kinds/levels of punishment (e.g., score penalties, complete invalidation of test results, forced retesting, or removal from an applicant pool; Burns & Christiansen, 2006; Hough, 1998) and their impact on deterring faking. Related to this, the meaningfulness of a punishment depends on the reward offered, and most studies have offered rewards that were probably too weak (e.g., $20) to be moderated by threats of penalties. Overall, the literature does support the use of punishments, but perhaps only as weak deterrents. Other factors have been proposed, but empirical support is limited. Motivation may increase as peers show more favorable attitudes toward faking and engage in it more often (McFarland & Ryan, 2006; Mueller-Hanson, Heggestad, & Thornton, 2006; Snell, Sydell, & Lueke, 1999). By contrast, organizations might be able to reduce faking by increasing the motivation to respond honestly. For example, “honesty contracts” (Bartram, 2009; Pino & Smith, 2003) might alter faking intentions, perhaps by reinforcing an organizational culture of ethics/integrity (Elangovan & Shapiro, 1998) and attracting applicants who feel they would fit in that culture (cf. Schneider, 1987). Although honesty contracts have not been in research, they resemble academic honor codes, which have received some empirical support (McCabe, Butter˜ 2001). field, & Trevino, Lastly, factors internal to test-takers based on demographics, personal values (e.g., integrity, ethics, or religiosity), and disposition (e.g., personality, propensity to feel guilt, a direct tendency to lie) can influence faking intentions. Current psychological theories offer little support for the notion that people will be inherently gratified by performing acts of deception with no external motivators; the concept of “pathological lying” has a dubious foundation (Dike, Baranoski, & Griffith, 2005; Ford, King, & Hollender, 1988; Henry & Raju, 2006; Newmark, Adityanjee, & Kay, 1999). Instead, faking may be an indicator of the general tendency to act with self-interest and a willingness
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
247
to manipulate others, often referred to as “Machiavellianism” (Christie & Geis, 1970). Snell, Sydell, and Lueke (1999) proposed that a willingness to break rules and to self-monitor might predict faking motivation. Studies have shown that Machiavellianism correlates positively with faking intentions (e.g., Levashina & Campion, 2007; Mueller-Hanson, Heggestad, & Thornton, 2006) and other deviant behaviors (Bennett & Robinson, 2000; Berry, Ones, & Sackett, 2007; Dahling, Whitaker, & Levy, 2009), although Cunningham, Wong, and Barbee (1994) found a moderate negative correlation between Machiavellianism and impression management. However, to the extent that faking is merely used to “cover up” past indiscretions, Machiavellianism would be a correlate, but not a cause, of faking. Regarding positive traits, Law, Mobley, and Wong (2002) found ethics and integrity values to be negatively related to faking although empirical support has been mixed (Mueller-Hanson, Heggestad, & Thornton, 2006; Weiss & Feldman, 2006). Many other proposed factors such as low self-esteem and need for approval have either been under-researched or received little support. Although personality may not cause faking, research indicates that socially desirable responding correlates with conscientiousness and emotional stability (Ones, Viswesvaran, & Reiss, 1996). McLeod and Genereux (2008) found significant relationships between deception use and personality traits, but these relationships varied depending on the type of lie. Faking Capability and Contextual Factors Individuals produce different degrees of score inflation through faking (e.g., McFarland & Ryan, 2000, 2006; Mueller-Hanson, Heggestad, & Thornton, 2006), and some even perform worse than if they had responded honestly (Griffith & McDaniel, 2006). Because these individual differences in faking can be observed even when participants are offered large incentives or are instructed to fake maximally, motivational factors cannot explain all variance in faked scores. Just a few studies have examined “faking ability” (including skill-based conceptualizations) and these studies suffer from tautological operational definitions of ability based on the score inflation outcomes it is supposed to predict (e.g., McFarland & Ryan 2000; Mueller-Hanson, Heggestad, & Thornton, 2006). Research in comparative (e.g., Premack, 1988; Whiten & Byrne, 1988) and developmental psychology (e.g., Banerjee & Yuill, 1999; Broomfield, Robinson, & Robinson, 2002; Vasek, 1986; Wimmer & Perner, 1983) support basic social-cognitive abilities (e.g., theory of mind) as prerequisites for the act of deception. At least in face-to-face communications, an understanding of cultural display rules (Lewis, Stanger, & Sullivan, 1989; Rogers, 1984, Saarni, 1984) and the ability to adhere to communication norms (Buller & Burgoon, 1996; Kashy & DePaulo, 1996) seem to enhance the quality of lies, but these factors reflect developmental rather than individual differences among adults.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
248 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Acting captures a broad skill set that is relevant to deception capability, particularly in face-to-face interactions. Just three published studies have examined this notion, finding that amateur and trained “method” actors produced the same lie detection rates as non-actors (Dawson, 1980; DePaulo & Rosenthal, 1979). Amateur acting by nurses produced stereotyped facial expressions that were not monitored for realism (Vrij, Edward, & Bull, 2001). However, these studies did not allow the deception to be planned and practiced, which might be necessary for acting well. Bond and DePaulo’s (2006) meta-analytic results indicate that lies are more difficult to detect when planned, but the relationship between planning and faking may be curvilinear because too much planning may lead to the creation of esoteric messages (Handel, 1989) or enhance leakage during person-to-person communications (Buller & Burgoon, 1994; DePaulo, Wetzel, Sternglanz, et al., 2003). Practice implies deceivers might accrue skills over time, and studies support a weak relationship between Machiavellianism and deception skill (Zuckerman, DePaulo, & Rosenthal, 1981) suggesting that some people take advantage of more opportunities to deceive others. Also, the high prevalence of deception ˜ 2001; Miller, Shoptaugh, & in academic (McCabe, Butterfield, & Trevino, Parkerson, 2008; Whitley, 1998) and social domains suggests that people have ample opportunities to develop social and communication skills through experience and trial-and-error learning (Buller & Burgoon, 1996; Porter & ten Brinke, 2009; Riggio, Tucker, Throckmorton, 1987). These notions might explain faking in selection retest situations, but Hogan, Barrett, and Hogan (2007) concluded that faking has little impact on personality retests, showing that retest scores improved little, or even decreased. However, their sample consisted only of people who failed the first testing. If people could fake effectively and pass the first test, it is possible that only people with low faking capability comprised the sample. Alternatively, retest applicants may have been motivated to produce similar answers across test administrations if concerned that large discrepancies would lead to accusations of dishonesty and subsequent penalties, thereby reducing the magnitude of faked responses. When deception must be performed by fabricating responses in real-time, self-consciousness, reality-monitoring, and self-control are important (DePaulo, Wetzel, Sternglanz, et al., 2003; McFarland & Ryan, 2001; Rowatt, Cunningham, & Druen, 1998; Vohs, Baumeister, & Chiarocco, 2005), although some studies provide contrary evidence (Riggio, Salinas, & Tucker, 1988; Zuckerman, DePaulo, & Rosenthal, 1981). Imagination and empathy skills might also be helpful for creating a false experience in which to conjure responses (DePaulo, 1992). Riggio, Salinas, and Tucker (1988) and Riggio, Tucker, and Throckmorton (1987) showed that extraversion is linked to higher quality lies, which may be helpful for expressing imagined states. Deceivers might also rely on social competence to decide what kinds of facial expressions and behaviors to perform (Feldman, Tomasian, & Coats, 1999).
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
249
Cognitive ability should improve faking attempts for a number of reasons (Gombos, 2006). Faking often involves recalling and applying prior knowledge about the job (Buller & Burgoon, 1994; Dunnette, McCartney, Carlson, et al., 1962; Godson & Wirtz, 2002; Raymark & Tafero, 2009; Shulsky, 2002; Snell, Sydell, & Lueke, 1999). Highly capable fakers probably consider how responses will be interpreted by the target (organization) and tailor responses such that they appear desirable and authentic (Ekman, 1985; Leary & Kowalski, 1990; Lewicki, 1983; Shulsky, 2002; Walczyk, Roper, Seemann, et al., 2003). To balance these competing goals, Vasilopoulos, Reilly, and Leaman (2005) proposed that greater cognitive effort is required (Vrij Edward, & Bull, 2001), and they found that correlations between content (personality) test scores and cognitive ability scores increased for applicants and students acting like applicants, who were all warned not to fake. Many studies have shown zero-order correlations between cognitive ability and socially desirable responding indices to be near zero (Law, Mobley, & Wong, 2002; Ones, Viswesvaran, & Reiss, 1996), although recent unpublished works have produced evidence of larger relationships (e.g., Clark & Biderman, 2006; Wrensen & Biderman, 2005). Logically, however, applicants cannot always rely on strategies like socially desirable responding and will need to decide which items to distort, and the direction and magnitude of distortion. McCornack (1997) formally proposed that deception involves problemsolving, but only a few (e.g., Heuer, 1982) have recognized that forms of faking can fit neatly into general models of cognitive decision making (Newell & Simon, 1972). Furthermore, that fakers need to engage simultaneously in all of the processes mentioned above implies the use of multi-tasking abilities (Porter & ten Brinke, 2009). Applicants might also able to enhance their capabilities through artificial means. Not all test coaching techniques encourage or teach faking (Sackett, Burris, & Ryan, 1989), but some provide test-takers with very specific rules about how to produce or select responses to enhance overall scores and/or to avoid detection (for examples see Cullen, Sackett, & Lievens, 2006; Storm & Graham, 2000). Generally, coaching seems to help people shift scores in the intended direction (e.g., Alliger & Dwight, 2000; Hurtz & Alliger, 2002; Ramsay, Schmitt, Oswald, et al., 2006; Rogers, Gillis, Bagby, et al., 1991; Sackett, Burris, & Ryan, 1989), but the impact of coaching surely depends on what exactly is taught. More importantly, one might ask: does coaching pose a greater threat than normal faking? Little is understood about whether coaching provides fakers with strategies/heuristics and tactics that they would not use spontaneously or without guidance (Kroger & Turnbull, 1975). Faking Strategies: Cognitions and Behaviors At some point, faking intentions must become behaviors. Behaviors such as voice inflection, speaking style, facial “micro-expressions,” and body
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
250 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY movements provide suggestive “cues,” but it is clear that no single behavior is a perfect indicator of deceit (Ekman, 2001). Of 88 potential behavioral cues to deception examined in DePaulo, Wetzel, Sternglanz, et al.’s (2003) meta-analysis, 24 were statistically significant, but only 6 had substantial effect sizes (Cohen’s d ≥0.35). Moreover, just one seems to represent a tactic enabling deception (i.e., claiming to have forgotten details). Moderator analyses suggested that some leakage behaviors (e.g., length of response, delays in speech, or fidgeting) can be reduced with planning and practice. For paper-and-pencil tests, faking strategies akin to cheating may include a complex sequence of physical behaviors (e.g., surreptitiously looking at another person’s responses), but explanations of how people circle multiple-choice responses or write sentences are less than insightful for understanding applicant faking. Instead, researchers can investigate the strategies that encompass sets of goal-directed decisions, which in turn determine how each response will be distorted in content and/or format (Zickar & Robie, 1999). It seems that all strategies require some assumption of rationality, given the intentional nature of faking. Even so, some strategies are surely easier to perform than others, which could explain variance in faked test scores. By describing faking as “claiming unlikely virtues, denying common faults and unpopular attitudes, exaggerating personal strengths, good impression, [and] self-enhancement,” Ones, Viswesvaran, and Reiss (1996: 660) captured the breadth of strategies that fakers may use, and the literature even includes broader propositions. As noted above, the literature supports a distinction between commissive and omissive strategies. Figure 7.1 includes a strategy component that resides at this level of detail. In Figure 7.1, I propose a way of organizing commissive strategies according to the information that fakers have about how the distorted responses should appear in form and content. Four general classes of strategies are ordered from information-rich to informationpoor: reproduction, rule application, extrapolation, and conditional guessing. In some instances, fakers may know exactly what form distorted responses should take and merely need to reproduce them. In other situations, a person may distort responses along a particular dimension (e.g., social desirability) according to a set of general responding rules. Yet at other times, a person may know only that the true response is undesirable and guess an alternative response. Reproduction The most information-rich strategies involve copying information from another source. People tend to conceptualize “cheating” narrowly to describe the acquisition of objectively correct answers to knowledge tests. Yet, education researchers conceptualize cheating broadly, as a set of tactics that includes actions such as obtaining a scoring key, stealing a test to research the answers, using “crib notes” or other aids, learning the best responses from a
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
251
previous test-taker, or copying another’s responses during the test administration (Davis, Grover, Becker, et al., 1992; McCabe & Trevino, 1993; Newstead, Franklyn-Stokes, & Armstead, 1996). In all cases, cheating enables test-takers to misrepresent themselves to obtain contingent rewards by providing false test responses. Also, cheating can occur in any context with any assessment, as when applicants “borrow” responses from another source to plagiarize a r´esum´e or to recite another person’s experience during an interview (Levashina & Campion, 2007). Cheating is a form of faking. In practice, cheating may be difficult to accomplish and thus occur infrequently, but practice-based research on test security reveals that some applicants do attempt it (e.g., Caveon Test Security, 2006). Although job applicants rarely obtain a test’s scoring key, they may find material to reproduce from general texts, Internet websites (Jobst, Hinshaw, & Harris, 2009), friends who have taken the same test, or coaching programs. O’Connell (2009) reported on individuals who were advised to take a personality test with one company to learn the items and develop fake responses, before taking the same test for the company at which they want the job. These examples highlight that reproduction strategies do not require an assumption that the information source is perfect. This strategy only provides fakers with clear information about what to fake. Rule application When people have less information about how to create faked responses, they can use general responding rules. Perhaps due to coaching or intuition, some applicants apply very technical rules (e.g., select moderate response options on a situational judgment test; Cullen, Sackett, & Lievens, 2006). More typically, applicants seem to rely on socially desirable responding (SDR). SDR on tests involves endorsing items/responses that reflect qualities generally regarded as positive in society (Paulhus, 2002). For many items on typical selection instruments, the socially desirable direction for a response scale is obvious, and using SDR may require little thought (cf. Ones, Viswesvaran, & Reiss, 1996). For items that do not link clearly to a (un)desirable quality, fakers must resort to judgment and decision-making processes to determine the direction and magnitude of distortion. Viswesvaran and Ones’ (1999) meta-analysis revealed that instructions to “fake good” were associated with substantially higher scores on SDR detection indices than instructions to respond honestly, in both between-group and within-group analyses. Others argue that SDR alone fails to capture more complex patterns of faked responses. No published study has directly examined how fakers use strategies, but people can produce “subtler” patterns of faked scores than is implied by SDR theories (Cunningham, Wong, & Barbee, 1994; Doll, 1971; Paulhus, Bruce, & Trapnell, 1995; Schrader & Osburn, 1977). One parsimonious explanation is that fakers use additional rules that moderate SDR (Clark &
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
252 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Biderman, 2006), such as: (i) avoid extreme responses, (ii) periodically answer an item in the socially undesirable manner, or (iii) check that all responses are plausible. Other rules can also be used to minimize leakage, such as “be consistent” and provide similar responses to similar items. Alternatively, applicants may use a more sophisticated rule instead of SDR. Job desirable responding (JDR) involves identifying and distorting (to a greater degree) only those items with apparent relevance to the job (Kluger & Colella, 1993; Levashina & Campion, 2006; Schmit & Ryan, 1993; Vasilopoulos, Reilly, & Leaman, 2000). To be effective, this strategy requires fakers to possess knowledge or stereotypes of the organization, occupation, or specific job that can be used to produce realistic and favorable responses (Kilcullen, White, Mumford, et al., 1995; Schmit & Ryan, 1993). JDR may be especially useful as a way to avoid detection. Instead of randomly downplaying some items to produce nonextreme test scores, JDR targets items that seem less important for obtaining the job. Extrapolation Borislow (1958) expanded thinking on SDR by instructing people to “respond as they believed a ‘perfect individual characterized by those traits that society considers highly desirable’.” By focusing on an actual or hypothetical person rather than traits measured by the test or specific job characteristics, fakers might be able to create a realistic profile of scores that is relatively indistinguishable from the profile of an ideal applicant or employee role (Abrahams, Neumann, & Githens, 1971; Dicken, 1959; Elliott, Lawty, & Jones, 1996; Furnham, 1990; Jackson, Wroblewski, & Ashton, 2000; Martin, Bowen, & Hunt, 2002). Also, this “prototype-matching” strategy (Leary & Kowlaski, 1990) can be used with any kind of ideal figure or set of role expectations, not just the socially desirable person that Borislow mentioned. When test items do not reveal an obvious match with the prototype, applicants can use existing knowledge of the prototype person and stereotypes to imagine how he or she would respond. As such, this strategy allows people to extrapolate new responses in information-poor conditions, when little is known about the job or organization. Unlike JDR, extrapolation may be especially helpful for faking tests that contain little job-relevant content, like personality tests. Martin, Bowen, and Hunt (2002) found that importance ratings for 29 of the 30 attributes of the Occupational Personality Questionnaire were similar between actual human resource managers and university students who only read a job description for a junior-level manager. Kroger and Turnbull (1975), however, showed that people were unable to recreate a realistic profile of an artist on the Minnesota Multiphasic Personality Inventory (MMPI) when they had a weak conception of “artist.” Overall, the research reveals little about this strategy beyond what can be hypothesized logically. It is not clear whether applicants use this strategy without prompting.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
253
Conditional guessing When applicants lack information about how responses should be altered, they probably resort to guessing strategies (Attali & Bar-Hillel, 2003; Furnham, 1990). This strategy describes situations when people can eliminate some potential responses from consideration, including the honest one, prior to guessing. (Completely random responding precludes attributions of faking because such behaviors are inconsistent with intentions to explicitly avoid the truth.) In addition, conditional guessing offers an explanation of how trialand-error learning can occur when people fake the same messages over time, as might be the case in multi-round interviews, multiple hurdle processes, or basic retesting situations. In the end, fakers may use more than one strategy as the situation allows or requires; the categories are not mutually exclusive. Some strategies capitalize on knowledge about the ideal test responses, while others involve problemsolving processes or information-poor guesses. These commissive strategies should also help to explain selection assessments requiring face-to-face interactions or voice/video recorded responses (e.g., Jones, Brasher, & Huff, 2002), although greater attention might be placed on ommissive strategies that suppress true information and conceal the fact that faking is being performed. Military research (e.g., Handel, 1982, 1989) offers an interesting expansion of faking theory that recognizes deception can be performed by groups. On a large scale, test developers have recognized attempts by organized groups to “harvest” items from repeated test administrations to recreate a test bank, by having a large number of test-takers each memorize one or two test questions (Burke, 2009; Jobst, Hinshaw, & Harris, 2009). For a specific applicant pool, practitioners need not be concerned with groups of applicants working together because, if for no other reason, applicants must compete against each other for positions to some degree. It may be premature to build theories around such behaviors, but faking models should be flexible enough to accommodate group-level phenomena.
Opportunities to Fake and Situational Constraints Even when sufficiently motivated and capable, situational factors can limit or enhance a faking applicant’s likelihood success (Garry, 1953). Organizations undeniably design test administrations to prevent general cheating (e.g., distancing applicants to prevent the sharing of responses). Other countermeasures prevent people from deriving information (e.g., constructs measured, scoring method) from the test itself to determine the “best” answers. Yet, as the previous section implies, countermeasures only be effective in so far as they address particular types of faking (Alliger & Dwight, 2001). Test developers may attempt to minimize test transparency by randomizing the order of items (McFarland, Ryan, & Ellis, 2002) or using indirect
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
254 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY scoring methods, such as empirical versus rational keying or option versus item keying (Kilcullen, White, Mumford, et al., 1995; Klein & Owens 1965; Kluger, Reilly, & Russell, 1991). Generally, studies have supported these investigations (Hough, 1998). Yet, Stokes, Hogan, and Snell (1993) found evidence contradicting Kluger, Reilly, and Russell’s (1991) finding that optionkeying was associated with less SDR. Until more is understood, the debate about whether such techniques are worth implementing will surely continue (Morgeson, Campion, Dipboye, et al., 2007; Ones, Dilchert, Viswesvaran, et al., 2007). Other individuals have exploited inherent properties of test items to reduce faking by: (i) discarding items for which the most socially/job desirable response was obvious, and (ii) increasing the proportion of test items that solicit objective and verifiable responses (Becker & Colquitt, 1992; Harold, McFarland, & Weekley, 2006; Kluger & Colella, 1993; McManus & Masztal, 1999). Some studies have supported the null hypothesis (e.g., McManus & Masztal, 1999) or indicated that all types of items can be faked (Graham, McDaniel, Douglas, et al., 2002), but results suggest that faking motivation can be reduced with this approach, although this reduction only has indirect effects on test scores. Also, this approach may not work unless applicants actually think that organizations will attempt to verify the responses in some manner. An extension of this approach uses a hybrid response format of multiplechoice and open-ended questions to solicit elaboration of responses (Schmitt & Kunce, 2002). Test-takers select a multiple-choice option and then proceed to write details in support of that option. Elaboration thus provides a weak form of corroboration between the multiple-choice and written answer, and provides a better opportunity for organizations to verify detailed responses. Initial studies have shown that elaboration requirements do lower test scores considerably (Schmitt & Kunce, 2002; Schmitt, Oswald, Kim, et al., 2003) and Lievens and Peeters (2008) recently connected this effect directly to differences in faking. Forced-choice response formats make it virtually impossible for a person to only select desirable responses because response alternatives are matched for desirability (Hirsch & Peterson, 2008; Pankratz & Binder, 1997). On the whole, investigations of forced-choice tests offer mixed results (see reviews by Berry, Sackett, & Wiemann, 2007; Rothstein & Goffin, 2006). Some studies support claims that forced-choice tests reduce score inflation due to faking and help to preserve an assessment’s criterion-related validity (Vasilopoulos, Cucina, Dynomina, et al., 2006; Jackson, Arnett, Feldman, et al., 2002; Martin, Bowen, & Hunt, 2002), while others indicate that the format has little effect on faked scores (Heggestad, Morrison, Reeve, et al., 2006). Because forced-choice tests are known to limit the psychometric quality of an assessment (Hicks, 1970; McCloy, Heggestad, & Reeve, 2005; Meade, 2004; Ones, Dilchert, Viswesvaran, et al., 2007), test developers actually tend to use partially ipsative tests that ensure normative scores can be derived from responses. Thus, the degree of ipsativity may moderate a test’s resistance to social
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
255
desirability, although Hirsch and Peterson (2008) found that three ipsative variants of a popular personality test (i.e., International Personality Item Pool; Goldberg, 1999) produced similar scores. Additional approaches manipulate response formats. As compared to the yes–no questions used with most lie detection instruments, multiple-choice or open-ended responses probably require fakers to exert more cognitive effort (Khorramdel & Kubinger, 2006; Van Iddekinge, Raymark, & Roth, 2005). At the same time, such responses might also be more difficult to verify, and accurately detecting a false response reveals little about the true response (Rodin, 1972). Implicit association tests (IATs) should be resistant to faking because they measure manifestations of the unconscious via response time (Greenwald, Nosek, & Banaji, 2003). For a personality test in IAT form, however, participants were able to distort their scores, but particularly for extraversion (Steffens, 2004). However, Greenwald (2008), creator of the IAT, responds to challenges that the IAT can be faked with this statement on his public website: “Clearly, the best strategy for faking the IAT (although few research subjects discover it on their own) is to respond slowly in whichever of the two combined tasks is easier.” Because the IAT can be faked, this begs the question: Does the IAT measure unconscious processes if people can intentionally distort responses? Conditional reasoning tests might limit response distortion by measuring constructs indirectly (LeBreton, Barksdale, Robin, et al., 2007), but a stronger empirical base must be established. Instead of masking the purpose of items, test developers might lure fakers into endorsing responses that are ostensibly desirable and face-valid, but actually keyed as undesirable or neutral, respectively (Dannenbaum & Lanyon, 1993), an approach also used for faking detection scales. Although sensible, this approach has not been clearly supported (Bagby, Nicholson, & Buis, 1998; Osberg & Harrigan, 1999) and might be feasible only with empirically based scoring systems. In these instances, researchers have also questioned whether using deception to detect fakers blurs ethical boundaries, or would at least draw legal attention. Contrary approaches could use direct instructions to specify the kinds of responses that will be accepted as legitimate. For instance, a personality item might ask about behaviors performed in the last month rather than “in the past,” providing less room for a person to interpret the instructions loosely and misrepresent the truth. Supporting this logic, situational judgment tests exhibit less response distortion when asking people about their understanding of different behaviors instead of their own behavioral tendencies (Nguyen, Biderman, & McDaniel, 2005). From a faker’s perspective, technology provides new ways to distort responses and to avoid detection (Naglieri, Drasgow, Schmit, et al., 2004; Tippins, 2009; Tippins, Beaty, Drasgow, et al., 2006). First, devices increasingly allow covert communications (e.g., cellular phones, radio transmitters) between test-takers or with sources outside of the testing locale and on the Internet (e.g., Farmer, 2007; Sommerville, 2009; Stephens, Young, & Calabrese,
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
256 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY 2007). For tests of declarative knowledge, applicants might also store information on permitted electronic devices (e.g., calculators). Second, people might “hack” into computer systems and access test banks, especially when tests are administered over the Internet. Third, the increasing use of unproctored tests may allow fakers to have impostors take their assessments, or allow cheating during the test. Research on these issues is still emerging, but early studies indicate that thoughtful test developers can address these issues to preserve test standards (e.g., Arthur, Glaze, Villado, et al., 2009; Bartram & Brown, 2004; Booth-Kewley, Edwards, & Rosenfeld, 1992; Grijalva, Nowell, & Kerkvliet, 2006; Hancock & Flowers, 2001; Nye, Dragow, & Fine, 2008). Fortunately, organizations may also find ways to use technology to administer tests more securely and to monitor applicant behaviors during testing (e.g., video cameras in place of proctors). A final caveat for limiting faking opportunities is the need to maintain a test’s basic psychometric quality (McFarland, Ryan, & Ellis, 2002). By limiting test transparency, requiring elaboration, or requiring forced-choice decisions, assessments may require even honest respondents to engage in more complex response processes, thus altering the nature of the test (Morgeson, Campion, Dipboye, et al., 2007; Vasilopoulos, Cucina, & McElreath, 2005). Innovations must be supported by evidence of psychometric quality and construct validity, in addition to demonstrations that they effectively reduce the influence of faking (Ones, Dilchert, Viswesvaran, et al., 2007; Robson, Jones, & Abraham, 2008; Vasilopoulos, Cucina, & McElreath, 2005). Additionally, applicants may develop negative reactions (Converse, Oswald, Imus, et al., 2008) to certain methods that seem invasive or lacking face validity, among other reasons.
PROXIMAL OUTCOMES OF APPLICANT FAKING Distorted responses and corresponding shifts in overall test scores away from honest levels represent the immediate outcomes of faking. Thus, faking can be regarded as a source of bias in measurement that necessarily reduces one aspect of construct validity (Komar, Brown, Komar, et al., 2008; Stark, Chernyshenko, Chan, et al., 2001). When successful, faked test scores should increase an applicant’s chance of being selected, but this more distal outcome depends on a number of factors, including other applicants’ honest and faked scores and how the organization responds to faking. Ultimately, practitioners should be concerned about whether the hiring of faking applicants affects organizational functioning, which has been indexed with estimates of criterion-related validity. Faking and Selection Test Score Shifts One can safely assume that high faking motivation and stronger intentions usually lead to “more” faking (in frequency of attempts and the magnitude of
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
257
distortion) and “better” test scores. Edwards (1953: 90) wrote, “the probability of endorsement of personality items is a monotonic increasing function of the scaled social desirability of the items.” Yet, as Levin and Zickar (2002: 261) point out that “there is no reward for increasing rank order incrementally, yet there is a substantial reward for increasing rank order in a quantum leap to the top of the pool.” Thus, one might expect to see evidence of nonlinear, stepwise relationships. While true in many situations, this second notion might fail to describe other situations. For assessments used to select-out applicants on the basis of cutoff scores (e.g., in multiple hurdle processes), low-level applicants only need to fake enough to “make the cut.” Therefore, faking may be expected to create all kinds of score shifts, ranging from small to moderate. Researchers must be mindful of the entire faking process when explaining score inflation effects (Paulhus, Bruce, & Trapnell, 1995; Raymark & Talfero, 2009). First and foremost, one must establish that applicants possessed the intent to fake, or else risk attributing inflation to irrelevant phenomena. Then, one must consider that not all people: 1. Have a need to fake; 2. Choose to fake (excluding studies with explicitly instructions to fake); 3. Aim to produce extreme distortions (when avoiding detection or using sophisticated strategies); 4. Are capable of faking well; and 5. Use strategies that will be effective in a given context. Unless all of these conditions are met, there is little reason to suspect extreme amounts of score inflation, though meeting these requirements partially can produce complex patterns (Komar, Brown, Komar, et al., 2008). In any case, it is clear that applicants and research participants instructed to act like applicants tend to produce “better” scores than other groups operationally defined as being more honest. In their meta-analysis of groups instructed to “fake good,” Viswesvaran and Ones (1999) found that between-subjects designs produced effect sizes of over half a standard deviation (d ∼ 0.6) for personality traits and about 1 standard deviation for SDR scales, compared with more “honest” groups. For studies using within-subjects designs, participants raised scores from honest conditions even more (d from 0.47 to 0.93 for personality and 2.26 for SDR scales). In a meta-analysis of applicant versus non-applicant groups, Birkeland, Manson, Kisamore, et al. (2006) found smaller effect sizes, but still found scores on personality dimensions to be higher for applicants than nonapplicants. One obvious explanation for the smaller effect sizes might be that laboratory-induced faking creates too much motivation to fake, often without a counteracting motivation to avoid detection (except in studies of warnings).
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
258 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Faking and Factor Structures Faking could alter the factor structure of content tests if enough applicants fake enough scale items to influence the relevant group-level statistics (Ellingson, Smith, & Sackett, 2001). Hypothesizing an alternate factor structure for fakers (i.e., people possessing clear intent to deceive) requires careful consideration of the complex nature of faking. Complete adherence to SDR or JDR could obliterate a normal factor structure if respondents endorse every item referring to a desirable action/quality and only those items. Yet, such a pattern could only be observed if test-takers are able to judge most test items along a desirability dimension. Alternatively, one might hypothesize that fakers using SDR or other rules (e.g., avoiding verifiable items) more loosely will create new factors, as would respondents using a new frame of reference (Lievens, De Corte, & Schollaert, 2008; Smith, Hanges, & Dickson, 2001; Stokes, Hogan, & Snell, 1993). Sophisticated strategies like prototype-matching, however, might leave factor structures relatively intact when fakers attempt to mimic a set of natural responses. The research evidence is mixed. While a five-factor personality model fit responses from a student sample, Schmit and Ryan (1993) concluded that an additional factor to index faking was needed to explain responses of job applicants. Ellingson, Sackett, and Hough (1999) demonstrated that intercorrelations between all 10 of the ABLE personality scales increased due to faking, indicating a unidimensional pattern. Other studies revealed altered factor structures with faking (e.g., Hirsch & Peterson, 2008; Vasilopoulos, Cucina, & McElreath, 2005; Zickar & Robie, 1999). Opposing research has revealed stability in factor structures and even evidence of measurement invariance between faking and honest responders (e.g., Ellingson, Smith, & Sackett, 2001; Marshall, De Fruyt, Rolland, et al., 2005; Smith & Ellingson, 2002; Smith, Hanges, & Dickson, 2001), though applicants (i.e., assumed fakers) tended to produce data that deviated from multivariate normality more than “honest” data. Overall, the literature has not proposed clear moderators of these research findings based on differences in study design or analyses, leaving the results open to interpretation.
Faking and Rank-ordering People who inflate scores by even a moderate degree can displace honest candidates at the top of a pool (Dwight & Donovan, 2003). Considering the entire faking process (Figure 7.1), most fakers should be able to improve their rankorder above more honest applicants. Skilled fakers should be able to rise above both honest and unskilled faking applicants. With student samples acting as applicants, Lievens and Peeters (2008) demonstrated that the percentage of fakers who were selected from the applicant pool increased as the selection ratio decreased; at a selection ratio of 0.2, 85% of those selected were fakers. Most
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
259
studies of rank-order reveal that fakers may comprise a large proportion of the selected applicants, and that the effect increases as the selection ratio decreases (Christiansen, Goffin, Johnston, et al., 1994; Griffith, Chmielowski, & Yoshita, 2007; Rosse, Stecher, Miller, et al., 1998; Winkelspecht, Lewis, & Thomas, 2006; Zickar, 2000). Moreover, Mueller-Hanson, Heggestad, and Thornton’s (2003) data supported a similar trend, but also suggest that selected applicants who were honest would perform at higher levels than selected fakers. Alternatively, Berry and Sackett’s (2009) simulation study suggests that cut-off scores specifically designed to select-in only honest individuals – to maintain fairness – may lead to lower levels of performance versus allowing more fakers to be selected. However, laboratory studies surely overestimate rank-order shifts because they manipulate too many people to fake (including those who would be honest as real applicants), though this notion pertains to the frequency of faking more than the degree of distortions to any given response. With statistical simulations, Converse, Peterson, and Griffith (2009) showed that under typical conditions, faking usually has a minimal impact on which candidates would be selected, regardless of whether some applicants fake. With an innovative design, Ellingson, Sackett, and Connelly (2007) compared scores for individuals completing a personality test (California Psychological Inventory) as actual applicants and participants in developmental programs. The authors observed small and unsystematic rank-order shifts and attributed them to measurement unreliability. Though suggestive that real applicants do not or cannot alter their rankings, the authors acknowledge that the results might generalize only to similar tests. Additionally, the results were based on correlational measures of rank-order change because the data represented a kind of virtual applicant pool containing participants from various test settings/times, which might have been affected by missing data (i.e., applicants in the real pool who did not have developmental test scores) or self-selection bias (i.e., fakers may be less inclined to develop themselves through work programs). Finally, Christiansen, Goffin, Johnston, et al.’s (1994) analysis of real applicants also revealed an interesting curvilinear effect, whereby the lowest selection ratios resulted in the selection of fewer fakers than slightly higher selection ratios, which hints that fakers might avoid detection by aiming for sub-maximal scores. Faking and Criterion-Related Validity If people respond dishonestly and thereby reduce a test’s construct validity, it follows logically that the test’s predictive power will be reduced. This notion has received mixed empirical support (Morgeson, Campion, Dipboye, et al., 2007), and researchers on both sides of the debate continually cite the “usual suspects” in support of their claims. Studies have confirmed that faking can decrease criterion-related validities, but few of them were conducted with actual applicants during selection (e.g., Anderson, Warner, & Spencer, 1984;
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
260 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Dunnette, McCartney, Carlson, et al., 1962; Harold, McFarland, & Weekley, 2006; Jackson, Wroblewski, & Ashton, 2000; Mueller-Hanson, Heggestad, & Thornton, 2003; Peeters & Lievens, 2005; Schmit, Ryan, Stierwalt, et al., 1995; Topping & O’Gorman, 1997). On the other side, numerous works (e.g., Barrick & Mount, 1996; Becker & Colquitt, 1992; Christiansen, Goffin, Johnston, et al., 1994; Ellingson, Sackett, & Connelly, 2007; Hough, 1998; Hough, Eaton, Dunnette, et al., 1990; Rosse, Stecher, Miller, et al., 1998), including meta-analyses (Li & Bagger, 2006; Ones & Viswesvaran, 1998; Ones, Viswesvaran, & Schmidt, 1993) and statistical simulations (Schmitt & Oswald, 2006), imply that faking has a negligible impact on criterion-related validities. Yet, Hough (1998: 211) points out that “If validities remained the same regardless of flagrant distortion, one should wonder what, in fact, was being measured.” Given evidence that individuals can respond dishonestly, inflate scores, and sometimes alter their rank-order, the burden of explanation must fall upon those concluding the null hypothesis (i.e., no relationship between faking and criterion-related validity). To date, a range of sensible explanations have been proposed to explain these findings, with some focusing on contextual boundary conditions while others attribute null results to methodological limitations. One plausible explanation emerges from work presented by Peterson and Griffith (2006) by accounting for the range-restriction typical of faking datasets. As indicated by Figure 7.2, fakers (shaded oval) exhibit higher observed scores on a selection test than honest applicants (hollow oval). Assuming that the mere act of faking does not improve job performance (Viswesvaran, Ones, & Hough, 2001), fakers will be normally distributed on job performance measures. Whether using predictive or concurrent validity designs (Hough, 1998), researchers can only obtain performance data for those applicants who Selection Cutoff Score
High
Job Performance
Acceptable Level
Low Low
High Observed Selection Test Score
Figure 7.2 Range restriction effects on estimates of criterion-related validities for honest and faking applicants.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
261
were selected (portion to the right of the vertical dashed line) and who retained employment by performing above a minimally acceptable level (portion above the horizontal dashed line). Thus, analyses of range-restricted samples only within this region will produce attenuated correlations for both groups, and both correlations will appear spuriously similar. By contrast, laboratory studies that can provide criterion data for all participants should be more likely to produce unattenuated validity estimates that differ between honest and faking groups. A second explanation involves the insensitivity of correlational statistics to detect changes in rank-order in one end of a distribution (Rosse, Stecher, Miller, et al., 1998), which suggests that criterion-related validities cannot be used to understand the performance of fakers specifically, separate from the broader set of selected applicants. A third, related explanation by Marcus (2006) claims that the variance in faking must be sufficiently large before criterion-related validities will decrease due to faking. Fourth, a statistical simulation by Komar, Brown, Komar, et al. (2008) revealed that many factors interact with each other, with the correlation between faking and criterion measures (across three levels: –0.20, 0 and +0.20) having the strongest and most numerous effects. Some studies (e.g., Viswesvaran, Ones, & Hough, 2001; Hough, 2001) suggest that faking and performance might be uncorrelated in many cases. Due to the complex nature of their findings, the reader is referred to Komar, Brown, Komar, et al.’s (2008) work for more details about how faking variability, distortion magnitude, proportion of fakers in a pool, faking-predictor correlation, faking-criterion correlation, and selection ratio can impact validity. In sum, the authors concluded that some combinations had a substantial impact while others decreased validity to a small degree, or even raised it. Converse, Peterson, and Griffith (2009) corroborated some findings in a similar simulation, but expanded the work to show that faking tends to have a smaller impact when multiple predictors are used. Another set of explanations identify reasons why faking should not impact validity due to test properties or boundary conditions imposed by the situation. First, the tests believed to be most susceptible to faking (e.g., personality) typically produce small to moderate criterion-related validities (Morgeson, Campion, Dipboye, et al., 2007). Irrespective of judgments about the practical significance of those validities (refer to Ones, Dilchert, Viswesvaran, et al., 2007; Tett & Christiansen, 2007), the percent of variance explained in performance criteria is rarely large enough for faking to hinder prediction. Second, faking may represent qualities of “substance” that help people perform well on the job (McCrae & Costa, 1983; Sackett & Wanek, 1996), although this claim depends on how one defines faking and selects performance criteria. The proportion of faking to honest applicants determines whether fakers can even have an impact on the criterion-related validities, particularly for between-groups analyses of faking impact. Assuming that fakers effectively produce high test scores, research supports the logical expectation that lower
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
262 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY selection ratios will lead to the selection of a higher proportion of fakers (Griffith, Chmielowski, & Yoshita, 2007; Lievens & Peeters, 2008; MuellerHanson, Heggestad, & Thornton, 2003), though simulations might reveal instances when the selection ratio is negatively related with validity (Converse, Peterson, & Griffith, 2009). Furthermore, low selection ratios tend to be associated with higher positions of greater organizational power and influence. Thus, it may be appropriate to weigh any decrements in performance outcomes due to faking more heavily in these jobs, to determine the impact of faking on organizational effectiveness. Organizational Reactions to Faking After people fake, tests scores will be subject to interpretation by hiring personnel or testing services that attempt to detect faking. At least some organizations exercise skepticism about the truthfulness of information self-reported by applicants (Robie, Tuzinski, & Bly, 2006), as evidenced by the use of background and reference checks (Gurchiek, 2008; Jobst, Hinshaw, & Harris, 2009) and response bias indicators (Goffin & Christiansen, 2003). Beyond that, no study has investigated characteristics that predispose some organizations to perceive faking as a problem and monitor its use. Organizations aiming to detect faking can use a variety of approaches, many of which have a long historical basis (Alder, 2007; Trovillo, 1939) – beginning with early psychological methods for detecting lie-based anxiety (Benussi, 1914; Lombroso 1895; Marston, 1917; Munsterberg, 1907). Seminal work by Hartshorne and May (1924) and Meehl and Hathaway (1946) established the basis for examining faking on psychological tests. Despite nearly a century of psychological research, however, no single detection method has been widely accepted as being sufficiently accurate. Psychophysiology-based approaches can detect “guilty” participants in experimental settings with fairly high accuracy (approximately within the range of 75–90%; Ben-Shakhar & Elaad, 2003; Inbau, Reid, Buckley, et al., 2004; MacLaren, 2004; Offe & Offe, 2007), and lower, but still respectable, accuracy rates in operational settings (Elaad, Ginton, & Jungman, 1992). However, the socio-political nature of accusing a person as a fake/liar requires a strict avoidance of false positive errors (i.e., false accusations), and polygraphs are perceived as producing enough errors to warrant their prohibition (i.e., Employee Polygraph Protection Act) in most work settings in the USA (Iacono & Lykken, 1997; Myers, Latter, Abdollahi-Arena, 2006; Sackett, Burris, & Callahan, 1989). In any case, the less precise nature of self-report measurements implies that faking scales (e.g., SDR measures) should be expected to achieve lower detection rates. General psychological research indicates that individuals lack the ability to detect lies without special instruments, achieving accuracy rates just above chance (Bond & DePaulo, 2006). Even professionals with formal training in detecting dissimulation (e.g., nurses, clinical psychologists, and secret service
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
263
agents) offer little evidence that humans can detect lies during direct social interactions (Bond & DePaulo, 2008; Ekman, O’Sullivan, & Frank, 1999); the most promising evidence (e.g., Ekman & O’Sullivan, 1991) only demonstrates that people can detect artificial lies that were designed to exhibit certain types of leakage (Porter & ten Brinke, 2008). Thus, there are few reasons to suspect that selection practitioners – who lack specific deception training – are able to detect faking accurately (Robie, Tuzinski, & Bly, 2006). Further complicating the matter, fakers may actively counter detection attempts by faking less (frequently and/or extremely), using sophisticated strategies, minimizing leakage, and faking the detection scale. Only a few methods allow for researchers to be highly confident that they indexed faking, such as applicants’ own admissions of faking, verifications of data against credible sources, and “bogus” claims that cannot possible be true. However, research clearly shows that faking can be detected at some level (Ones & Viswesvaran, 1998), though measures may be imperfect. The rest of this chapter rests on the assumption that organizations can detect faking with some degree of accuracy. Generally, individuals that discover they are the targets of deception tend to experience negative affect (Bies, & Tripp, 1996; Boon & McLeod, 2001; McCornack & Levine, 1990; Trovillo, 1939) and a desire to impose punitive sanctions, including termination of the relationship. Organizational representatives that detect applicant/employee deception may react similarly, at their own discretion or in accordance with formal organizational policies (Carter v. Tennant Co., 2004; Goffin & Christiansen, 2003). Yet, surveys also reveal that many organizations still hire applicants known to be dishonest (Gurchiek, 2008). Hough (1998) demonstrated that correcting test scores for faking and removing detected fakers from the applicant pool entirely can, under certain conditions (e.g., low selection ratio, valid measure of faking), produce very similar results. Figure 7.1 incorporates these notions to account for the influence that organizational targets have on both favorable and unfavorable outcomes. Correcting Content Test Scores for the Impact of Faking People might attempt to recover true scores on content scales by statistically “correcting” for faking (Cronbach, 1950). In theory, faking scales should represent suppressor variables that allow irrelevant variance to be partialled from predictor scores, before those scores are used to predict a criterion. Alternatively, faking indices can be used in moderator analyses to determine whether different criterion-related validities are achieved for people at different levels of faking. (For a more-detailed discussion see Burns & Christiansen, 2006.) Considering both scenarios, Schmitt (in Morgeson, Campion, Dipboye, et al., 2007: 709) noted that “in order for corrections to make an impact on criterionrelated validity or standardized performance measures, the faking measure that you are using has to be correlated with the outcome, the predictor, or both.”
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
264 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY In at least some situations, this set of requirements can be met. For instance, meta-analyses do not support a correlation between impression management and managerial performance ratings (Viswesvaran, Ones, & Hough, 2001), but do support correlations between SDR and some personality traits (McCrae & Costa, 1983; Ones, Viswesvaran, & Reiss, 1996; Paulhus, 2002). Corrections can affect scores for a large percentage of applicants (Burns & Christiansen, 2006) and do affect which individuals are hired (Christiansen, Goffin, Johnston, et al., 1994; Hough, 1998). Simultaneously, research consistently and strongly supports claims that such corrections have a negligible impact on validity (Barrick & Mount, 1996; Burns & Christiansen, 2006; Christiansen, Goffin, Johnston, et al., 1994; Ellingson, Sackett, & Hough, 1999; Hough, 1998; Li & Bagger, 2006; Ones, Viswesvaran, & Reiss, 1996). As a result, Ones, Viswesvaran, and Reiss (1996) found explanations of SDR as a suppressor to be unsupported. Moreover, a statistical simulation revealed that even discarding up to 30% of an applicant pool due to faking produced small improvements to criterion-related validities, under typical selection conditions (Schmitt & Oswald, 2006). These conclusions also follow logically from demonstrations that faking has little effect on criterionrelated validity. Another problem lies in the fact that corrections can only be effective to the extent that faking scales are valid. Unlike studies of lie detection that produce dichotomous decision outcomes (i.e., liar vs. innocent), faking scales must be able to measure the type and magnitude of faking. Most scales fail to capture more than one faking strategy (usually SDR) and will be limited when used in isolation. Even when multiple scales are used, the accumulated evidence supports few scales as sufficiently construct-valid. Using a more robust faking index, however, Anderson, Warner, and Spencer’s (1984) study provides a rare exception where corrections to real applicants’ scores increased validity. With student samples taking a new refined measure of response distortion, Hakstian and Ng (2005) also improved criterion-related validities for certain performance criteria. Effective Selection Decisions The literature offers tentative conclusions for a number of topics previously studied, but at least one thing is clear: organizations cannot avoid hiring at least some fakers (Peterson & Griffith, 2006). Whether analyzing mean scores, convergent/divergent validity evidence, factor patterns, rank-orders, criterionrelated validity, or other indicators of construct validity, researchers and practitioners want to know whether hiring applicants who fake will impact organizations negatively (or even positively). Thus far, the literature exhibits a clear reliance on job performance criteria, but the hiring of known fakers could lead applicants and current employees to view the selection process as unfair, or at least irresponsible (Hough, 1998; Morgeson, Campion, Dipboye, et al.,
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
265
2007). McFarland (2003) explored organizational justice perceptions in relation to warnings, but researchers might need to examine perceptions of current employees regarding the effectiveness of selection decisions, especially if they believe that peers faked those assessments. To evaluate the impact of faking more effectively, hypotheses must be formulated from clear theories. For instance, some believe personality constructs predict organizational citizenship performance (Borman & Motowidlo, 1997; Organ, 1997) partly because such outcomes depend more on behavioral tendencies and volition than on knowledge or skill (Motowidlo, Borman, & Schmit, 1997). That personality tests and measures of behavioral tendency seem particularly susceptible to faking (Nguyen, Biderman, & McDaniel, 2005) implies that faking might decrease validities with citizenship performance (Hoffman, Blair, Meriac, et al., 2007), but not with task performance. This hypothesis requires further exploration (Peterson & Griffith, 2006), though Ones and Viswesvaran (1998) reported near-zero correlations between SDR and contextual performance criteria from Project A. In other work, Ones, Viswesvaran, and Reiss (1996) have shown that SDR relates to learning (i.e., training performance and school success), but not to job performance (overall and task) or counterproductive behaviors. With a broader focus on organizational behavior, Moorman and Podsakoff (1992) presented meta-analytic correlations between SDR and a range outcomes rarely considered by faking researchers, showing small positive relationships with job satisfaction and organizational commitment, but negative relationships with role conflict, role ambiguity, and locus of control. Although constructs were defined somewhat broadly, the findings suggest that faking, to the extent it relates to SDR, might enhance communications within organizations. However, one can also hypothesize that dishonesty is detrimental to organizational communications in many instances. The following section explores potential faking outcomes that would occur after the selection phase.
DISTAL OUTCOMES OF APPLICANT FAKING Employees tell many kinds of work-related lies (Gurchiek, 2007; Haefner, 2007), and some resemble specific responses given during faking, like exaggerating skills or claiming false accomplishments. The motivation to be dishonest may resemble faking motivation (e.g., need for a favorable performance evaluation). Given that past behaviors predict future behaviors, applicants who fake should be more likely to use similar forms of deception as employees. Basic reinforcement theory supports this claim as people rewarded for being dishonest will learn to repeat the behavior. Also, people who “get away” with faking may need to maintain that deception as employees (e.g., pretending to have a skill or knowledge set). For example, a dean of the Massachusetts Institute of Technology recently resigned after a lie on her resume was discovered 28 years
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
266 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY later (Lewin, 2007). Conversely, applicants who are caught faking may be less likely to use deception in the future. Studies have yet to examine this potential relationship directly, perhaps because employee (dis)honesty has been viewed as just one aspect of broader concepts, namely integrity (Berry, Ones, & Sackett, 2007) and employee deviance (Bennett & Robinson, 2000). Conceptual and operational measurements have provided only loose distinctions between dishonesty and other acts such as employee theft (e.g., stealing physical property, claiming extra work hours), violence, drug use (e.g., Jones, 1981; Lewicki & Stark, 1996; Murphy, 1993; O’Bannon, Goldfinger, & Appleby, 1989; Sackett & Decker, 1979; Scott & Jehn, 2003), and ingratiation (e.g., Drory & Zaidman, 2007). Dishonesty may be one indicator of general integrity or deviance, but other counterproductive behaviors do not fit basic definitions of deception. Furthermore, dishonesty may appear to be a strong indicator of integrity partly because it is often caused by employee deviance, as a way for employees to “cover up” any wrongdoing and to avoid punishment. Faking and/or employee dishonesty might also influence the quality of intraorganizational relationships, particularly ones dependent upon trust. To explain relationships between individuals or work groups, trust can be defined as a willingness to depend on (or “be vulnerable to”) another party based on cognitive expectations about that party’s intended behaviors (Dirks, 2000; Mayer, Davis, & Schoorman, 1995; Rousseau, Sitkin, Burt, et al., 1998), particularly in situations of uncertainty or risk that require the initial party to expose a vulnerability (Bigley & Pearce, 1998; Colquitt, Scott, & LePine, 2007). Because dishonesty poses a direct threat to trust, propositions linking these concepts (e.g., Kim, Ferrin, Cooper, et al., 2004; Mishra, 1996) may seem rather unremarkable. Nonetheless, more needs to be understood about how different types and degrees of deception affect trust and perceptions of “trustworthiness” over time (Levin, Whitener, & Cross, 2006; Lewicki & Bunker, 1996; Rousseau, Sitkin, Burt, et al., 1998). Employee dishonesty varies considerably, ranging from assertions about false limitations (Becker & Martin, 1995) to claiming credit for another’s work (Bratton & Kacmar, 2004), and targets who detect faking and dishonesty may exhibit a variety of reactions, ranging from complete forgiveness to the imposition of a severe penalty (Elangovan & Shapiro, 1998; Kim, Ferrin, Cooper, et al., 2004). Some reactions may only influence the relationship between two employees, but others may affect trust among other peers, work groups, or even within the broader organizational culture (Elangovan & Shapiro, 1998; Ferrin, Dirks, & Shah, 2006). Research on trust in co-worker relationships has grown rapidly (Colquitt, Scott, & LePine, 2007; Lau & Liden, 2008). Acknowledging that the direction of causality could not be established, Colquitt, Scott, and LePine (2007) found consistent meta-analytic support for moderate correlations between trust and performance behaviors (risk-taking, task, citizenship, and counterproductive).
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
Actions
Pre-Entry
Consequences for Org.
General impression management
Assessment errors increases
Faking qualifications to meet minimum requirements
Selection errors increases
Faking characteristics to appear more desirable (e.g., committed) than competing applicants
267
Poorer quality salary negotiations Reputation of assessments’ validity decreases Applicants perceptions of injustice increases
Distorting situation negotiate greater compensation or benefits
Post-Entry
General impression management
Job performance poorer for tasks requiring faked KSAOs
Maintaining pre-hiring deception
Fakers’ commitment increases via dissonance, perhaps resulting in improved performance
Lying about violation of organizational policies Extreme careerism Sacrifice of personal impression for coworker or collective group Faking new qualifications for promotion & progress
Performance losses by employees who maintain deception; organization resources wasted trying to detect deception Coordination losses due to communication of false information Trust between deceiver & coworkers decreased Cohesion between sacrificing deceiver & peers increased Organizational culture promotes distrust and/or lack of integrity & ethics Coworkers’ perceptions of injustice increase Need to re-hire after fakers are identified & dismissed Public reputation of organization decreases Organizational effectiveness impacted
Figure 7.3 Potential actions and outcomes of deception in work settings. KSAO, knowledge, skills, abilities, and other characteristics.
Others have suggested that trust between employees, or a lack of it, may have a direct impact on group-level outcomes, such as cohesion, cooperation, and coordination resulting from the predictability of others’ actions (e.g., Dirks & Ferrin, 2001; Jones & George, 1998; Klimoski & Carol, 1976; Rousseau, Sitkin, Burt, et al., 1998; Simons & Peterson, 2000). Mayer, Davis, and Schoorman (1995) also proposed that group functioning among diverse members may be particularly dependent on trust. Additionally, Salamon and Robinson (2008) showed that aspects of organizational culture (i.e., responsibility norms) can mediate perceived trust to influence organizational-level sales. Figure 7.3 lists just a few possible outcomes of faking and employee deception, distinguishing them based on whether they occur pre- or post-hiring. For brevity, Figure 7.3 does not include positive outcomes, but reasonable hypotheses can be formed about how deception might facilitate group functioning (Rozell & Gundersen, 2003) – at least for the short-term. Especially when viewed over a long period of time, the negative impact of faking on organizational outcomes may increase as employees who use dishonesty rise to higher positions and gain greater influence over subordinates, work processes, and organizational culture.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
268 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY
FAKING AND CULTURAL CONCERNS The literature contains few discussions of faking across cultural groups (Frei, Yoshita, & Isaacson, 2006). Current practices (Robie, Tuzinski, & Bly, 2006) warrant the consideration of at least two issues: 1. Cross-cultural use of faking scales; and 2. Fairness issues in selection. Similar Conceptions of Faking/Social Desirability Most relevant studies pertain to assessments of SDR, which tends to be defined more broadly than in faking research. Here, SDR encompasses lies and other behaviors aimed at social approval (e.g., Mwamwenda, 1996). Notwithstanding this caveat, the majority of studies indicate judgments about the social desirability of behaviors and lies are fairly similar across cultures (Collazo, 2005; Fioravanti, Gough, & Frere 1981; Fu, Xu, Cameron, et al., 2007; Tripathi & Shahjahan, 1987; Tsushima, 1969). Of course, some cultures interpret specific behaviors differently, but identifiable cultural characteristics have not been linked to general interpretations of SDR (e.g., Diamantopolous, Reynolds, & Simintiras, 2006; Li & Reb, 2009). Consequently, SDR measures of faking can be administered to applicants of different cultural backgrounds. One possible exception to this conclusion comes from studies of Japanese samples that tended to produce SDR scores with less variance than Western samples. However, this effect seems to result from a Japanese rating bias to select responses near the midpoint of a scale (Stening & Everett, 1984), in contrast to an extremity bias shown by Americans (Iwawaki & Cowen, 1964). When associated with specific cultures, these biases may cause those groups to (i) be more likely to avoid detection, but also to (ii) fake in a manner that distorts content test scores insufficiently. Fairness Issues Applicant pools containing members from multiple cultural groups may lead to issues of unfairness (Sandal & Endresen, 2002). Just as Sackett and Wanek (1996) proposed for integrity tests, the use of faking scales to select applicants may disadvantage sub-groups that are less willing to fake or less capable of faking. Only two studies have examined ethnic group differences in faking by job applicants. For police applicants, Hough (1998) found that Black people scored slightly higher than Caucasians on an unlikely virtues scale (d = 0.13), and “minorities” in two other samples scored lower than Caucasians (average d was 0.14 and 0.41, respectively). Dudley, McFarland, Goodman, et al. (2005) found that Asians scored higher on SDR than Caucasians, by an appreciable extent (average d = 1.04), as did Hispanics (average d = 0.47)
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
269
and Black people (average d = 0.37). They also showed that fewer minorities would be selected if SDR scores were considered. These results imply that adverse impact or procedural fairness issues could result in either of two situations: 1. Minorities are allowed to fake and enhance their chances of being hired; or 2. Minorities are punished for faking and selected at lower rates. Dudley, McFarland, Goodman, et al. (2005) did not provide evidence that their results reflected cultural differences but surmised that SDR increases resulted because Black people distrust testing procedures and limit selfdisclosure, Hispanics acquiesce, and Asians seek harmony. Thus, SDR might be regarded as a negative quality for Black people, but as a neutral or positive quality for Hispanics and Asians. These claims require further investigation, but some evidence supports the supposition about Asians. Cross-cultural examinations indicate that SDR is used more by collectivistic cultures (Hofstede, 1983; Triandis, McCusker, & Hui, 1990) in which individuals are encouraged to dismiss personal desires for group needs, including the maintenance of harmony (Lalwani, Shavitt, & Johnson, 2006; Middleton & Jones, 2000). East Asian nations tend to be automatically classified as collectivistic, but studies also show greater deception by nations with similar values, including Israeli-Jews and Israel-Arabs (Sohlberg, 1976), South Africans (Mwamwenda, 1996), Samoans (Aune & Waters, 1994), and Portuguese (in relation to UK members; Lopes & Fletcher, 2004). Hypothesizing that Norway discourages self-promotion and assertiveness, Sandal and Endresen (2002) found that Norwegian students scored significantly lower on a faking scale than US students. Because collectivist societies encourage members to make interdependent self-construals (Gelfand, Nishii, Holcombe, et al., 2001; Markus & Kitayama, 1991), they may promote deception as a means of subordinating personal attitudes and goals to those of the group (Fu, Xu, Cameron, et al., 2007). Another explanation comes from evidence that individuals experience pressure to conform to group expectations (Heine & Lehman, 1995; Hewlin, 2009) and not appear statistically deviant (or that members should self-enhance only when it benefits the group; Heine & Hamamura, 2007). Consequently, Asian applicants may fake more, but perhaps in ways that lower content test scores when faking within their cultural group. However, all applicants will be motivated to enhance their appearance to some degree and Asian applicants may not perceive an employer as representing their cultural group, so it is not clear how people reconcile collectivist influences with the personal need to compete with other candidates or need to represent their own group to out-group members (e.g., hiring organizations). As implied by Figure 7.3, cultural differences in faking may translate into differences in employee deception. Xin (2004) revealed that Asian American
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
270 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY managers disclose less information and use different IM tactics (being more job-focused) than European American managers, and Asians’ behaviors led to lower quality relationships with supervisors. In other cases, people may claim (falsely) responsibility for a group failure to help another member maintain his or her social standing and avoid public shame (Komarraju, Dollinger, & Lovell, 2008; Triandis, Brislin, & Hui, 1988). This act of helping another to “save face” (Gil-del-real & Brown, 1980; Merkin, 2006) could tighten social bonds (Berg & Jaya, 1993; Covelman & Covelman, 1993; Kim, 1985), reduce conflict, and reinforce power structures (Randall, Huo, & Pawelk, 1993). Minority group members might also use deception to adapt and fit in, though only a few studies have explored this issue (e.g., Foldes, Ones, & Sinangil, 2006; Hewlin, 2009; Montagliani & Giacalone, 1998). The examples illustrate different ways in which cultures promote the use of deception, and some might actually increase organizational effectiveness. Faking as Privacy Organizations that detect and punish faking may also affect a disproportionate number of people from groups sensitive to violations of privacy. Most organizations presumably avoid using assessment questions that would offend applicants, or at least make earnest attempts to exclude items of questionable legality (e.g., per Title VII of the Civil Rights Act of 1964 in the USA). However, applicants might still be asked illegitimate questions (e.g., about age, sexual orientation, or family life) during selection and after hiring. Thus, they may feel the need to respond with fake answers as a way of maintaining privacy (cf. DePaulo, Wetzel, Sternglanz, et al., 2003). As an example, Hall (1986) revealed that lesbian professionals actively managed their work communications to conceal or withhold information (as omissive deception) to avoid discrimination. Employees may act similarly when dealing with issues that they perceive to be beyond the organization’s purview, such as domestic violence (O’Leary-Kelly, Lean, Reeves, et al., 2008). More broadly, certain ethnic cultures may place a particularly high value on rights to privacy and the maintenance of a social reputation (and avoidance of public humiliation). When applicant pools contain individuals from a mix of cultures, organizations may also disadvantage certain sub-groups by expecting or requiring full disclosure.
FAKING AND VARIATIONS IN RESEARCH METHODOLOGY The literature provides many interesting findings, but contains a mixed bag of null, simple, and complex results that can obscure the broader process of faking, especially in the absence of theory. At least some differences in past
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
271
findings are due to variations in the methodology used to study different effects, but exploring potential moderators (e.g., Birkeland, Manson, Kisamore, et al., 2006) would require deeper coverage than can be offered here. I highlight just a few points to illustrate how design differences make it difficult to derive clear interpretations. A major criticism of many faking studies stems from the use of imperfect measures of faking. As noted here, early works defined faking, SDR, and other terms differently (Mesmer-Magnus, Viswesvaran, Deshpande, et al., 2006; Ones, Viswesvaran, & Reiss, 1996; Zickar & Gibby, 2006), which makes direct comparisons of study results difficult. Acknowledging that no measure can be perfect and that no applicant will fake completely (with no response indicating any truth) may alleviate the most extreme concerns. However, there is a growing consensus that different faking indicators have unique strengths and critical weaknesses (Alliger & Dwight, 2000; Barger, 2002; Burns & Christiansen, 2006; Morgeson, Campion, Dipboye, et al., 2007; Reid-Seiser & Fritzsche, 2001; Tett & Christiansen, 2007). At the outset, it must be acknowledged that SDR and commonly used faking indices correlate moderately, indicating some common construct seems to be tapped (Mesmer-Magnus, Viswesvaran, Deshpande, et al., 2006; Moorman & Podsakoff, 1992; Mueller-Hanson, Heggestad, Thornton, et al., 2003; Smith & Ellingson, 2002). Beyond this, even the widely used Balanced Inventory of Desirable Responding (BIDR) measure of SDR has an unclear factor structure and ambiguous empirical support (e.g., Helmes & Holden, 2003; Leite & Beretvas, 2005). Many researchers continue to rely on a two-dimension model for BIDR responses, despite Paulhus’ (2002) introduction of a two-tiered, fourdimensional model to update his earlier, often-cited work (Zerbe & Paulhus, 1987). Also, Li and Bagger’s (2007) quantitative review indicated that mean reliability estimates for the two BIDR scales commonly used were acceptable, but low (around 0.7, ranging from 0.27 to 0.92). Another criticism of faking scales is that they are susceptible to faking (Levin & Zickar, 2002; Pauls & Crost, 2004), just like content tests. Unless test designers can disguise faking scale items as “regular” items, applicants may be able to work around them, or even endorse them in an undesirable manner to fake an appearance of honesty and genuineness. Also, items in SDR scales most often refer to improbable, but not impossible events (e.g., “unlikely virtues”), meaning that the best, honest applicants may endorse them and be regarded as fakers, at least in theory. To address the limitations above, researchers might use a battery of faking scales to capture the use of different strategies and resulting response patterns (Baer, Wetter, & Berry, 1992; Bagby, Nicholson, Buis, et al., 2000; Roger, 1997). Bogus item scales (Anderson, Warner, & Spencer, 1984; Pannone, 1984) lure people into endorsing statements about fictitious events and qualities that cannot possibly be true, unlike the unlikely (but possible) virtues of SDR scales. This cannot prevent false positive errors, but should reduce them considerably. Researchers could also solicit self-reports of faking
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
272 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY strategies and behaviors, whenever participants have the cognitive capacity to report them accurately (e.g., Weiss & Feldman, 2006). Alternatively, one could corroborate test responses against ratings from other credible sources (e.g., Topping & O’Gorman, 1997), like co-workers, friends, or past employers (Ones, Dilchert, Viswesvaran, et al., 2007). Other methods for detecting deceit (Ford, 2006) that have received less attention in the industrial and organizational literature include:
r Polygraphs (Ben-Shakur & Elaad, 2003; Iacono, 2008; Inbau, Reid, Buckley, et al., 2004; National Research Council, 2003):
r Intuitive behavioral observations (Bond & DePaulo, 2006; Ekman & O’Sullivan, 1991);
r Facial expression coding systems (Ekman, 2001; Ekman & O’Sullivan, 2006);
r Response latency/time (Farwell & Donchin, 1991; Gronau, Ben-Shakhar, & Cohen, 2005; Holden, Kroner, Fekken, et al., 1992; Khorramdel & Kubinger, 2006; Seymour & Kerlin, 2008; Walczyk, Schwartz, Clifton, et al., 2005); r Linguistic style analyses (Newman, Pennebaker, Berry, et al., 2003; Pennebaker, Mehl, & Niederhoffer, 2003; Tangney, 1992); and r Neuroimaging techniques (Ganis, Kosslyn, Stose, et al., 2003; Langleben, Loughead, Bilker, et al., 2005; Rosenfeld, 2005; Spence & Kaylor-Huges, 2008). Also, new analytical techniques for identifying faking based on patterns in test responses might be useful (e.g., Kuncel & Borneman, 2007; Sotaridona & Meijer, 2003; Sotaridona, van der Linden, & Meijer, 2006; van der Linden & Sotaridona, 2004; Wollack, 2007). In particular, methods of detecting faking based on item response theory (IRT) have the potential to identify specific patterns of faking in test responses, and can be applied without altering traditional administration procedures. As opposed to examining score distortion at the scale-level, Zickar (e.g., Zickar, Gibby, & Robie, 2004; Zickar & Robie, 1999) has promoted analyses of specific response options to determine whether tendencies imply that individuals are distorting scores away from latent trait standings (true scores) in a systematic and predictable way (Zickar, 2000). Preliminary findings support the use of various IRT models to capture different types of distortion (e.g., Stark, Chernyshenko, Chan, et al., 2001; Zickar, Gibby, & Robie, 2004), and LaHuis and Copeland (2009) introduced an approach that may help test developers and selection personnel identify items most likely to be faked for a given person, whose ability level is discrepant from the item’s difficulty level. The initial approaches appear promising and offer a direct method of testing propositions about specific types of faking decisions and behaviors, but also require further incorporation of substantive theories and recognition that applicant deception encompasses a wide range of strategies.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
273
The other major concern about methodology pertains to the fidelity of simulated versus actual selection settings (Frei, Yoshita, & Isaacson, 2006). Tett and Christiansen (2007) point out that many design features have been confounded with whether or not real applicant samples were observed during selection, and these features may affect the type of faking that occurred. For instance, many have concluded that experimental instructions typically induce much more faking than would be performed by real applicants (Doll, 1971; Griffith, Chmielowski, & Yoshita, 2007; Smith & Ellingson, 2002; Viswesvaran & Ones, 1999) and that this difference explains why content and faking scale scores tend to be lower in field research (Hough, 1998). However, the same pattern of results might be observed for alternative reasons. Kim (2007) found that a number of individuals from a student sample automatically assumed, without receiving a warning, that faked answers in a simulated selection procedure would be detected and penalized. If such assumptions are more likely in real selection settings, field research may be confounded with particular punishments. Furthermore, field settings commonly rely on between-groups designs that allow only for inferences that applicants performed faking without direct evidence, resulting in conclusions with caveats such as: “to the extent that the incomparability between the applicant and incumbent samples is due to faking” (Stokes, Hogan, & Snell, 1993: 754). By contrast, simulated selection procedures allow for within-subjects analyses of faking, but conclusions are valid only when accurate faking indices or clear instructional sets (honest vs. faked) are used, which may be difficult as noted above. Finally, operational definitions of faking manipulations and groups vary enough that the comparability of conditions across studies should be questioned. Not all applicant samples are equal, as some studies include “real applicants” taking selection tests while other samples completed tests “for research purposes” with no bearing on selection decisions (e.g., Hough, Eaton, Dunnette, et al., 1990; Schmitt & Kunce, 2002). Outside of field settings, researchers have created faking conditions by telling people to fake in a particular way (e.g., “fake good” or fake subtlely”), having participants imagine applying for jobs, offering small monetary incentives for top test scores, and implying that real job opportunities might be presented at a future date for top test scores. For between-group analyses of faking magnitude, one must also decide whether “honest” conditions are comparable. Studies have had applicants, job incumbents, and nonapplicants respond to tests with standard/“straighttake” content test instructions, instructions to be “as honest as possible,” and instructions “to be somewhat modest” (Paulhus, Bruce, & Trapnell, 1995).
REFERENCES Abrahams, N.M., Neumann, I., & Githens, W.H. (1971). Faking vocational interests: Simulated versus real life motivation. Personnel Psychology, 24, 5–12.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
274 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Alder, K. (2007). The History of an American Obsession: The Lie Detectors. New York: The Free Press. Alliger, G.M., & Dwight, S.A. (2000). A meta-analytic investigation of the susceptibility of integrity tests to response distortion. Educational and Psychological Measurement, 60, 59–72. Alliger, G.M., & Dwight, S.A. (2001). Invade or evade? The trade-off between privacy invasion and item fakability. Applied H. R. M. Research, 6, 95–104. Anderson, C.D., Warner, J.L., & Spencer, C.C. (1984). Inflation bias in self-assessment examinations: Implications for valid employee selection. Journal of Applied Psychology, 69, 574–80. Arthur, W.A. Jr., Glaze, R.M., Villado, A.J., & Taylor, J.E. (2009). Unproctored Internet-based tests of cognitive ability and personality: Magnitude of cheating and response distortion. Industrial and Organizational Psychology: Perspectives on Science and Practice, 2, 39–45. Ash P. (1987). Honesty lest scores, biographical data, and delinquency indicators. Presented at the Annual Meeting of the Academy of Criminal Justice Sciences, St. Louis, MO. Attali, Y., & Bar-Hillel, M. (2003). Guess where: The position of correct answers in multiple-choice test items as a psychometric variable. Journal of Educational Measurement, 40, 109–28. Aune, R.K., & Waters, L.L. (1994). Cultural differences in deception: Motivations to deceive in Samoans and North Americans. International Journal of Intercultural Relations, 18, 159–72. Baer, R.A., Wetter, M.W., & Berry, D.T.R. (1992). Detection of underreporting of psychopathology on the MMPI: A meta-analysis. Clinical Psychology Review, 12, 509–25. Bagby, R.M., Nicholson, R.A., & Buis, T. (1998). Utility of the deceptive-subtle items in the detection of malingering. Journal of Personality Assessment, 70, 405–15. Bagby, R.M., Nicholson, R.A., Buis, T., & Bacchiochi, J.R. (2000). Can the MMPI-2 validity scales detect depression feigned by experts? Assessment, 7, 55–62. Banerjee, R., & Yuill, N. (1999). Children’s understanding of self-presentational display rules: Associations with mental-state understanding. British Journal of Developmental Psychology, 17, 111–24. Barger, S.D. (2002). The Marlowe–Crowne affair: Short-forms, psychometric structure, and social desirability. Journal of Personality Assessment, 79, 286–305. Barrick, M.R., & Mount, M.K. (1996). Effects of impression management and selfdeception on the predictive validity of personality constructs. Journal of Applied Psychology, 81, 261–72. Bartram, D. (2009). The international test commission guidelines on computer-based and internet-delivered testing. Industrial and Organizational Psychology: Perspectives on Science and Practice, 2, 11–3. Bartram, D., & Brown, A. (2004). Online testing: Mode of administration and the stability of OPQ 32i scores. International Journal of Selection and Assessment, 12, 278–84. Beck, L., & Ajzen, I. (1991). Predicting dishonest actions using the theory of planned behavior. Journal of Research in Personality, 25, 285–301. Becker, T.E., & Colquitt, A.L. (1992). Potential versus actual faking of a biodata form: An analysis along several dimensions of item type. Personnel Psychology, 45, 389– 406. Becker, T.E., & Martin, C.L. (1995). Trying to look bad at work: Methods and motives for managing poor impressions in organizations. Academy of Management Journal, 38, 174–99.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
275
Ben-Shakhar, G., & Elaad, E. (2003). The validity of psychophysiological detection of information: with the Guilty Knowledge Test – A meta-analytic review. Journal of Applied Psychology, 88, 131–51. Bennett, R.J., & Robinson, S.L. (2000). Development of a measure of workplace deviance. Journal of Applied Psychology, 85, 349–60. Benussi, V. (1914). Beitr¨age zur psychologie der gestalt- und Bewegungserlebnisse: ¨ die gesamte. Psychologie, 32, 50–7. Archiv fur Berg, I.K., & Jaya, A. (1993). Different and same: Family therapy with Asian-American families. Journal of Marital and Family Therapy, 19, 31–8. Berry, C.M., Ones, D.S., & Sackett, P.R. (2007). Interpersonal deviance, organizational deviance, and their common correlates: A review and meta-analysis. Journal of Applied Psychology, 92, 410–24. Berry, C.M., & Sackett, P.R. (2009). Faking in personnel selection: Tradeoffs in performance versus fairness resulting from two cut-score strategies. Personnel Psychology, 62, 835–63. Berry, C.M., Sackett, P.R., & Wiemann, S. (2007). A review of recent developments in integrity test research. Personnel Psychology, 60, 271–301. Bies, R.J., & Tripp, T.M. (1996). Beyond distrust: “Getting even” and the need for revenge. In R.M. Kramer & T.R. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research (pp. 246–60). Thousand Oaks, CA: Sage. Bigley, G.A., & Pearce, J.L. (1998). Straining for shared meaning in organization science: Problems of trust and distrust. Academy of Management Review, 23, 405– 21. Birkeland, S.A., Manson, T.M., Kisamore, J.L., Brannick, M.T., & Smith, M.A. (2006). A meta-analytic investigation of job applicant faking on personality measures. International Journal of Selection and Asssessment, 14, 317–35. Bok, S. (1978). Lying: Moral Choice in Public and Private Life. New York: Vintage. Bolino, M.C., & Turnley, W.H. (2003). More than one way to make an impression: Exploring profiles of impression management. Journal of Management, 29, 141–60. Bond, C.F. Jr., & DePaulo, B.M. (2006). Accuracy of deception judgments. Personality and Social Psychology Review, 10, 214–34. Bond, C.F. Jr., & DePaulo, B.M. (2008). Individual differences in judging deception: Accuracy and bias. Psychological Bulletin, 134, 477–92. Boon, S.D., & McLeod, B.A. (2001). Deception in romantic relationships: Subjective estimates of success at deceiving and attitudes toward deception. Journal of Social and Personal Relationships, 18, 463–75. Booth-Kewley, S., Edwards, J.E., & Rosenfeld, P. (1992). Impression management, social desirability, and computer administration of attitude questionnaires: Does the computer make a difference? Journal of Applied Psychology, 77, 562–6. Borislow, B. (1958). The Edwards Personal Preference Schedule (EPPS) and fakability. Journal of Applied Psychology, 42, 22–7. Borman, W.C., & Motowidlo, S.J. (1997). Task performance and contextual performance: The meaning for personnel selection research. Human Performance, 10, 99–109. Broomfield, K.A., Robinson, E.J., & Robinson, W.P. (2002). Children’s understanding about white lies. British Journal of Developmental Psychology, 20, 47–65. Bratton, V.K., & Kacmar, K.M. (2004). Extreme careerism: The dark side of impression management. In R.W. Griffin & A.M. O’Leary-Kelly (Eds), The Dark Side of Organizational Behavior (pp. 291–308). San Francisco, CA: Jossey-Bass. Buller, D.B., & Burgoon, J.K. (1994). Deception: Strategic and nonstrategic communication. In J.A. Daly & J.M. Wiemann (Eds), Strategic Interpersonal Communication (pp. 191–223). Hillsdale, NJ: Lawrence Erlbaum Associates.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
276 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Buller, D.B., & Burgoon, J.K. (1996). Interpersonal deception theory. Communication Theory, 3, 203–42. Buller, D.B., Burgoon, J.K., Buslig, A., & Roiger, J. (1996). Testing interpersonal deception theory: The language of interpersonal deception. Communication Theory, 6, 268–88. Burke, E. (2009). Preserving the integrity of online testing. Industrial and Organizational Psychology: Perspectives on Science and Practice, 2, 35–8. Burns, G.N., & Christiansen, N.D. (2006). Sensitive or senseless: On the use of social desirability measures in selection and assessment. In R. Griffith & M.H. Peterson (Eds), A Closer Examination of Applicant Faking Behavior (pp. 115–50). Geenwich, CT: Information Age. Carter v. Tennant Co., 383 F.3d 673 (2004). Caveon Test Security (2006). 2006 state education test security survey results. http://www.caveon.com/pr/2006 ed test security.pdf (accessed 30 June 2009). Christiansen, N.D., Goffin, R.D., Johnston, N.G., & Rothstein, M.G. (1994). Correcting the 16PF for faking: Effects on criterion-related validity and individual hiring decisions. Personnel Psychology, 47, 847–60. Christie, R., & Geis, E.L. (1970). Studies in Machiavellianism. New York: Academic Press. Clark, J.M. III., & Biderman, M.D. (2006). A structural equation model measuring faking propensity and faking ability. Paper presented at the 21st Annual Conference of the Society for Industrial and Organizational Psychology, Dallas, TX. Collazo, A.A. (2005). Translation of the Marlowe–Crowne Social Desirability Scale into an equivalent Spanish version. Educational and Psychological Measurement, 65, 780–806. Colquitt, J.A., Scott, B.A., & LePine, J.A. (2007). Trust, trustworthiness, and trust propensity: A meta-analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 92, 909–27. Converse, P.D., Oswald, F.L., Imus, A., Hedricks, C., Roy, R., & Butera, H. (2008). Comparing personality test formats and warnings: Effects on criterion-related validity and test-taker reactions. International Journal of Selection and Assessment, 16, 155–69. Converse, P.D., Peterson, M.H., & Griffith, R.L. (2009). Faking on personality measures: Implications for selection involving multiple predictors. International Journal of Selection and Assessment, 17, 47–60. Covelman, K.W., & Covelman, S. (1993). Saving face: A neglected dynamic in couples and family therapy. The Family Journal: Counseling and Therapy for Couples and Families, 1, 331–6. Cronbach, L.J. (1946). Response sets and test validity. Educational and Psychological Measurement, 6, 475–94. Cronbach, L.J. (1950). Response sets and test validity. Educational and Psychological Measurement, 10, 3–31. Cullen, M.J., Sackett, P.R., & Lievens, F. (2006). Threats to the operational use of situational judgment tests in the college admission process. International Journal of Selection and Assessment, 14, 142–55. Cunningham, M.R., Wong, D.T., & Barbee, A.P. (1994). Self-presentation dynamics on overt integrity tests: Experimental studies of the Reid Report. Journal of Applied Psychology, 79, 643–58. Dahling, J.J., Whitaker, B.G., & Levy, P.E. (2009). The development and validation of a new Machiavellianism scale. Journal of Management, 35, 219–57. Daniel, D.C., & Herbig, K.L. (1982). Propositions on military deception. In J. Gooch & A. Perlmutter (Eds), Military Deception and Strategic Surprise (pp. 155–77). London: Rutledge.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
277
Dannenbaum, S.E., & Lanyon, R.I. (1993). The use of subtle items in detecting deception. Journal of Personality Assessment, 61, 501–10. Davis, S.F., Grover, C.A., Becker, A.H., & McGregor, L.N. (1992). Academic dishonesty: Prevalence, determinants, techniques, and punishments. Teaching of Psychology, 19, 16–20. Dawson, M.E. (1980). Physiological detection of deception: Measurement of responses to questions and answers during countermeasure maneuvers. Psychophysiology, 17, 8–17. DePaulo, B.M. (1992). Nonverbal behavior and self-presentation. Psychological Bulletin, 111, 203–43. DePaulo, B.M., Kashy, D.A., Kirkendol, S.E., Wyer, M.M., & Epstein, J.A. (1996). Lying in everyday life. Journal of Personality and Social Psychology, 70, 979–95. DePaulo, B.M., & Pfeifer, R.L. (1986). On-the-job experience and skill at detecting deception. Journal of Applied Social Psychology, 16, 249–67. DePaulo, B.M., & Rosenthal, R. (1979). Telling lies. Journal of Personality and Social Psychology, 37, 1713–22. DePaulo, B.M., Wetzel, C., Sternglanz, R.W., & Wilson, M.J.W. (2003). Verbal and nonverbal dynamics of privacy, secrecy, and deceit. Journal of Social Issues, 59, 391–410. Diamantopoulos, A., Reynolds, N.L., & Simintiras, A.C. (2006). The impact of response styles on the stability of cross-national comparisons. Journal of Business Research, 59, 925–35. Dicken, C.F. (1959). Simulated patterns on the Edwards Personal Preference Schedule. Journal of Applied Psychology, 43, 372–8. Dike, C.C., Baranosk, M., & Griffith, E.E.H. (2005). Pathological lying revisited. Journal of the American Academy of Psychiatry and the Law, 33, 342–9. Dilchert, S., Ones, D.S., Viswesvaran, C., & Deller, J. (2006). Response distortion in personality measurement: Born to deceive, yet capable of providing valid selfassessments? Psychology Science, 48, 209–25. Dirks, K.T. (2000). Trust in leadership and team performance: Evidence from NCAA basketball. Journal of Applied Psychology, 85, 1004–12. Dirks, K.T., & Ferrin, D.L. (2001). The role of trust in organizational settings. Organization Science, 12, 450–67. Doll, R.E. (1971). Item susceptibility to attempted faking as related to item characteristic and adopted fake set. Journal of Psychology, 77, 9–16. Donovan, J.J., Dwight, S.A., & Hurtz, G.M. (2003). An assessment of the prevalence, severity, and verifiability of entry-level applicant faking using the randomized response technique. Human Performance, 16, 81–106. Drory, A., & Zaidman, N. (2007). Impression management behavior: Effects of the organizational system. Journal of Managerial Psychology, 22, 290–308. Dudley, N.M., McFarland, L.A., Goodman, S.A., Hunt, S.T., & Sydell, E.J. (2005). Racial differences in socially desirable responding in selection contexts: Magnitude and consequences. Journal of Personality Assessment, 85, 50–64. Dunnette, M.D., McCartney, J., Carlson, H.C., & Kirchner, W.K. (1962). A study of faking behavior on a forced-choice self-description checklist. Personnel Psychology, 15, 13–24. Dwight, S.A., & Donovan, J.J. (2003). Do warnings not to fake reduce faking? Human Performance, 16, 1–23. Edwards, A.L. (1953). The relationship between the judged desirability of a trait and the probability that the trait will be endorsed. Journal of Applied Psychology, 37, 90–3. Ekman, P. (1985). Telling Lies: Clues to Deceit in the Marketplace, Marriage, and Politics. New York: Norton.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
278 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Ekman, P. (2001). Telling Lies: Clues to Deceit in the Marketplace, Politics, and Marriage (3rd edn). New York: Norton. Ekman, P., & Friesen, W.V. (1969). Nonverbal leakage and clues to deception. Psychiatry, 32, 88–105. Ekman, P., Friesen, W.V., O’Sullivan, M. (1988). Smiles when lying. Journal of Personality and Social Psychology, 54, 414–20. Ekman, P., & O’Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46, 913–20. Ekman, P., & O’Sullivan, M. (2006). From flawed self-assessment to blatant whoppers: The unitary of voluntary and involuntary behavior in detecting deception. Behavioral Sciences and the Law, 24, 673–86. Ekman, P., O’Sullivan, M., & Frank, M. (1999). A few can catch a liar. Psychological Science, 10, 263–5. Elaad, E., Ginton, A., & Jungman, N. (1992). Detection in measures in real-life criminal Guilty Knowledge Tests. Journal of Applied Psychology, 77, 757–67. Elangovan, A.R., & Shapiro, D.L. (1998). Betrayal of trust in organizations. Academy of Management Review, 23, 547–66. Ellingson, J.E., Sackett, P.R., & Connelly, B.S. (2007). Personality assessment across selection and development contexts: Insights into response distortion. Journal of Applied Psychology, 92, 386–95. Ellingson, J.E., Sackett, P.R., & Hough, L.M. (1999). Social desirability corrections in personality measurement: Issues of applicant comparison and construct validity. Journal of Applied Psychology, 84, 155–66. Ellingson, J.E., Smith, D.B., & Sackett, P.R. (2001). Investigating the influence of social desirability on personality factor structure. Journal of Applied Psychology, 86, 122–33. Elliot, S., Lawty-Jones, M., & Jackson, C. (1996). Effect of dissimulation on selfreport and objective measures of personality. Personality and Individual Differences, 21, 335–43. Farmer, B. (2007). Exam cheat probe into hidden earpieces. Telegraph (17 August 2007) http://www.telegraph.co.uk/news/uknews/1560521/Exam-cheat-probeinto-hidden-earpieces.html (accessed 30 June 2009). Farwell, L.A., & Donchin, E. (1991). The truth will out: Interrogative polygraphy (“lie detection”) with event-related brain potentials. Psycholphysiology, 28, 531–47. Fathi, N. (2008). Minister’s dismissal is setback for Iranian leader. New York Times (4 November 2008) http://www.nytimes.com/2008/11/05/world/ middleeast/05iran.html (accessed 30 June 2009). Feldman, R.S., Tomasian, J.C., & Coats, E.J. (1999). Nonverbal deception abilities and adolescents’ social competence: Adolescents with higher social skills are better liars. Journal of Nonverbal Behavior, 23, 237–49. Ferrin, D.L., Dirks, K.T., & Shah, P.P. (2006). Direct and indirect effects of third-party relationships on interpersonal trust. Journal of Applied Psychology, 91, 870–83. Fioravanti, M., Gough, H.G., & Frere, L.J. (1981). English, French, and Italian adjective check lists: A social desirability analysis. Journal of Cross-Cultural Psychology, 12, 461–72. Foldes, H.J., Ones, D.S., & Sinangil, H.K. (2006). Psychology Science, 48, 357–68. Ford, E.B. (2006). Lie detection: Historical, neuropsychiatric, and legal dimensions. International Journal of Law and Psychiatry, 29, 157–77. Ford, C.V., King, B.H., & Hollender, M.H. (1988). Lies and liars: Psychiatric aspects of prevarication. American Journal of Psychiatry, 145, 554–62. Frankfurt, H.G. (2005). On Bullshit. Princeton, NJ: Princeton University Press. Frei, R., Yoshita, Y., & Isaacson, J. (2006). Let’s go faking!: Culture and response distortion in international employment testing. In R. Griffith & M.H. Peterson (Eds),
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
279
A Closer Examination of Applicant Faking Behavior (pp. 333–55). Geenwich, CT: Information Age. Fu, G., Xu, F., Cameron, C.A., Heyman, G., & Lee, K. (2007). Cross-cultural differences in children’s choices, categorizations, and evaluations of truths and lies. Developmental Psychology, 43, 278–93. Furnham, A. (1990). Faking personality questionnaires: Fabricating different profiles for different purposes. Current Psychology: Research and Reviews, 9, 46–55. ´ Galasinksi, D. (2000). The Language of Deception: A Discourse Analytical Study. Thousand Oaks, CA: Sage. Ganis, G., Kosslyn, S.M., Stose, S., Thompson, W.L., & Yurgelun-Todd, D.A. (2003). Neural correlates of different types of deception; an fMRI investigation. Cerebral Cortex, 13, 830–6. Garry, R. (1953). Individual differences in ability to fake vocational interests. Journal of Applied Psychology, 37, 33–7. Gelfand, M.J., Nishii, L.H., Holcombe, K.M., Dyer, N., Ohbuchi, K.I., & Fukuno, M. (2001). Cultural influences on cognitive representations of conflict: Interpretations of conflict episodes in the United States and Japan. Journal of Applied Psychology, 86, 1059–74. Gil-del-real, M.T., & Brown, B.R. (1980). Potlatching and face-maintenance among the Kwakiutl of British Columbia. Journal of Psychoanalytic Anthropology, 3, 295– 308. Godson, R., & Wirtz, J.J. (2002). Strategic denial and deception. In R. Godson & J.J. Wirtz (Eds), Strategic Denial and Deception: The Twenty-first Century Challenge (pp. 1–14). New Brunswick, US: Transaction. Goffin, R.D., & Christiansen, N.D. (2003). Correcting personality tests for faking: A review of popular personality tests and an initial survey of researchers. International Journal of Selection and Assessment, 11, 340–4. Goffman, E. (1959). The Presentation of Self in Everyday Life. New York: Double Day. Goldberg, L.R. (1999). A broad-bandwidth, public domain, personality inventory measuring the lower-level facets of several five-factor models. In I. Mervielde, I. Deary, F. De Fruyt, & F. Ostendorf (Eds), Personality Psychology in Europe (volume 7), (pp. 7–28). Tilburg, The Netherlands: Tilburg University Press. Gombos, V.A. (2006). The cognition of deception: The role of executive processes in producing lies. Genetic, Social, and General Psychology Monographs, 132, 197– 214. Graham, K.E., McDaniel, M.A., Douglas, E.F., & Snell, A.F. (2002). Biodata validity decay and score inflation with faking: Do item attributes explain variance across items? Journal of Business and Psychology, 16, 573–92. Greenwald, A.G. (2008, June 9). Implicit association test: Validity debates. http://faculty.washington.edu/agg/iat validity.htm (accessed 30 June 2009). Greenwald, A.G., Nosek, B.A., & Banaji, M.R. (2003). Understanding and using the Implicit Association Test: I. An improved scoring algorithm. Journal of Personality and Social Psychology, 85, 197–216. Griffith, R.L., Chmielowski, T.S., & Yoshita, Y. (2007). Do applicants fake? An examination of the frequency of applicant faking behavior. Personnel Review, 34, 341–55. Griffith, R.L., & McDaniel, M. (2006). The nature of deception and applicant faking behavior. In R. Griffith & M.H. Peterson (Eds), A Closer Examination of Applicant Faking Behavior (pp. 1–19). Charlotte, NC: Information Age. Griffith, R.L., & Peterson, M.H. (2006). A Closer Examination of Applicant Faking Behavior. Charlotte, NC: Information Age. Grijalva, T.C., Nowell, C., & Kerkvliet, J. (2006). Academic honesty and online courses. College Student Journal, 40, 180–5.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
280 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Gronau, N., Ben-Shakhar, G., & Cohen, A. (2005). Behavioral and physiological measures in the detection of concealed information. Journal of Applied Psychology, 90, 147–58. Gurchiek, K. (2007). Sorry I’m late. A raccoon stole my work shoe.... SHRM News (29 May 2007) http://www.shrm.org/Publications/HRNews/Pages/CMS 021684.aspx (accessed 30 June 2009). Gurchiek, K. (2008, August 14). Mensa membership, Kennedy kinship among outrageous resume lies. SHRM News (14 August 2008). http://moss07.shrm.org/ Publications/HRNews/Pages/OutrageousResumeLies.aspx (accessed 30 June 2009). Haefner, R. (2007). One in five workers admit to telling lies at work. HR Magazine (1 May 2006). http://www.careerbuilder.com/Article/ CB-473-The-WorkplaceSurvey-One-in-Five-Workers-Lie/ (accessed 30 June 2009). Hall, M. (1986). The lesbian corporate experience. Journal of Homosexuality, 12, 59–74. Hakstian, A.R., & Ng, E. (2005). Employment-related motivational distortion: Its nature, measurement, and reduction. Educational and Psychlogical Measurement, 65, 405–41. Hancock, D.R., & Flowers, C.P. (2001). Comparing social desirability responding on World Wide Web and paper-and-pencil surveys. Educational Technology, Research and Development, 49, 5–13. Handel, M.I. (1982). Intelligence and deception. In J. Gooch & A. Perlmutter (Eds), Military Deception and Strategic Surprise (pp. 122–54). London: Rutledge. Handel, M.I. (1989). Military Deception in Peace and War. Jerusalem: Magnes Press. Harold, C.M., McFarland, L.A., & Weekley, J.A. (2006). The validity of verifiable and non-verifiable biodata items: An examination across applicants and incumbents. International Journal of Selection and Assessment, 14, 336–46. Hartshorne, H., & May, M.A. (1928). Studies on Deceit. New York: Macmillan. Heggestad, E.D., Morrison, M., Reeve, C.L., & McCloy, R.A. (2006). Forced-choice assessments of personality for selection: Evaluating issues of normative assessment and faking resistance. Journal of Applied Psychology, 91, 9–24. Heine, S.J., & Hamamura, T. (2007). In search of East Asian self-enhancement. Personality and Social Psychology Review, 11, 4–27. Heine, S.J., & Lehman, D.R. (1995). Cultural variation in unrealistic optimism: Does the West feel more invulnerable than the East? Journal of Personality and Social Psychology, 68, 595–607. Helmes, E., & Holden, R.R. (2003). The construct of social desirability: One or two dimensions? Personality and Individual Differences, 34, 1015–23. Henry, M.S., & Raju, N.S. (2006). The effects of traited and situational impression management on a personality test: An empirical analysis. Psychology Science, 48, 247–67. Heron, A. (1956). The effects of real-life motivation on questionnaire response. Journal of Applied Psychology, 40, 65–8. Heuer, R.J. Jr. (1982). Cognitive factor in deception and counterdeception. In D.C. Daniel & K.L. Herbig (Eds), Strategic Military Deception (pp. 31–69). New York: Pergamon. Hewlin, P.F. (2009). Wearing the cloak: Antecedents and consequences of creating facades of conformity. Journal of Applied Psychology, 9, 727–41. Hicks, L.E. (1970). Some properties of ipsative, normative, and forced-choice normative measures. Psychological Bulletin, 74, 167–84. Hilton, J.L., Fein, S., & Miller, D.T. (1993). Suspicion and dispositional influence. Personality and Social Psychology Bulletin, 19, 501–12. Hirsch, J.B., & Peterson, J.B. (2008). Predicting creativity and academic success with a “fake-proof ” measure of the Big Five. Journal of Research in Personality, 42, 1323–33.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
281
Hoffman, B.J., Blair, C.A., Meriac, J.P., & Woehr, D.J. (2007). Expanding the criterion domain? A quantitative review of the OCB literature. Journal of Applied Psychology, 92, 555–66. Hofstede, G. (1983). National cultures in four dimensions. International Studies of Management and Organizations, 13, 46–74. Hogan, J., Barrett, P., & Hogan, R. (2007). Personality measurement, faking, and employment selection. Journal of Applied Psychology, 92, 1270–85. Holden, R.R., Kroner, D.G., Fekkern, G.C., & Popham, S.M. (1992). A model of personality test item response dissimulation. Journal of Personality and Social Psychology, 63, 272–9. Hopper, R., & Bell, R.A. (1984). Broadening the deception construct. Quarterly Journal of Speech, 70, 288–302. Hough, L.M. (1998). Effects of intentional response distortion in personality measurement and evaluation of suggested palliatives. Human Performance, 11, 209–44. Hough, L.M., Eaton, N.K., Dunnette, M.D., Kamp, J.D., & McCloy, R.A. (1990). Criterion-related validities of personality constructs and the effect of response distortion on those validities. Journal of Applied Psychology, 75, 581–95. Hurtz, G.M., & Alliger, G.M. (2002). Influence of coaching on integrity test performance and unlikely virtues scale scores. Human Performance, 15, 255–73. Hyman, R. (1989). The psychology of deception. Annual Review of Psychology, 40, 133–54. Iacono, W.G. (2008). Effective policing: Understanding how polygraph tests work and are used. Criminal Justice and Behavior, 35, 1295–308. Iacono, W.G., & Lykken, D.T. (1997). The validity of the lie detector: Two surveys of scientific opinion. Journal of Applied Psychology, 82, 426–33. Inbau, F.E., Reid, J.E., Buckley, J.P., & Jayne, B.C. (2004). Criminal Interrogation and Confessions (4th edn). Boston, MA: Jones and Bartlett. Iwawaki, S., & Cohen, E.L. (1964). The social desirability of trait-descriptive terms: Applications to a Japanese sample. Journal of Social Psychology, 63, 199–205. Jackson, D.N., & Messick, S. (1958). Content and style in personality assessment. Psychological Bulletin, 55, 243–52. Jackson, D.N., Wroblewski, V.R., & Ashton, M.C. (2000). The impact of faking on employment tests: Does forced choice offer a solution? Human Performance, 13, 371–88. Jacobs, S., Dawson, E.J., & Brashers, D. (1996). Information manipulation theory: A replication and assessment. Communication Monographs, 63, 70–82. Jensen, L.A., Arnett, J.J., Feldman, S.S., & Cauffman, E. (2002). It’s wrong, but everybody does it: Academic dishonesty among high school students and college students. Contemporary Educational Psychology, 27, 209–28. Jobst, P., Hinshaw, K., & Harris, D. (2009). Admissions and selection testing: New techniques for facing test security challenges and protecting program integrity. 10th Annual ATP Innovations in Testing Palm Springs, CA. Jones, J.W. (1981). Dishonesty, burnout, and unauthorized work break extensions. Personality and Social Psychology Bulletin, 7, 406–9. Jones, J.W., Brasher, E.E., & Huff, J.W. (2002). Innovations in integrity-based personnel selection: Building a technology-friendly assessment. International Journal of Selection and Assessment, 10, 87–97. Jones, G.R., & George, J.M. (1998). The experience and evolution of trust: Implications for cooperation and teamwork. Academy of Management Review, 23, 531–46. Jones, E.E., & Pittman, T.S. (1982). Toward a general theory of strategic selfpresentation. In J. Suls (Ed.), Psychological Perspectives on the Self (volume 1), (pp. 231–62). Hillsdale, NJ: Lawrence Erlbaum Associates.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
282 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Kashy, D.A., & DePaulo, B.M. (1996). Who lies? Journal of Personality and Social Psychology, 70, 1037–51. Khorramdel, L., & Kubinger, K.D. (2006). The effect of speediness on personality questionnaires: An experiment on applicants with a job recruiting procedure. Psychology Science, 48, 378–97. Kilcullen, R.N., White, L.A., Mumford, M.D., & Mack, H. (1995). Assessing the construct validity of rational biodata scales. Military Psychology, 7, 17–28. Kim, B.H. (2007). Building a theory of strategic deception. Doctoral dissertation, Michigan State University, 2007. Dissertation Abstracts International, 68, 660. Kim, S.C. (1985). Family therapy for Asian Americans: A strategic-structural framework. Psychotherapy, 22, 342–8. Kim, P.H., Ferrin, D.L., Cooper, C.D., & Dirks, K.T. (2004). Removing the shadow of suspicion: The effects of apology versus denial for repairing competence- versus integrity-based trust violations. Journal of Applied Psychology, 89, 104–18. Kisamore, J.L., Stone, T.H., & Jawahar, I.M. (2007). Academic integrity: The relationship between individual and situational factors on misconduct contemplations. Journal of Business Ethics, 75, 381–94. Klein, S.P., & Owens, W.A. Jr. (1965). Faking of a scored life history blank as a function of criterion objectivity. Journal of Applied Psychology, 49, 452–4. Klimoski, R.J., & Carol, B.L. (1976). The impact of trust on creative problem solving groups. Journal of Applied Psychology, 61, 630–3. Kluger, A.N., & Colella, A. (1993). Beyond the mean bias: The effect of warning against faking on biodata item variances. Personnel Psychology, 46, 763–80. Kluger, A.N., Reilly, R.R., & Russell, C.J. (1991). Faking biodata tests: Are optionkeyed instruments more resistant? Journal of Applied Psychology, 76, 889–96. Komar, S., Brown, D.J., Komar, J.A., & Robie, C. (2008). Faking and the validity of conscientiousness: A monte carlo investigation. Journal of Applied Psychology, 93, 140–54. Komarraju, M., Dollinger, S.J., & Lovell, J.L. (2008). Individualism-collectivism in horizontal and vertical directions as predictors of conflict management styles. International Journal of Conflict Management, 19, 20–35. Kroger, R.O., & Turnbull, W. (1975). Invalidity of validity scales: The case of MMPI. Journal of Consulting and Clinical Psychology, 43, 48–55. Kuncel, N.R., & Borneman, M.J. (2007). Personality tests: The use of idiosyncratic item responses. International Journal of Selection and Assessment, 15, 220–31. Kuncel, N.R., & Tellegen, A. (2009). A conceptual and empirical reexamination of the measurement of social desirability of items: Implications for detecting desirable response style and scale development. Personnel Psychology, 62, 201–28. LaFreni`ere, P. (1988). The ontogeny of tactical deception in humans. In R.W. Byrne & A. Whiten (Eds), Machiavellian Intelligence: Social Expertise and the Evolution of Intellect in Monkeys, Apes, and Humans (pp. 238–52). Oxford, UK: Clarendon Press. LaHuis, D.M., & Copeland, D. (2009). Investigating faking using a multilevel logistic regression approach to measuring person fit. Organizational and Research Methods, 12, 293–319. Lalwani, A.K., Shavitt, S., & Johnson, T. (2006). What is the relation between cultural orientation and socially desirable responding? Journal of Personality and Social Psychology, 90, 165–78. Langleben, D.D., Loughead, J.W., Bilker, W.B., Ruparel, K. Childress, A.R., Busch, S.I., et al. (2005). Telling truth from lie in individual subjects with fast event-related fMRI. Human Brain Mapping, 26, 262–72. Lau, D.C., & Liden, R.C. (2008). Antecedents of coworker trust: Leaders’ blessings. Journal of Applied Psychology, 93, 1130–8.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
283
Law, K.S., Mobley, W.H., & Wong, C. (2002). Impression management and faking in biodata scores among Chinese job-seekers. Asia Pacific Journal of Management, 19, 541–56. Leary, M.R., & Kowalski, R.M. (1990). Impression mnagement: A literature review and two-component model. Psychological Bulletin, 107, 34–47. LeBreton, J.M., Barksdale, C.D., Robin, J.D., & James, L.R. (2007). Measurement issues associated with conditional reasoning tests: Indirect measurement and test faking. Journal of Applied Psychology, 92, 1–16. Lee, S. (2007). Revelations of false credentials shake South Korea. New York Times (1 September 2007). http://www.nytimes.com/2007/09/01/world/asia/ 01korea.html (accessed 30 June 2009). Lee, S.T. (2004). Lying to tell the truth: Journalists and the social context of deception. Mass Communication and Society, 7, 97–120. Leite, W.L., & Beretvas, S.N. (2005). Validation of scores on the Marlowe–Crowne Social Desirability Scale and the Balanced Inventory of Desirable Responding. Educational and Psychological Measurement, 65, 140–54. Levashina, J., & Campion, M.A. (2006). A model of faking likelihood in the employment interview. International Journal of Selection and Assessment, 14, 299–316. Levashina, J., & Campion, M.A. (2007). Measuring faking in the employment interview: Development and validation of an interview faking behavior scale. Journal of Applied Psychology, 92, 1638–56. Levin, R.A., & Zickar, M.J. (2002). Investigating self-presentation, lies, and bullshit: Understanding faking and its effects on selection decisions using theory, field research, and simulation. In J.M. Brett & F. Drasgow (Eds), The Psychology of Work (pp. 253–76). Mahwah, NJ: Lawrence Erlbaum Associates. Levin, D.Z., Whitener, E.M., & Cross, R. (2006). Perceived trustworthiness of knowledge sources: The moderating impact of relationship length. Journal of Applied Psychology, 91, 1163–71. Levine, T.R., Park H.S., & McCornack, S.A. (1999). Accuracy in detecting truths and lies: Documenting the “veracity effect.” Communication Monographs, 66, 199– 218. Lewicki, R.J. (1983). Lying and deception: A behavioral model. In M. Bazerman & R. Lewicki (Eds), Negotiating in Organizations. Newbury Park, CA: Sage. Lewicki, R.J., & Bunker, B.B. (1996). Developing and maintaining trust in work relationships. In R.M. Kramer & T.R. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research (pp. 114–39). Thousand Oaks, CA: Sage. Lewicki, R.J., & Stark, N. (1996). What is ethically appropriate in negotiations: An empirical examination of bargaining tactics. Social Justice Research, 9, 69–95. Lewin, T. (2007). Dean at M.I.T. resigns, ending a 28-year lie. New York Times (27 April 2007) http://www.nytimes.com/2007/04/27/us/27mit.html (accessed 30 June 2009). Lewis, M., Stanger, C., & Sullivan, M.W. (1989). Deception in 3-year-olds. Developmental Psychology, 25, 439–43. Li, A., & Bagger, J. (2006). Using the BIDR to distinguish the effects of impression management and self-deception on the criterion validity of personality measures: A meta-analysis. International Journal of Selection and Assessment, 14, 131–41. Li, A., & Bagger, J. (2007). The Balanced Inventory of Desirable Responding (BIDR): A reliability generalization study. Educational and Psychological Measurement, 67, 525–44. Li, A., & Reb, J. (2009). A cross-nations, cross-cultures, and cross-conditions analysis on the equivalence of the Balanced Inventory of Desirable Responding. Journal of Cross-Cultural Psychology, 40, 214–33.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
284 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Lievens, F., De Corte, W., & Schollaert, E. (2008). A closer look at the frame-ofreference effect in personality scale scores and validity. Journal of Applied Psychology, 93, 268–79. Lievens, F., & Peeters, H. (2008). Impact of elaboration on responding to situational judgment items. International Journal of Selection and Assessment, 16, 345–55. Lombroso, C. (1895). L’Homme Crimine (2nd edn). Lopes, J., & Fletcher, C. (2004). Fairness of impression management in employment interviews: A cross-country study of the role of equity and Machiavellianism. Social Behavior and Personality, 32, 747–68. MacLaren, V.V. (2004). A quantitative review of the Guilty Knowledge Test. Journal of Applied Psychology, 86, 674–83. Marcus, B. (2006). Relationships between faking, validity, and decision criteria in personnel selection. Psychology Science, 48, 226–46. Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 90, 339–63. Marshall, M.B., De Fruyt, F., Rolland, J., & Bagby, R.M. (2005). Socially desirable responding and the factorial stability of the NEO PI-R. Journal of Applied Psychology, 17, 379–84. Marston, W.M. (1917). Systolic blood pressure symptoms of deception. Journal of Experimental Psychology, 2, 117–63. Martin, B.A., Bowen, C.C., & Hunt, S.T. (2002). How effective are people at faking on personality questionnaires? Personality and Individual Differences, 32, 247–56. Mayer, R.C., Davis, J.H., & Schoorman, F.D. (1995). An intergrative model of organizational trust. Academy of Management Review, 20, 709–34. McCabe, D.L. (2005). Cheating among college and university students: A North American perspective. International Journal for Educational Integrity, 1. http://www.ojs.unisa.edu.au/index.php/IJEI/index (accessed 1 November 2010). ˜ L.K. (2001). Cheating in academic inMcCabe, D.L., Butterfield, K.D., & Trevino, stitutions: A decade of research. Ethics & Behavior, 11, 219–32. ˜ L.K. (2006). Academic dishonesty in McCabe, D.L., Butterfield, K.D., & Trevino, graduate business programs: Prevalence, causes, and proposed action. Academy of Management Learning and Education, 5, 294–305. McCabe, D.L., & Trevino, L.K. (1993). Academic dishonesty: Honor codes and other contextual influences. Journal of Higher Education, 64, 522–38. McCabe, D.L., & Trevino, L.K. (1997). Individual and contextual influences on academic dishonesty: A multicampus investigation. Research in Higher Education, 38, 379–96. McCloy, R.A., Heggestad, E.D., & Reeve, C.L. (2005). A silk purse from a sow’s ear: Retrieving normative information from multidimensional forced-choice items. Organizational Research Methods, 8, 222–48. McCornack, S.A. (1997). The generation of deceptive messages: Laying the groundwork for a viable theory of interpersonal deception. In J.O. Greene (Ed.), Message Production: Advances in Communication Theory (pp. 91–126). Hillsdale, NJ: Lawrence Erlbaum. McCornack, S.A., & Levine, T.R. (1990). When lies are uncovered: Emotional and relational outcomes of discovered deception. Communication Monographs, 57, 119– 38. McCrae, R.R., & Costa, P.T. (1983). Social desirability scales: More substance than style. Journal of Consulting and Clinical Psychology, 51, 882–8. McFarland, L.A. (2003). Warning against faking on a personality test: Effects on applicant reactions and personality test scores. International Journal of Selection and Assessment, 11, 265–76.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
285
McFarland, L.A., & Ryan, A.M. (2000). Variance in faking across noncognitive measures. Journal of Applied Psychology, 85, 812–21. McFarland, L.A., & Ryan, A.M. (2001). Toward an integrated model of applicant faking behavior. Paper presented at the 2001 conference of the Society for Industrial and Organizational Psychology. San Diego, CA. McFarland, L.A., & Ryan, A.M. (2006). Toward an integrated model of applicant faking behavior. Journal of Applied Social Psychology, 36, 979–1016. McFarland, L.A., Ryan, A.M., & Ellis, A. (2002). Item placement on a personality measure: Effects on faking behavior and test measurement properties. Journal of Personality Assessment, 78, 348–69. McManus, M.A., & Masztal, J.J. (1999). The impact of biodata item attributes on validity and socially desirable responding. Journal of Business and Psychology, 13, 437–46. McLeod, B.A., & Genereux, R.L. (2008). Predicting the acceptability and likelihood of lying: The interaction of personality with type of lie. Personality and Individual Differences, 45, 591–6. Meade, A.W. (2004). Psychometric problems and issues involved with creating and using ipsative measures for selection. Journal of Occupational and Organizational Psychology, 77, 531–52. Meehl, P.E., & Hathaway, S.R. (1946). The K factor as a suppressor variable in the Minnesota Multiphasic Personality Inventory. Journal of Applied Psychology, 30, 525– 64. Merkin, R.S. (2006). Uncertainty avoidance and facework: A test of the Hofstede model. International Journal of Intercultural Relations, 30, 213–28. Mesmer-Magnus, J., Viswesvaran, C., Deshpande, S., & Joseph, J. (2006). Social desirability: The role of over-claiming, self-esteem, and emotional intelligence. Psychology Science, 48, 336–56. Middleton, K.L., & Jones, J.L. (2000). Socially desirable response sets: The impact of country culture. Psychology and Marketing, 17, 149–63. Miller, G.R. (1983). Telling it like it isn’t and not telling it like it is: Some thoughts on deceptive communication. In J.I. Sisco (Ed.), The Jensen Lectures: Contemporary Communication Studies (pp. 91–116). Tampa, FL: University of South Florida. Miller, A., Shoptaugh, C., & Parkerson, A. (2008). Under reporting of cheating in research using volunteer college students. College Student Journal, 42, 326–39. Miller, G.R., & Stiff, J.B. (1993). Deceptive Communication. Newbury Park, CA: Sage. Mishra, A.K. (1996). Organizational response to crisis: The centrality of trust. In R.M. Kramer & T.R. Tyler (Eds), Trust in Organizations: Frontiers of Theory and Research (pp. 261–87). Thousand Oaks, CA: Sage. Montagliani, A., & Giacalone, R.A. (1998). Impression management and cross-cultural adaptation. Journal of Social Psychology, 138, 598–608. Moorman, R.H., & Podsakoff, P.M. (1992). Journal of Occupational and Organizational Psychology, 65, 131–49. Morgeson, F.P., Campion, M.A., Dipboye, R., Hollenbeck, J.R., Murphy, K., & Schmitt, N. (2007). Reconsidering the use of personality tests in personnel selection contexts. Personnel Psychology, 60, 683–729. Mosel, J.N., & Cozan, L.W. (1952). The accuracy of application blank work histories. Journal of Applied Psychology, 36, 365–9. Motowidlo, S.J., Borman, W.C., & Schmit, M.J. (1997). A theory of individual difference in task and contextual performance. Human Performance, 10, 71–83. Mueller-Hanson, R.A., Heggestad, E.D., Thornton, G.C. III. (2003). Faking and selection: Considering the use of personality from select-in and select-out perspectives. Journal of Applied Psychology, 88, 348–55.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
286 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Mueller-Hanson, R.A., Heggestad, E.D., Thornton, G.C. III. (2006). Individual differences in impression management: An exploration of the psychological processes underlying faking. Psychology Science, 3, 288–312. Munsterberg, H. (1908). On the Witness Stand: Essays on Psychology and Crime. New York: Doubleday. Murdock, T.B., & Anderman, E.M. (2006). Motivational perspectives on student cheating: Toward an integrated model of academic dishonesty. Educational Psychologist, 41, 129–45. Murphy, K.R. (1993). Honesty in the Workplace. Pacific Grove, CA: Brooks/Cole. Mwamwenda, T.S. (1996). Social desirability scores of South African and Canadian students. Psychological Reports, 78, 723–6. Myers, B., Latter, R., & Abdollahi-Arena, M.K. (2006). The court of public opinion: Lay perceptions of polygraph testing. Law and Human Behavior, 30, 509–23. Naglieri, J.A., Drasgow, F., Schmit, M., Handler, L., Prifitera, A., Margolis, A., et al. (2004). Psychological testing on the Internet: New problems, old issues. American Psychologist, 59, 150–69. National Research Council. (2003). The Polygraph and Lie Detection. Washington, D.C.: National Research Council. Newell, A., & Simon, H. (1972). Human Problem Solving. Englewood Cliffs, NJ: Prentice-Hall. Newman, M.L., Pennebaker, J.W., Berry, D.S., & Richards, J.M. (2003). Lying words: Predicting deception from linguistic styles. Personality and Social Psychology Bulletin, 29, 665–75. Newmark, N., Adityanjee, & Kay, J. (1999). Pseudologia fantastica and factitious disorder: review of the literature and a case report. Comprenhsive Psychiatry, 40, 89–95. Newstead, S.E., Franklyn-Stokes, A., & Armstead, P. (1996). Individual differences in student cheating. Journal of Educational Psychology, 88, 229–41. Nguyen, N.T., Biderman, M.D., & McDaniel, M.A. (2005). Effects of response instructions on faking a situational judgment test. International Journal of Selection and Assessment, 13, 250–60. Nye, C.D., Do, B., Drasgow, F., & Fine, S. (2008). Two-step testing in employee selection: Is score inflation a problem? International Journal of Selection and Assessment, 16, 112–20. O’Bannon, R.M., Goldfinger, L.A., & Appleby, J.D. (1989). Honesty and Integrity Testing: A Practical Guide. Atlanta, GA: Applied Information Resources. O’Connell, V. (2009). Test for dwindling retail jobs spawns a culture of cheating. http://online.wsj.com/article/SB123129220146959621.html (accessed 16 July 2009). O’Leary-Kelly, A., Lean, E., Reeves, C., & Randel, J. (2008). Coming into the light: Intimate partner violence and its effects at work. Academy of Management Perspectives, 22, 57–72. Offe, H., & Offe, S. (2007). The Comparison Question Test: Does it work and if so how? Law and Human Behavior, 31, 291–303. Ones, D.S., Dilchert, S., Viswesvaran, C., & Judge, T.A. (2007). In support of personality assessment, in organizational settings. Personnel Psychology, 60, 995–1027. Ones, D.S., & Viswesvaran, C. (1998). The effects of social desirability and faking on personality and integrity assessment for personnel selection. Human Performance, 11, 245–69. Ones, D.S., Viswesvaran, C., & Reiss, A.D. (1996). Role of social desirability in personality testing for personnel selection: A red herring. Journal of Applied Psychology, 81, 660–79.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
287
Ones, D.S., Viswesvaran, C., & Schmidt, F.L. (1993). Comprehensive meta-analysis of integrity test validities: Findings and implications for personnel selection and theories of job performance. Journal of Applied Psychology, 78, 679–703. Organ, D.W. (1997). Organizational citizenship behavior: It’s construct clean-up time. Human Performance, 10, 85–97. Osberg, T.M., & Harrigan, P. (1999). Comparative validity of the MMPI-2 WienerHarmon Subtle-Obvious scales in male prison inmates. Journal of Personality Assessment, 72, 36–48. Pankratz, L., & Binder, L.M. (1997). Malingering on intellectual and neuropsychological measures. In R. Rogers (Ed.), Clinical Assessment of Malingering and Deception (pp. 223–36). New York: Guilford Press. Pannone, R.D. (1984). Predicting test performance: A content valid approach to screening applicants. Personnel Psychology, 37, 507–14. Paulhus, D.L. (2002). Socially desirable responding. The evolution of a construct. In H.I. Braun, D.N. Jackson, & D.E. Wiley (Eds), The Role of Constructs in Psychological and Educational Measurement (pp. 49–69). Mahwah, NJ: Erlbaum. Paulhus, D.L., Bruce, M.N., & Trapnell, P.D. (1995). Effects of self-presentation strategies on personality profiles and their structure. Personality and Social Psychology Bulletin, 21, 100–8. Pauls, C.A., & Crost, N.W. (2004). Effects of faking on self-deception and impression management scales. Personality and Individual Differences, 37, 1137–51. Peeters, H., & Lievens, F. (2005). Situational judgment tests and their predictiveness of college students’ success: The influence of faking. Educational and Psychological Measurement, 65, 70–89. Pennebaker, J.W., Mehl, M.R., & Niederhoffer, K.G. (2003). Psychological aspects of natural language use: Our words, our selves. Annual Review of Psychology, 54, 547– 77. Peterson, M.H., & Griffith, R.L. (2006). Faking and job performance: A multi-faceted issue. In R.L. Griffith & M.H. Peterson (Eds), A Closer Examination of Applicant Faking Behavior (pp. 233–61). Greenwich, CT: Information Age. Pino, N.W., & Smith, W.L. (2003). College students and academic dishonesty. College Student Journal, 37, 490–500. Porter, S., & ten Brinke, L. (2008). Reading between the lies: Identifying concealed and falsified emotions in universal facial expressions. Psychological Science, 19, 508–14. Porter, S., & ten Brinke, L. (2009). The truth about lies: What works in detecting high-stakes deception? Legal and Criminological Psychology, 15, 57–75. Premack, D. (1988). ‘Does the chimpanzee have a theory of mind’ revisited. In R. Byrne & A. Whiten (Eds), Machiavellian Intelligence: Social Expertise and the Evolution of Intellect in Monkeys, Apes, and Humans (pp. 160–79). Oxford, UK: Clarendon Press. Ramsay, L.J., Schmitt, N., Oswald, F.L., Kim, B.H., & Gillespie, M.A. (2006). The impact of situational context variables on responses to biodata and situational judgment inventory items. Psychology Science, 48, 268–87. Randall, D.M., Huo, Y.P., & Pawelk, P. (1993). Social desirability bias in cross-cultural ethics research. International Journal of Organizational Analysis, 1, 185–202. Raymark, P.H., & Tafero, T.L. (2009). Individual differences in the ability to fake on personality measures. Human Performance, 22, 86–103. Reid-Seiser, H.L., & Fritzsche, B.A. (2001). The usefulness of the NEO PI-R Positive Presentation Management Scale for detecting response distortion in employment contexts. Personality and Individual Differences, 31, 639–50. Riggio, R.E., Salinas, C., & Tucker, J. (1988). Personality and deception ability. Personality and Individual Differences, 9, 189–91.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
288 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Riggio, R.E., Tucker, J., & Throckmorton, B. (1987). Social skills and deception ability. Personality and Social Psychology Bulletin, 13, 568–77. Robie, C. (2006). Effects of perceived selection ratio on personality test faking. Social Behavior and Personality, 34, 1233–44. Robie, C., Tuzinski, K.A., & Bly, P.R. (2006). A survey of assessor beliefs and practices related to faking. Journal of Managerial Psychology, 21, 669–81. Robson, S.M., Jones, A., & Abraham, J. (2008). Personality, faking and convergent validity: A warning concerning warning statements. Human Performance, 21, 89– 106. Rodin, M. (1972). A further look at social desirability. Psychonomic Science, 29, 103–5. Rogers, R. (1984). Towards an empirical model of malingering and deception. Behavioral Sciences and the Law, 2, 93–111. Rogers, R. (1997). Clinical Assessment of Malingering and Deception (2nd edn). New York: Guilford Press. Rogers, R., Gillis, J.R., Bagby, R.M., & Monteiro, E. (1991). Detection of malingering on the Structured Interview of Reported Symptoms (SIRS): A study of coached and uncoached simulators. Psychological Assessment, 3, 673–7. Rosenfeld, J.P. (2005). ‘Brain Fingerprinting’: A critical analysis. Scientific Review of Mental Health Practice, 4. http://www.fbi.gov/hq/lab/fsc/backissu/july2003/ podlesny.htm (accessed 30 June 2009). Rosse, J.G., Stecher, M.D., Miller, J.L., & Levin, R.A. (1998). The impact of response distortion on preemployment personality testing and hiring decisions. Journal of Applied Psychology, 83, 634–44. Rothstein, M.G., & Goffin, R.D. (2006). The use of personality measures in personnel selection: What does the research support? Human Resource Management Review, 16, 155–80. Rousseau, D.M., Sitkin, S.B., Burt, R.S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23, 393–404. Rowatt, W.C., Cunningham, M.R., & Druen, P.B. (1999). Facial physical attractiveness on the willingness to deceive prospective dating partners. Journal of Social and Personal Relationships, 16, 209–23. Rozell, E.J., & Gundersen, D.E. (2003). The effects of leader impression management on group perceptions of cohesion, consensus, and communication. Small Group Research, 34, 197–222. Saarni, C. (1984). An observational study of children’s attempts to monitor their expressive behavior. Child Development, 55, 1504–13. Sackeim, H.A., & Gur, R. (1978). Self-deception, self-confrontation, and consciousness. In G.E. Schwartz & D. Shaprio (Eds), Consciousness and Self-Regulation: Advances in Research (volume 2), (pp. 139–97). New York: Plenum Press. Sackett, P.R., Burris, L.R., & Callahan, C. (1989). Integrity testing for personnel selection: An update. Personnel Psychology, 42, 491–529. Sackett, P.R., Burris, L.R., & Ryan, A.M. (1989). Coaching and practice effects in personnel selection. In C.L. Cooper & I. Robertson (Eds), International Review of Industrial and Organizational Psychology. Chichester, NY: John Wiley & Sons. Sackett, P.R., & Decker, P.J. (1979). Detection of deception in the employment context: A review and critical analysis. Personnel Psychology, 32, 487–506. Sackett, P.R., & Wanek, J.E. (1996). New developments in the use of measures of honesty, integrity, conscientiousness, dependability, trustworthiness, and reliability for personnel selection. Personnel Psychology, 49, 787–829. Salamon, S.D., & Robinson, S.L. (2008). Trust that binds: The impact of collective felt trust on organizational performance. Journal of Applied Psychology, 93, 593– 601.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
289
Sandal, G.M., & Endresen, I.M. (2002). The sensitivity of the CPI Good Impression Scale for detecting ‘faking good’ among Norwegian students and job applicants. International Journal of Selection and Assessment, 10, 304–11. Schab, F. (1991). Schooling without learning: Thirty years of cheating in high school. Adolescence, 26, 839–47. Schmit, M.J., & Ryan, A.M. (1993). The Big Five in personnel selection: Factor structure in applicant and non-applicant populations. Journal of Applied Psychology, 78, 966–74. Schmit, M.J., Ryan, A.M., Stierwalt, S.L., & Powell, A.B. (1995). Frame-of-reference effects on personality scale scores and criterion-related validity. Journal of Applied Psychology, 80, 607–20. Schmitt, N., & Kunce, C. (2002). The effects of required elaboration of answers to biodata questions. Personnel Psychology, 55, 569–87. Schmitt, N., & Oswald, F.L. (2006). The impact of corrections for faking on the validity of noncognitive measures in selection settings. Journal of Applied Psychology, 91, 613–21. Schmitt, N., Oswald, F.L., Kim, B.H., Gillespie, M.A., Ramsay, L.J., & Yoo, T. (2003). Impact of elaboration on social desirability and the validity of biodata measures. Journal of Applied Psychology, 88, 979–88. Schneider, B. (1987). The people make the place. Personnel Psychology, 40, 437–53. Schrader, A.D., & Osburn, H.G. (1977). Biodata faking: Effects of induced subtlety and position specificity. Personnel Psychology, 30, 395–404. Scott, E., & Jehn, K. (2003). Multiple stakeholder judgments of employee behaviors: A contingent prototype model of dishonesty. Journal of Business Ethics, 46, 235– 50. Seymour, T.L., & Kerlin, J. R. (2008). Successful detection of verbal and visual concealed knowledge using an RT-based paradigm. Applied Cognitive Psychology, 22, 475–90. Shulsky, A. (2002). Elements of strategic denial and deception. In R. Godson & J.J. Wirtz (Eds), Strategic Denial and Deception: The Twenty-first Century Challenge (pp. 15–32). New Brunswick, NJ: Transaction. Simons, T.L., & Peterson, R.S. (2000). Task conflict and relationship conflict in top management teams: The pivotal role of intragroup trust. Journal of Applied Psychology, 85, 102–11. Smith, D.B., & Ellingson, J.E. (2002). Substance versus style: A new look at social desirability in motivating contexts. Journal of Applied Psychology, 87, 211–9. Smith, D.B., Hanges, P.J., & Dickson, M.W. (2001). Personnel selection and the fivefactor model: Reexamining the effects of applicant’s frame of reference. Journal of Applied Psychology, 86, 304–15. Snell, A.F., Sydell, E.J., & Lueke, S.B. (1999). Towards a theory of applicant faking: Integrating studies of deception. Human Resource Management Review, 9, 219–42. Sohlberg, S.C. (1976). Social desirability responses in Jewish and Arab children in Israel. Journal of Cross-Cultural Psychology, 7, 301–14. Sommerville, Q. (2009). China’s hi-tech answer to cheats. BBC. http://news.bbc.co.uk/ 2/hi/asia-pacific/8084582.stm (accessed 30 June 2009). Sotaridona, L.S., & Meijer, R.R. (2003). Two new statistics to detect copying. Journal of Educational Measurement, 40, 53–69. Sotaridona, L.S., van der Linden, W.J., & Meijer, R.R. (2006). Detecting answer copying using the kappa statistic. Applied Psychological Measurement, 30, 412–31. Spence, S.A., & Kaylor-Hughes, C.J. (2008). Looking for the truth and finding lies: The prospects for a nascent neuroimaging of deception. Neurocase, 14, 68– 81.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
290 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Stark, S., Chernyshenko, O.S., Chan, K., Lee, W.C., & Drasgow, F. (2001). Effects of the testing situation on item responding: Cause for concern. Journal of Applied Psychology, 86, 943–53. Steffens, M.C. (2004). Is the Implicit Association Test immune to faking? Experimental Psychology, 51, 165–79. Steinberg, J. (2002). Officials link foreign web sites to cheating on graduate admissions exams. New York Times. http://www.nytimes.com/2002/08/08/us/officials-linkforeign-web-sites-to-cheating-on-graduate-admission-exams.html (accessed 30 June 2009). Stening, B.W., & Everett, J.E. (1984). Response styles in a cross-cultural managerial study. Journal of Social Psychology, 122, 151–6. Stephens, J.M., Young, M.F., & Calabrese, T. (2007). Does moral judgment go offline when students are online? A comparative analysis of undergradutes’ beliefs and behaviors related to conventional and digital cheating. Ethics and Behavior, 17, 233–54. Stokes, G.S., Hogan, J.B., & Snell, A.F. (1993). Comparability of incumbent and applicant samples for the development of biodata keys: The influence of social desirability. Personnel Psychology, 46, 739–62. Stokes, G.S., & Toth, C.S. (1996). Background data for personnel selection. In R.S. Barrett (Ed.), Fair Employment Strategies in Human Resource Management (pp. 171–9). Westport, CT: Quorom Books. Storch, E.A., & Storch, J.B. (2002). Fraternities, sororities, and academic dishonesty. College Student Journal, 36, 247–52. Storm, J., & Graham, J.R. (2000). Detection of coached general malingering on the MMPI-2. Psychological Assessment, 12, 158–65. Tangney, J.P. (1992). Situational determinants of shame and guilt in young adulthood. Personality and Social Psychology Bulletin, 18, 199–206. Taylor, L., Gitttes, M., O’Neal, E.C., & Brown, S. (1994). The reluctance to expose dangerous lies. Journal of Applied Social Psychology, 24, 301–15. Tett, R.P., Anderson, M.G., Ho, C., Yang, T.S., Huang, L., & Hanvongse, A. (2006). Seven nested questions about faking on personality tests: An overview and interactions model of item-level response distortion. In R.L. Griffith & M.H. Peterson (Eds), A Closer Examination of Applicant Faking Behavior (pp. 43–84). Greenwich, CT: Information Age. Tett, R.P., & Christiansen, N.D. (2007). Personality tests at the crossroads: A response to Morgeson, Campion, Dipboye, Hollenbeck, Murphy, and Schmitt (2007). Personnel Psychology, 60, 967–93. Tippins, N.T. (2009). Where is the unproctored Internet testing train headed now? Industrial and Organizational Psychology: Perspectives on Science and Practice, 2, 69–76. Tippins, N.T., Beatty, J., Drasgow, F., Gibson, W.M., Pearlman, K., Segall, D.O., et al. (2006). Unproctored Internet testing in employment settings. Personnel Psychology, 59, 189–225. Topping, G.D., & O’Gorman, J.G. (1997). Effects of faking set on validity of the NEO-FFI. Personality and Individual Differences, 23, 117–24. Triandis, H.C., Brislin, R., & Hui, C.H. (1988). Cross-cultural training across the individualism-collectivism divide. International Journal of Intercultural Relations, 12, 269–89. Triandis, H.C., McCusker, C., & Hui, C.H. (1990). Multimethod probes of individualism and collectivism. Journal of Personality and Social Psychology, 59, 1006–20. Tripathi, R.R., & Shahjahan, M. (1987). Constancy of social desirability judgments across Indian, Bangladeshi, and American samples. Psychological Studies, 32, 33–5. Trovillo, P.V. (1939). A history of lie detection. American Journal of Police Science, 29, 848–81.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
DECEPTION AND APPLICANT FAKING
291
Tsushima, W.T. (1969). Responses of Irish and Italians of two social classes on the Marlowe-Crowne Social Desirability Scale. Journal of Social Psychology, 77, 215– 9. Turner, R.E., Edgley, C., & Olmstead, G. (1975). Information control in conversations: Honesty is not always the best policy. Kansas Journal of Sociology, 11, 69–89. USA Today (2008). Chef fired after resume fraud. (3 March 2008). http://www.usatoday.com/news/nation/2008-03-03-chef-fired N.htm (accessed 30 June 2009). van der Linden, W.J., & Sotaridona, L.S. (2004). A statistical test for detecting answer copying on multiple-choice tests. Journal of Educational Measurement, 41, 361–77. Van Iddekinge, C.H., Raymark, P.H., & Roth, P.L. (2005). Assessing personality with a structured employment interview: Construct-related validity and susceptibility to response inflation. Journal of Applied Psychology, 90, 536–52. Vandehey, M.A., Diekhoff, G.M., & LaBeff, E.E. (2007). College cheating: A twentyyear follow-up and the addition of an honor code. Journal of College Student Development, 48, 468–80. Vasek, M.E. (1986). Lying as a skill: The development of deception in children. In R.W. Mitchell & N.S. Thompson (Eds), Deception: Perspectives on Human and Nonhuman deceit (pp. 271–92). Albany, NY: State University of New York Press. Vasilopoulos, N.L., Cucina, J.M., Dyomina, N.V., Morewitz, C.L., & Reilly, R.R. (2006). Forced-choice personality tests: A measure of personality and cognitive ability? Human Performance, 19, 175–99. Vasilopoulos, N.L., Cucina, J.M., & McElreath, J.M. (2005). Do warnings of response verification moderate the relationship between personality and cognitive ability? Journal of Applied Psychology, 90, 306–22. Vasilopoulos, N.L., Reilly, R.R., & Leaman, J.A. (2005). The influence of job familiarity and impression management on self-report measure scale scores and response latencies Journal of Applied Psychology, 85, 50–64. Viswesvaran, C., & Ones, D.S. (1999). Meta-analyses of fakability estimates: Implications for personality measurement. Educational and Psychological Measurement, 59, 197–210. Viswesvaran, C., Ones, D.S., & Hough, L.M. (2001). Do impression management scales in personality inventories predict managerial job performance ratings? International Journal of Selection and Assessment, 9, 277–89. Vohs, K.D., Baumeister, R.F., & Ciarocco, N.J. (2005). Self-regulation and selfpresentation: Regulatory resource depletion impairs impression management and effortful self-presentation depletes regulatory resources. Journal of Personality and Social Psychology, 88, 632–57. Volpe, R., Davidson, L., & Bell, M.C. (2008). Faculty attitudes and behaviors concerning student cheating. College Student Journal, 42, 164–75. Vrij, A., Edward, K., & Bull, R. (2001). Stereotypical verbal and nonverbal responses while deceiving others. Personality and Social Psychology Bulletin, 27, 889–909. Vroom, V.H. (1964). Work and Motivation. New York: Wiley. Walczyk, J.J., Roper, K.S., Seemann, E., & Humphrey, A.M. (2003). Cognitive mechanisms underlying lying to questions: Response time as a cue to deception. Applied Cognitive Psychology, 17, 755–74. Walczyk, J.J., Schwartz, J.P., Clifton, R., Adams, B., Wei, M., & Zha, P. (2005). Lying person-to-person about life events: A cognitive framework for lie detection. Personnel Psychology, 58, 141–70. Weiss, B., & Feldman, R.S. (2006). Looking good and lying to do it: Deception as an impression management strategy in job interviews. Journal of Applied Social Psychology, 36, 1070–86.
P1: OTA/XYZ P2: ABC c07 JWST046-Hodgkinson
January 21, 2011
13:3
Printer Name: Yet to Come
292 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Whaley, B. (1982). Toward a general theory of deception. In J. Gooch & A. Perlmutter (Eds), Military Deception and Strategic Surprise (pp. 178–92). London, UK: Rutledge. Whiten, A., & Byrne, R.W. (1988). Tactical deception in primates. Behavioural and Brain Sciences, 11, 233–44. Whitley, B.E. Jr. (1998). Factors associated with cheating among college students: A review. Research in Higher Education, 39, 235–74. Williams, A.E., & Janosik, S.M. (2007). An examination of academic dishonesty among sorority women and nonsorority women. Journal of College Student Development, 48, 706–14. Wilson, S.R. (1997). Developing theories of persuasive message production: The next generation. In J.O. Greene (Ed.), Message Production: Advances in Communication Theory (pp. 15–43). Hillsdale, NJ: Lawrence Erlbaum. Wimmer, H. & Perner, J. (1983). Belief about beliefs: Representation and constraining functions of wrong beliefs in young children’s understanding of deception. Cognition, 13, 103–28. Winkelspecht, C., Lewis, P., & Thomas, A. (2006). Potential effects of faking on the NEO-P-I-R: Willingness and ability to fake changes who gets hired in simulated selection decisions. Journal of Business and Psychology, 21, 243–59. Wollack, J.A. (2007). Computer software review. Cheating detection at your fingertips: A review of INTEGRITY. Applied Psychological Measurement, 31, 233–9. Wowra, S.A. (2007). Moral identities, social anxiety, and academic dishonesty among American college students. Ethics and Behavior, 17, 303–21. Wrensen, L.B., & Biderman, M.D. (2005). Factors related to faking ability: A structural equation model application. Paper presented at the 20th Annual Conference of the Society for Industrial and Organizational Psychology, Los Angeles, CA. Xiaofeng, G., & Lie, M. (2006). High-tech products used for cheating in examination. China Daily. http://www.chinadaily.com.cn/cndy/2006-06/20/content 620882.htm (accessed 30 June 2009). Xin, K.R. (2004). Asian American Managers: An impression gap?: An investigation of impression management and supervisor-subordinate relationships. Journal of Applied Behavioral Science, 40, 160–81. Yeung, L.N.T., Levine, T.R., & Nishiyama, K. (1999). Information manipulation theory and perceptions of deception in Hong Kong. Communication Reports, 12, 1–11. Zerbe, W.J., & Paulhus, D.L. (1987). Socially desirable responding in organizational behavior: A reconception. Academy of Management Review, 12, 250–64. Zickar, M.J. (2000). Modeling faking on personality tests. In D. Ilgen, & C.L. Hulin (Eds), Computational Modeling of Behavioral Processes in Organizations (pp. 95–108). Washington, DC: American Psychological Association. Zickar, M.J., & Gibby, R.E. (2006). A history of faking and socially desirable responding on personality tests. In R.L. Griffith & M.H. Peterson (Eds), A Closer Examination of Applicant Faking Behavior (pp. 21–42). Greenwich, CT: Information Age. Zickar, M.J., Gibby, R.E., & Robie, C. (2004). Uncovering faking samples in applicant, incumbent, and experimental data sets: An application of mixed-model item response theory. Organizational Research Methods, 7, 168–90. Zickar, M.J. & Robie, C. (1999). Modeling faking good on personality items: An itemlevel analysis. Journal of Applied Psychology, 84, 551–63. Zuckerman, M., DePaulo, B., & Rosenthal, R. (1981). Verbal and non-verbal communication of deception. In L. Berkowitz (Ed.), Advances in Experimental Psychology (volume 14), (pp. 1–59). New York: Academic Press.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
Chapter 8 ACTIONS SPEAK TOO: UNCOVERING POSSIBLE IMPLICIT AND EXPLICIT DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS Therese Macan and Stephanie Merritt University of Missouri–St. Louis, St. Louis, Missouri, USA The employment interview is a social exchange between applicants seeking employment and interviewers gathering information on which to make selection decisions. Both use this encounter to collect information, make judgments, and manage impressions. This interaction holds multiple opportunities for verbal and nonverbal behaviors to be displayed and communicated, some of which may not be perceived as being inclusive of all applicants. In this chapter, we examine issues that focus on possible discrimination in interviews. A reader may question: why study this area? After all, legislation exists in many countries that protects individuals from employment discrimination. In fact, Myors, Lievens, Schollarert, et al. (2008) found that all 22 countries they examined had some law or directive that outlawed discrimination for members of specified groups. Discrimination claims, lawsuits, and court decisions continue to send the message that formal, blatant discrimination will not be tolerated. However, discrimination can and does occur in various forms within the employment interview process. Discrimination in the interview may be manifested in formal, overt ways such as in differential evaluations of applicants’ interview performances and subsequent hiring decisions based on group membership, or it may be manifested in more subtle ways. The modern view of discrimination focuses on nonverbal, paralinguistic and, sometimes, verbal behaviors that occur in social interactions with stigmatized persons (Dovidio & Gaertner, 2000; Hebl, Foster, Mannix, et al., 2002). Stigmatized persons are those who are perceived to possess attributes that are seen as undesirable and distinct from others (Goffman, 1963). Thus, if interviewers react to and treat stigmatized individuals differently than non-stigmatized International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
294 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY persons during the interview, even unconsciously, the result is that interviewers obtain less accurate and fair applicant information – thus compromising the quality of their hiring decisions. While interpersonal discrimination may not always constitute illegal discrimination under some of the present legal language, legal scholars have taken note of these more subtle forms of discrimination and their potential to impact legal standards and precedent (e.g., Banks & Ford, 2009; Bartlett, 2009; Lee, 2005; Mitchell & Tetlock, 2006; Ware, 2007–2008). Consider three recent examples of legal cases specific to the interview: A male academic has won a sex discrimination case against the University of Surrey. . . Dr. Gilbert was rejected for the job after an interview with Professor Ogden and the associate dean for learning and teaching. . . The pair said that Dr. Gilbert had performed poorly at the interview and that Professor Ogden had found him ‘arrogant and annoying’. But the tribunal said Professor Ogden could provide no evidence of her assertion. It also drew an ‘adverse inference’ from the fact that the two professors did not take notes during the interview. (Newman, 10 December 2009) The Equal Employment Opportunity Commission of the USA alleges that Orkin’s agent asked Kokezas his age, then cut the interview short after learning Kokezas was 51. (EEOC Press Release, 20 May 2010) A woman who said she was turned down for a shop assistant’s job because she was expecting her second child has won her sex discrimination case. Ms. Tobin said that during an interview she told Mr. Majid she wanted a temporary post because she was pregnant. She claimed his demeanor and manner changed, he appeared annoyed, became very blunt and said ‘no’. The tribunal accepted that Mr. Majid told Ms. Tobin he was not prepared to employ her when she told him she was pregnant and that her pregnancy was the reason why he did not want to employ her. (Story from BBC News, 5 November 2007)
Given these incidences of interviewer behaviors, we concur with many organizational researchers that future research needs to integrate the subtle forms of discrimination with research on overt discrimination (Dipboye & Colella, 2005). Therefore, we focus this chapter on how both overt and interpersonal components of employment discrimination can occur in the interview. Researchers have investigated interviewers’ judgments of applicants of different demographics, with disabilities, or those who are pregnant, overweight, Lesbian Gay Bisexual Transgender (LGBT), or stigmatized in other ways. In this chapter we highlight recent research publications on these topics (as previous reviews have summarized earlier work, e.g., Arvey & Campion, 1982; Harris, 1989; Macan, 2009; Posthuma, Morgeson, & Campion, 2002; Schmitt, 1976). We also endeavor to expand thinking by connecting the implicit attitude research into the interview process. Within this conceptual framework, we present many ideas for future research (Table 8.1). We aim to share with the reader the following information: 1. What we know about these interviewer–interviewee dynamics for protected and stigmatized applicants in the interview; 2. What research findings suggest we can do to reduce biases; and
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS Table 8.1
Suggested areas for future research on implicit cognition and the employment interview
Pre-interview Phase R´esum´e screening Intergroup forecasting error
Stereotype threat
Interview Phase Behavioral processes
Perceptual processes
Post-interview Phase Ratings – shifting standards and status characteristics
Selection and constructed criteria All Phases Individual differences and motivation Accountability Context Training
295
Does removing identifying information from r´esum´es decrease adverse impact in interview invitations? What effect does the intergroup forecasting error have in interview settings? Can interventions making similarities salient decrease the intergroup forecasting error for interviewers? Can mental imagery reduce discrimination at the pre-interview and interview phases? How can organizations decrease interviewer stereotype threat and its effects? Could diversity framing and diversity climate efforts have the ironic effect of inducing interviewer stereotype threat? How can automatic avoidance behaviors and assimilation be prevented in order to reduce self-fulfilling prophesies in the interview? How do interviewer individual differences (e.g., prejudice level) affect the allocation of attention, memory, and attributions for stereotypical and counter-stereotypical individuating information? What effects do various note taking strategies and rating forms (e.g., subjective/objective) have on subgroup differences in interviews? To what extent do motivated interviewers attempt to mentally correct for potential biases in their ratings of stigmatized group members? Are their corrections accurate, or do they over- or under-correct? What are the effects of cognitive load? What are the effects of interviewer attempts to suppress stereotypes during the interview under low and high cognitive load? Are subsequent interviews affected by stereotype rebound? Do adverse impact effect sizes differ on subjective vs. objective rating forms? On ratings of minimum standards (e.g., cut-off selection model) versus competence assessment (e.g., compensatory model)? How do decision makers construct criteria to favor the preferred applicant(s)? How can constructed criteria bias be avoided (e.g., actuarial model, pre-commitment to information weights)? What interviewer individual differences and motivations are associated with greater degrees of bias? Can such individual differences and motivations be trained? What are the effects of interventions designed to increase interviewer accountability in decision making? How can organizational context (e.g., diversity climate, physical environment) be leveraged to minimize biases in interviews? How can training programs for interviewers be designed to minimize/change implicit cognition in order to reduce bias? How do individuals react to such change attempts?
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
296 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY 3. Where we believe researchers need to focus their efforts in order to create a more inclusive workforce. This chapter serves two main aims. First in the first section, we review recent research demonstrating that discrimination can and does occur in the job interview. Secondly, we provide an overview of research on implicit cognition which we believe may be relevant to discrimination in the interview context in the hope of providing directions for future research on diversity and inclusion in the job interview process. We believe this focus on the employment interview is imperative. As one of the most frequently used techniques for selecting applicants, it is unusual that an individual is hired without an interview. It appears employers want to meet individuals face-to-face or speak with them before extending an offer of hire. We suspect that when other selection techniques may be waived because of cost cutting, scheduling issues, or reasonable accommodation challenges, the employment interview remains as a selection tool. If this occurs, organizations may be weighting the employment interview more than other sources of information in reaching a final hiring decision. Furthermore, we argue that structured interviews may not necessarily be immune to interviewer biases. A structured interview can take many forms and there is much variability among researchers when using the term “structured” to describe the interview. Quantitative and qualitative reviews of employment interview research have concluded that adding structure to the interview process can enhance the reliability and validity of interviewer evaluations without the adverse impact typically found for cognitive ability tests (e.g., Conway, Jako, & Goodman, 1995; Huffcutt & Arthur, 1994; Huffcutt & Culbertson, 2010; Huffcutt & Woehr, 1999; Posthuma, Morgeson, & Campion, 2002). Several moderators of this relationship, however, have been noted (for a review see Macan, 2009). In addition, despite the evidence showing that structured interviews can be valid predictors of job performance, surveys show that managers, human resource (HR) professionals, and organizations infrequently use them (Huffcutt & Culbertson, 2010; Klehe, 2004; Lievens & De Paepe, 2004; Simola, Taggar, & Smith, 2007; van der Zee, Bakker, & Bakker, 2002). Researchers have struggled with how to encourage organizations to use structured interviews (e.g., Dipboye, 1994, 1997; Lievens & De Paepe, 2004). While it appears the trend is to use some of the structured interview components, maintaining the intended level of standardization of the selection interview can be challenging. When standardization is compromised, it is possible for biases to enter the interview process (McKay & Davis, 2008).
REVIEW OF EMPLOYMENT INTERVIEW RESEARCH It is critical to understand how interviewers’ judgments are made so we can ensure fair, accurate, and valid evaluations for all. As organizations become
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
297
global, we need to consider the interviewer–interviewee dynamics across cultures and countries and their effects on protected and stigmatized groups for which there may be prejudices, stereotypes, preconceived expectations, and even unconscious or implicit attitudes that creep into one’s judgments. In this section, we review some of the more recent employment interview research that has examined a number of applicant characteristics protected by laws in most countries (i.e., gender, race, ethnicity, age, disabilities) as well as those covered in only a few of the 22 countries Myors, Lievens, Schollaert, et al. (2008) examined (i.e., marital/family status, weight, sexual orientation, political opinion). Applicant Demographics Researchers have investigated the effects of a variety of applicant demographics and other characteristics on interviewer judgments for over 40 years. Details of these studies can be found in the major reviews of the interview literature (Arvey & Campion, 1982; Harris, 1989; Macan, 2009; Posthuma, Morgeson, & Campion, 2002; Schmitt, 1976; Wright, 1969). In general, these early studies examined the direct effects of gender, race, or ethnicity; found small and inconsistent effects; and concluded that these demographics are not major factors in interviewers’ decisions. However, this conclusion requires re-examination given more recent research. For example, Roth, Van Iddekinge, Huffcutt, et al. (2002) found “fairly large ethnic group differences” for the use of the interview as an initial screening. They found methodological problems with the effect size estimates used in prior studies as these estimates were calculated using restricted samples (e.g., job applicants who had passed a previous selection test or samples of job incumbents) resulting in standardized ethnic group differences not reflective or directly applicable to the entire referent population of job applicants. Most surprising is that these “differences were observed despite the fact that the interview was structured, with job-related questions based on job analyses, and trained interviewers” (Roth, Van Iddekinge, Huffcutt, et al. (2002: 375). Additional research also points to the importance of examining the underlying interview processes related to demographics. Contextual factors such as: (i) job status level (e.g., Singer & Sewell, 1989), (ii) job sex-typing (e.g., Raza & Carpenter, 1987) or age-typing (e.g., Perry & Bourhis, 1998), and (iii) the percentage of minorities in the applicant pool (e.g., Cleveland, Festa, & Montgomery, 1988) have been found to have a role in whether less favorable ratings are given for minority candidates. Interviewers’ expectations for applicants based on demographics also have affected interviewer ratings (Purkiss, Perrewe, Gillespie, et al., 2006). Demographic similarity between interviewer and applicant additionally has been considered. Both McCarthy, Van Iddekinge, and Campion (2010) and Sacco, Scheu, Ryan, et al. (2003) investigated the effects of actual demographic similarity between interviewer and applicant (i.e., gender and race)
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
298 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY on interviewer ratings and found no evidence of similarity rating effects. The interviewers’ perceived similarity with the applicant based on both demographics and attitudinal similarity attitudes, however, may influence the interactions during the interview (Posthuma, Morgeson, & Campion, 2002). Given the steady rise in the number of older employees in the workforce, the effects of applicant age on employment interview processes and outcomes have also been explored (for a review see Morgeson, Reider, Campion, et al., 2008). In general, age stereotypes have been found. The extent to which these stereotypes translate into age discrimination in the employment interview requires further investigation. In general, more evidence of age discrimination in the interview has been demonstrated in laboratory studies than field studies. Applicants with Disabilities Researchers have examined the effects of applicant disability on interviewers’ judgments across a wide variety of disabilities (e.g., blindness, hearing impairment, HIV-positive, leg amputee, mental illness, paraplegia, schizophrenia, recovering substance abusers, transverse myelitis). Overall, researchers have found that interviewers’ ratings and hiring recommendations for applicants with disabilities depend on the type of disability (for reviews see Arvey & Campion, 1982; Macan, 2009; Posthuma, Morgeson, & Campion, 2002). Applicant disclosure of their disability and discussion of the disability during the interview have received attention. The notion is that applicants who choose to disclose or acknowledge the disability may reduce interviewer uncertainty about the disability and any interviewer tension and discomfort in the interaction. This acknowledgment would serve to directly address the negative stereotypes and control the quality and quantity of information about the disability (Thompson, 1982). Previous research, however, has found mixed results on the ratings of applicants who disclose their disability. For example, Macan and Hayes (1995; Hayes & Macan, 1997) found applicant discussion of some types of disability-related information (e.g., encouraging the interviewer to ask questions about the disability, stating how they would perform the job) was favorably related to how well the interviewer evaluated the interview performance. However, Herold (2000) showed applicants who disclosed received lower ratings than those who did not disclose. Other research findings suggest possible moderators between applicant disability and interviewer ratings such as whether interviewers perceive applicants as responsible for the disability (Hebl & Kleck, 2002), the nature of the disability (e.g., physical or mental; Dalgin & Bellini, 2008), whether the disability is obvious or not, the level of applicant qualifications (for a review see Posthuma, Morgeson, & Campion, 2002), and when during the interview session (early, later) the disability is discussed (Hebl & Skorinko, 2005; Roberts & Macan, 2006). Research has also begun to examine the effects of imposing structure to the interview in the types of questions asked (e.g., job-related situational and
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
299
behavioral) and use of behaviorally anchored rating scales on interviewer evaluations of applicants with disabilities. Results are mixed, with one study reporting no significant difference in hiring ratings for applicants with and without disabilities in the structured interview condition (Brecher, Bragger, & Kutcher, 2006) and the other finding differences among the types of disabilities on recommendation-for-hire scores but also some support for using behavioral anchors on a total interview score (Reilly, Bocketti, Maser, et al., 2006). However, each study chose different types of applicant disabilities, making comparisons difficult. Researchers have examined a number of different disabilities. Despite the benefit to such breadth, one result of different choices of disabilities is that it is challenging to compare across studies and establish a theoretical rationale for differences found. Stone and Colella (1996) identified a number of elements that distinguish disabilities and can determine how a person with a disability is judged (e.g., physical, psychological, or sensory; concealable; disruptiveness; or origin (i.e., cause)). Researchers need to consider these particular dimensions and articulate them in their work. At times, previous research has confounded the visible–nonvisible aspect with general disability type (i.e., physical/visible disabilities have been compared with mental/nonvisible disabilities). A more systematic framework would advance our understanding of the underlying processes influencing discrimination towards applicants with disabilities in employment interviews and lead to more effective practical recommendations for interviewers and applicants. Pregnant and Working Mother Applicants In a survey of over 1000 pregnant women and new mothers by the UK’s Equal Opportunities Commission, it was found that almost half reported experiencing discrimination at work (Payne, 2006). The number of pregnancy discrimination claims in the USA continues to climb, with a 33% increase from 2000 to 2009 (EEOC, 2010). A few studies have begun to examine the effects of applicant pregnancy on interviewer decision making and have found less favorable interviewer hiring ratings when the female applicant was presented pregnant (wearing a pregnancy prosthesis) compared with when she provided the identical interview performance (using videos) but was shown not pregnant (Bragger, Kutcher, Morgan, et al., 2002; Cunningham & Macan, 2007). Interviewer concerns about the pregnant applicant needing time off, missing work, and quitting may provide some explanation for the differences (Cunningham & Macan, 2007). Important for the job interview interaction, Hebl, King, Glick, et al. (2007) found that women posing as applicants asking about job openings experienced more interpersonal discrimination (i.e., were treated in a more hostile manner) when ostensibly pregnant than when not pregnant. Also, working mothers have been perceived to be less competent, less committed to and involved in their work, and less flexible for advancement than
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
300 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY fathers (Correll, Benard, & Paik, 2007; Heilman & Okimoto, 2008; King, 2008). Thus, the majority of current evidence suggests that pregnant women and working mothers may face significant discrimination in hiring contexts. Future research needs to examine the influence of these perceptions within the employment interview process. The effect of applicant discussion of these concerns during the interview on ameliorating interviewers’ fears should be explored to ascertain the impact on interviewers’ evaluations and decisions. Overweight and Obese Applicants Roehling (1999) provided a comprehensive review of research examining weight-based discrimination in employment from a psychological and USbased legal perspective. Much of the prior work investigating hiring decisions asked participants to review information (e.g., applications, r´esum´es), with weight manipulated through pictures, videos, or written descriptions. Overwhelmingly, overweight or obese applicants in studies typically conducted in the USA, especially those who were female, were recommended for hire significantly less than normal weight applicants (see review by Roehling, 1999). This finding also held in samples from New Zealand (Ding & Stillman, 2005) and Britain (Swami, Chan, Wong, et al., 2008). In fact, a meta-analytic review of weight-related empirical studies showed an effect size of d = –0.70 for hiring outcomes (Rudolph, Wells, Weller, et al., 2009). We found only a few studies that specifically evaluated applicant weight in an employment interview setting, all of which showed evidence for weight-based biased judgments. In two of the studies, videotaped interviews were used in which normal weight applicants were made to look overweight through clothing, make-up, and/or prostheses. Both studies found that overweight applicants were less likely to be recommended for hire than the equally qualified normal weight applicant (Kutcher & Bragger, 2004; Pingitore, Dugoni, Tindale, et al., 1994). Weight has had a larger effect for women than men (Pingitore, Dugoni, Tindale, et al., 1994) and in general women have reported more experiences of weight-related discrimination than men (Roehling, Roehling, & Pichler, 2007). Evidence, however, suggests that men are stigmatized too (Hebl & Turchin, 2005). Furthermore, evaluations have differed based on the level of interview structure, with a behaviorally based evaluation resulting in no significant difference in ratings but a significant difference between the overweight and normal weight applicants in an unstructured interview (Kutcher & Bragger, 2004). Finkelstein, Demuth, and Sweeney (2007) found no evidence for type of job or applicant race as moderators of weight-based hiring ratings, although there is evidence that people may be more accepting of overweight African-Americans (e.g. Hebl & Heatherton, 1998; Hebl & Turchin, 2005). In addition, no differences in hiring ratings were found by weight for highly qualified applicants, while marginal effects on weight resulted for the moderately qualified applicant.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
301
Acknowledging the stigma in the interview, a strategy that has been shown to sometimes be beneficial for applicants with disabilities, was not necessarily found to be advantageous for obese individuals. In fact, a better strategy for obese individuals was to say nothing about the condition, unless the information revealed that the applicant was not directly responsible for the disability (e.g., “has a thyroid condition”; Hebl & Kleck, 2002). Given the social interaction aspect of the employment interview, it is important for future research to help us understand what transpires during job interviews for all applicants including those who may be overweight and obese. Previous interview studies have typically used videotaped scripted interviews resulting in uncertainty regarding whether interviewers might react to and treat overweight/obese applicants differently in the interview process. If so, this differential treatment may lead candidates themselves to act differently, in ways that may negatively impact how they are evaluated. Overweight job applicants have been perceived to differ from normal weight applicants on personality traits, although current findings provide no evidence to substantiate these commonly held stereotypes (Roehling, Roehling, & Odland, 2008). Physical attractiveness is another route by which applicant weight may bias interview decisions. Overweight people have been perceived as less attractive, suggesting that some of the bias against people who are overweight could be attributed to biases generally found for less attractive individuals (Rothblum, Miller, & Garbutt, 1988; see also Roehling, 1999, for a model of this and other processes that may lead to weight-based employment discrimination). Lesbian Gay Bisexual Transgender Applicants King and Cortina (2010) noted that LGBT individuals are another group that has been shown to experience discrimination at work. Applying for jobs in the Greek private sector, gay men had a significantly lower chance of obtaining a job interview than heterosexuals for the same job (Drydakis, 2009). While we were not able to locate any studies that examined interviewers’ evaluations of LGBT applicants in the employment interview, relevant studies of the job application context point to possible differences in how the interviewer might interact with these applicants. Hebl, Foster, Mannix, et al. (2002) had confederates acting as job applicants enter stores in a mall to enquire about applying for a job. The stigmatized applicants wore a hat labeled, “Gay and Proud,” while the other group wore a “Texan and Proud” hat. No evidence of formal discrimination was found (i.e., the proportion of stigmatized individuals who were told of job openings and given an application was not significantly different from non-stigmatized individuals). However, the stigmatized group experienced more interpersonal discrimination, as store employees tended to shorten the interaction, avoid eye contact, and express more overall negativity compared to the nonstigmatized group. Using this same methodology, Singletary and Hebl (2009)
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
302 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY replicated the findings and demonstrated some evidence that compensatory strategies (i.e., increased positivity, acknowledgment, individuating information) reduced interpersonal discrimination for stigmatized persons. In contrast to visible stigmas, one’s sexual orientation can be concealed. Not only would it be important for future research to examine the social interaction elements of the employment interview process for LGBT job candidates who disclose, but also any effects if someone is perceived to be LGBT when in fact they are not. Research Directions Examining Discrimination Issues Given the protected status across many countries of a wide range of individuals (Myors, Lievens, Schollaert, et al., 2008), it is important to recognize that researchers are examining a wider variety of groups beyond age, race, and gender. In general, the evidence suggests that discrimination in the employment interview can occur. Further examinations of multiply stigmatized applicants (e.g., gender and obesity as in Pingitore, Dugoni, Tindale, et al., 1994) would be informative to discern whether these additional stigmas have an additive or multiple jeopardy effect (e.g., Landrine, Klonoff, Alcaraz, et al., 1995) or whether one stigma may dominate interviewer evaluations (such as proposed in the ethnic prominence hypothesis; Levin, Sinclair, Veniegas, et al., 2002). It is also important for researchers to examine the effects of interviewer perceived (rather than actual) similarity to applicants (e.g., Garcia, Posthuma, & Colella, 2008). With some exceptions, much of the work on the effects of various applicant demographics and other characteristics has been conducted with videotaped interviews and in less structured interview situations. Clearly, this work is important, but future research should examine these effects in actual faceto-face interviews with greater degrees of structure (McKay & Davis, 2008) and with a focus on what actually transpires during the interview session. That is, employment interviews could present an opportunity for subtle cues to affect interviewer perceptions, behaviors, and judgments as well as applicants’ perceptions and interview performance, even in interviews that include components of structure as we describe next.
IMPLICIT BIASES Although explicit prejudice and stereotypes have been the focus of much research in our field, we believe that subtle, implicit biases deserve increased attention. While Landy (2008) argued that there is currently no evidence to suggest that implicit (or explicit) biases have any large effect in the field, he further argued that implicit biases are likely to have their largest effects in stranger-to-stranger interactions. The job interview is, in essence, a strangerto-stranger interaction in which implicit attitudes and stereotypes are likely to
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
303
have effects on interviewer behavior, interviewer information processing, and interviewee performance. To this end, we present an overview of some of this literature, as applicable to the interview process. We begin by briefly discussing the conceptual meaning of implicit cognition. Next, we present an interview process model which we adopt as a framework for discussing effects on the job interview. Finally, we generate some propositions regarding the potential process role that implicit cognition might have in each stage of the job interview and offer suggestions for future research in these areas. Implicit/Automatic Attitudes and Cognition For many years, researchers in social psychology and social cognition have explored the effects of implicit, or automatic, intergroup stereotypes and attitudes. Implicit constructs are rooted in connectionist theory, where concepts are associated in the mind based on learned associations, and in dual process models which propose that human judgments and behavior can be produced by either automatic or controlled processes. While definitions of implicit attitudes vary, there is a growing argument that these “implicit” constructs may not necessarily be unconscious, as some individuals may have awareness of these mental associations (cf. Bargh, 1994; Gawronski, Hofmann, & Wilbur, 2006; Gawronski, LeBel, & Peters, 2007). For this reason, the term “implicit” attitudes may be somewhat of a misnomer, with “automatic” being perhaps a more appropriate term. However, due to the tradition and popularity of the term “implicit,” we use the labels “implicit” and “automatic” interchangeably, with the recognition that these constructs may not always be unconscious. Typically, explicit constructs are assessed via traditional self-report measures such as Likert-type attitudinal scales (e.g., the Modern Racism Scale; McConahay, 1986). Several methods for assessing implicit constructs exist, including response latency measures such as the Implicit Association Test1 (IAT; Greenwald, McGhee, & Schwartz, 1998), Go/No-Go Association Task (GNAT; Nosek & Banaji, 2001), Evaluative Priming Task (Fazio, Jackson, Dunton, et al., 1995; Fazio, Sanbonmatsu, Powell, et al., 1986), Extrinsic Affective Simon Task (EAST; De Houwer, 2003), and semantic prim˜ ing paradigms (for reviews see Petty, Fazio, & Brinol, 2008; Wittenbrink & Schwartz, 2007); and nonresponse latency measures such as the Affect Misattribution Procedure (Payne, Cheng, Govorun, et al., 2005) and the Stereotype Explanatory Bias (Sekaquaptewa, Espinoza, Thompson, et al., 2003).
1
The validity of the IAT has been the subject of considerable debate (c.f., Nosek, Greenwald, & Banaji, 2007). However, our review does not focus solely on research using the IAT. We incorporate findings from studies using several implicit measures.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
304 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY The theoretical and empirical relationships between implicit and explicit attitudes are the subject of debate. Early researchers endorsed the notion that when implicit attitude measures showed racial bias but explicit (i.e., self-report) measures did not, the discrepancy was due to social desirability effects on explicit measures and that the implicit measures reflected respondents’ “true” attitudes (e.g., “bona fide pipeline”; Fazio, Jackson, Dunton, et al., 1995). However, as research has progressed, a perspective based in dual process models has gained popularity. This perspective states that implicit and explicit attitudes may be two separate (but related) constructs, each with differential validity for predicting different types of outcomes. For example, the Associative-Propositional Evaluation Model (APE; Gawronski & Bodenhausen, 2006) suggests that implicit attitudes operate purely via associative processes and are independent of the individual’s values or perceived truth. Thus, implicit attitudes may affect the individual’s behavior regardless of whether or not the individual believes the attitude to be “true” or “correct.” In contrast, the APE model suggests that explicit attitudes operate via propositional processes in which the attitude is subjected to truth values (“true” or “false”) and will be endorsed only when the individual believes the attitude to be true. Empirical Evidence Empirically, there seems to be some support for each perspective. Supporting the “social desirability” hypothesis, research has found that implicit and explicit attitudes correlate more strongly in non-socially sensitive domains than in sensitive domains such as racial attitudes (cf. Greenwald, Poehlman, Uhlmann, et al., 2009; Nosek, 2005), while other research has reached different conclusions (Hofmann, Gawronski, Gschwendner, et al., 2005). Greenwald, Poehlman, Uhlmann, et al.’s finding suggests that social desirability may have a role in the divergence between an individual’s implicit and explicit attitudes. However, supporting the second perspective, it has typically been found that implicit attitudes are better predictors of spontaneous and difficult-tocontrol behaviors, such as nonverbal behaviors and biased information processing, whereas explicit attitudes are better predictors of deliberate and ver¨ bal behaviors (Asendorpf, Banse, & Mucke, 2002; Dovidio, Kawakami, & Gaertner, 2002; Dovidio, Kawakami, Johnson, et al., 1997; Hofmann, Rauch, & Gawronski, 2007; Hugenberg & Bodenhausen, 2003, 2004; Neumann, ¨ Hulsenbeck, & Seibt, 2004). Also, some studies have found that specific manipulations produced changes in explicit, but not implicit, attitudes (e.g., Gawronski & Strack, 2004), whereas other manipulations have produced changes in implicit, but not explicit, attitudes (e.g., Dasgupta & Greenwald, 2001; Karpinski & Hilton, 2001; Olson & Fazio, 2006), suggesting that these types of attitudes are separate constructs. Consistent with dual process models, implicit attitudes are thought to control behavior when the individual is under high cognitive load, whereas when
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
305
under low load conditions, explicit attitudes can control behavior if the individual possesses sufficient motivation to override the automatic effects (e.g., Fazio & Towles-Schwen, 1999). These results have been bolstered by a neuroimaging study finding that different brain areas are associated with processing of ingroup and outgroup faces under conditions associated with automatic and controlled processing (Cunningham, Johnson, Raye, et al., 2004). The implication of the notion that implicit and explicit processes are related but distinct is that by turning a lens on the role of implicit cognition we may gain information about interview processes that cannot be obtained via traditional explicit measures. Researchers in industrial/organizational (I/O) psychology have begun to take note of the implicit cognition research. A study examining the role of implicit prejudice in simulated hiring decisions (Ziegert & Hanges, 2005), a discussion of potential uses of implicit measures in organizational research (Haines & Sumner, 2006), and a spirited debate on the applicability of implicit cognition research to I/O (Landy, 2008) have appeared in top I/O journals. We believe that much of the implicit cognition research does have potential to be applied to workplace settings, particularly in the domain of job interviews. In fact, these implicit processes may affect various stages of the interview process.
INTERVIEW PROCESS MODEL Dipboye and Macan’s (1988) interview process model provides an organizing framework in which to examine the various paths through which potential biases may enter the interview process, even for interviews that are structured. According to the model, there are three phases to the interview process: preinterview, interview, and post-interview (Figure 8.1). The relationships among both behavioral and cognitive events that occur between the interviewer and interviewee are delineated within each phase of the model. With regard to the pre-interview phase, Dipboye and Macan posit that interviewers hold notions of the ideal applicant for the job based on a variety of information including job-relevant knowledge and skills as well as stereotypical attributes that interviewers believe are important for success on the job. Even before interviewers meet the applicants, they typically have access to information about them and tend to make judgments of the applicants’ fit to the ideal candidate based on materials available (e.g., r´esum´es, applications, test scores, letters of recommendation, social networking profiles). That is, interviewers form pre-interview impressions of applicants, which may shape their expectations for interviewee performance and affect subsequent information processing and behavior. The interview phase encompasses the actual social interaction that takes place, frequently conducted face-to-face but also possibly remotely with videoconference technology (e.g., Chapman, Uggerslev, & Webster, 2003; Straus,
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
306 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Pre-interview stage Résumé screening Pre-interview expectations Intergroup forecasting error stereotype threat
Interview stage Behavioral processes Avoidance/distancing Assimilation Self-fulfilling prophesies Perceptual processes Interpretation of ambiguous info. Attention Memory Attributions
Post-interview stage Shifting standards Status characteristics Constructed criteria Figure 8.1
Interview process model and implicit processes entering at each stage.
Miles, & Levesque, 2001). Dipboye and Macan (1988) propose that interviewers’ pre-interview impressions can influence both their cognitive processing of information and their behaviors during this exchange. One example is that the questions interviewers choose to ask may be related to how qualified they think the applicant is for the job based on their pre-interview impression (Macan & Dipboye, 1988). If a structured interview is used in which job-related questions are provided to the interviewer, questions may be standardized across applicants. However, even in this case, interviewer paralinguistic and nonverbal behaviors may affect the interaction. For example, how interviewers ask the question (i.e., their intonation, expressed interest); which follow-up questions are used and to what extent; and what their nonverbals communicate to the applicants may differ across applicants. In turn, applicants may be able to
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
307
detect interviewers’ attitudes regarding their qualifications from these behaviors, and this could affect applicants’ interview performances. Applicants who believe the interviewer sees them in a favorable light may feel encouraged, less anxious, and better able to answer the interview questions. If not viewed favorably, applicants may experience increased anxiety (McCarthy & Goffin, 2004; Sieverding, 2009) and find it extremely challenging to provide answers that overcome this negativity and leave the interviewer with a positive impression. Dipboye and Macan (1988) suggest pre-interview impressions can affect the interviewers’ level of attention and what information they recognize about the applicant, even in unconscious ways. Interviewers’ pre-interview impressions can also influence what information they recall from the interview and also how that recalled information is interpreted and causally attributed (Macan & Dipboye, 1994). In the post-interview stage, interviewers make decisions about the applicants (e.g., hire/not hire, invite for another interview, site visit, further testing). Interviewers are expected to integrate the information they gathered during these first two phases to form these decisions. Interviewers tend to form a general impression of how well the applicant performed in the interview based on their verbal and nonverbal behaviors. While this overall evaluation may not be formally rated (depending on whether it is a dimension on a structured rating form), it can still influence interviewers’ ratings (similar to how contextual performance has been found to affect supervisors’ overall performance ratings; e.g., Motowidlo & Van Scotter, 1994). In the remainder of the chapter, we present research we believe to be highly relevant to how implicit attitudes can enter at each of these stages. Although much of this work has focused on racial issues, we believe it could apply to all protected classes, as well as all stigmatized individuals. Therefore, our hope is to inspire research on applied implicit cognition, including studies of effects in field settings and intervention studies designed to reduce workplace prejudice and discrimination across all subgroups of individuals within the interview process. Our overview of this research is necessarily broad rather than deep. However, by introducing some of the seminal and/or recent research in each topic area below, our goal is to spark interest and creativity in extending research on these topics to the interview setting.
Pre-Interview Phase R´esum´e screening Bias may enter the selection system at the pre-interview phase. One type of study focused on this issue is the “r´esum´e audit,” in which r´esum´es are sent to employers in response to actual job ads (e.g., Banerjee, Bertrand, Datta, et al., 2009; Bendick, Jackson, & Romero, 1997; Galgano, 2009; Neumark, Bank, & Van Nort, 1996). The quality and content of these r´esum´es are typically
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
308 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY controlled, with only group membership varying, and the dependent variable is rate of callbacks for interviews for the members of different subgroups. In one of the most widely cited of these studies, Bertrand and Mullainathan (2004) sent equivalently qualified r´esum´es to job ads posted in Boston and Chicago and, on these r´esum´es, manipulated the names to sound either European-American (e.g., Jill, Brett) or African-American (e.g., Ebony, Tyrone). They found that r´esum´es with White-sounding names received 50% more interview callbacks than r´esum´es with Black-sounding names and calculated that having a White-sounding name was as beneficial as having an additional 8 years of work experience. Furthermore, they provided evidence that this difference was not attenuated by higher quality r´esum´es (in fact, the difference in callback rate was even larger with higher quality r´esum´es) and that the difference did not seem to be caused by perceived socio-economic status differences or by differential familiarity with the names. Laboratory studies have also found bias based on the implied ethnicity of applicants’ names on r´esum´es, particularly depending on job status (King, Madera, Hebl, et al., 2006). Given the evidence, one might conclude that bias affected the callback rates for these fictitious applicants; however, whether that bias operated at an implicit or explicit level was not assessed. A follow-up study by Rooth (2007) assessed implicit and explicit bias in the r´esum´e screening process. Using a procedure similar to that of Bertrand and Mullainathan, Rooth found that in response to Swedish job ads, r´esum´es with Swedish-sounding names received 50% more callbacks than r´esum´es with Arab/Muslim-sounding names. Following the callback period, Rooth contacted the hiring managers who had actually made the callback decisions and asked them to take a Swedish–Muslim IAT and measures of explicit preference for hiring Swedish and Arab/Muslim workers. Although the hiring managers were willing to report explicit preferences for Swedish applicants, these preferences were not significantly associated with callback rates. However, the IAT scores did significantly predict callback rates for r´esum´es with Arab/Muslim sounding names, suggesting that implicit (but not explicit) stereotypes and attitudes may drive discrimination at the r´esum´e screening stage. Given these findings, and the fact that implicit attitudes are difficult to control consciously, the ideal solution for organizations may be to prevent implicit processes from operating at the r´esum´e screening stage by removing identifying information from the r´esum´e prior to screening. For example, an automated r´esum´e screening system might be used, or an administrative assistant might remove names, e-mails, and addresses from the r´esum´es, replacing them with an identifying code. The hiring manager could then screen the r´esum´es and select r´esum´es for callbacks free of contamination from names and other information that might identify members of protected groups, and applicants could be contacted by individuals other than those who will be conducting the interviews. Removing group-identifying information from r´esum´es should prevent
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
309
implicit stereotypes and prejudices from being activated and from influencing the r´esum´e screening process. A second potential remedy may be to increase accountability for r´esum´e screeners. Ford, Gambino, Lee, et al. (2005) found that telling participants that they would need to justify their r´esum´e screening decisions eliminated racial bias against applicants with Black-sounding names. When no such instructions were provided, racial bias was found. However, for this technique to be effective, organizations must consistently enforce the policy that justification be provided and ensure that the justification provided is free of subtle groupbased biases. Furthermore, when group-identifying information is salient to screeners, such as it would be with this strategy, the interviewer’s preparations for interviewing the candidate may be affected. Preparing to interview Even when applicants survive the screening process, interviewers may have differential expectations for these applicants when preparing for the interview. Research suggests that preparations to interact with outgroup members are subject to the intergroup forecasting error, in which people have unrealistically negative expectations for intergroup interactions (Mallett, Wilson, & Gilbert, 2008), and people tend to underestimate outgroup members’ interest in interacting with them (Shelton & Richeson, 2005). Thus, interviewers might have negative expectations for the quality of an upcoming interview when the interviewee is known to be an outgroup member. This “intergroup forecasting error” seems to stem from a “default” tendency to focus on the ways that outgroup members are different from oneself. In contrast, the “default” tendency when preparing to interact with ingroup members is to focus on similarities (Mallett, Wilson, & Gilbert, 2008). It was found that by inducing participants to focus on the ways that they were similar to outgroup members, participants’ expectations for positive intergroup interactions matched their expectations for positive within-group interactions (Mallett, Wilson, & Gilbert, 2008: studies 2 and 3). This suggests that one way to encourage positive expectations for outgroup interviewees might be to encourage interviewers to focus on expectations for similarities. Another potential technique for reducing the intergroup forecasting error may be the use of mental imagery. Research has found that simply imagining counter-stereotypic individuals (i.e., a strong woman) can reduce stereotyping as measured by several techniques, including the IAT, GNAT, and false memory measures (Blair, Ma, & Lenton, 2001). A program of research examining the effects of imagined positive intergroup contact found that mental simulation of positive intergroup contact resulted in decreased subsequent stereotyping and prejudice, increased projection of positive traits to the outgroup, increased perceived outgroup heterogeneity, and decreased intergroup anxiety (Crisp & Turner, 2009). Those authors propose that imagined intergroup
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
310 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY interaction can substitute for actual intergroup interaction when face-to-face contact with outgroup members is not available (e.g., in highly segregated environments). The benefits of mental imagery on preventing discrimination in the job interview may be a viable avenue for future research on diversity interventions. Some research has suggested that stereotype threat processes may also affect majority group interviewers preparing to interact with an outgroup candidate, particularly in situations where racial identity is salient – for example, in organizations with a strong diversity focus in selection. Evidence suggests that White Americans and Canadians are aware of the possibility that they are stereotyped as racially prejudiced (e.g., Vorauer, Hunter, Main, et al., 2000; Vorauer, Main, & O’Connell, 1998), and they show stereotype threat effects in this domain (e.g., Frantz, Cuddy, Burnett, et al., 2004). Goff, Steele, and Davies (2008) told White participants that they would be required to discuss either a racially charged topic (racial profiling) or a nonracially charged topic (love and relationships) with a Black partner and assessed their stereotype threat. They found that those who believed they were going to discuss a racially charged topic scored significantly higher on stereotype threat and also generated significantly more stereotype-relevant thoughts (e.g., “My first thought when I saw ‘racial profiling’ as a topic, and my partner was of a different ethnicity was that I might want to be cognizant of this and be somewhat careful in my remarks”). Furthermore, individuals in this condition significantly physically distanced themselves from their interaction partners. Thus, organizations seeking to communicate the value of diversity to hiring managers representing the majority group may inadvertently frame the interview scenario as racially charged – ironically thereby triggering stereotype threat and physical distancing behavior in interviews. Thus, framing of diversity-related values should be considered carefully, and research focusing on how best to do this is needed. Interview Phase Upon the initial encounter of another person, automatic preferences and stereotypes may activate immediately, effortlessly, uncontrollably, and oftentimes unconsciously. Even if pre-interview expectations have been controlled by the removal of group-identifying information, automatic biases can affect interviewer behavior at this point in time in several ways, including behaviorally and cognitively. Behavioral processes Encountering a member of a negatively stereotyped outgroup tends to automatically prime avoidance behavior. Paladino and Castelli (2008) demonstrated that participants were faster to perform approach behaviors when presented with ingroup members and faster to perform avoidance behaviors when
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
311
presented with outgroup members. This effect may lead to distancing behavior during the interview. A seminal study by Word, Zanna, and Cooper (1974) found that in mock job interviews, White job “interviewers” placed their chairs significantly closer to White “interviewees” than to Black “interviewees,” by an average distance of almost 4 inches. In addition, although the interviewers were given the same list of 15 questions to ask each interviewee, and although the interviewees were confederates trained to give equivalent responses to these questions, the interviewers spent 25% less time interviewing Black (9.42 minutes) than White interviewees (12.77 minutes). Thus, it is possible that interviewers may demonstrate avoidance behaviors such as greater interpersonal distance and terminating the interview sooner when interviewing outgroup members than ingroup members – which, in essence, would create differential treatment. In their second experiment, Word, Zanna, and Cooper trained confederate job interviewers to demonstrate either the immediate or non-immediate behaviors identified in their first study (i.e., in the non-immediate condition, confederates sat farther away than in the immediate condition). Independent coders analyzed videotapes showing only the interviewees to determine whether the interviewer’s behavior elicited a self-fulfilling prophecy. Consistent with the self-fulfilling prophecy, participants in the nonimmediate condition placed their chairs significantly farther away from the interviewer than participants in the immediate condition. Importantly, participants in the nonimmediate condition were rated by the independent coders to be significantly less composed and less adequate for the job than participants in the immediate condition. This research highlights the potential importance of standardization of interviewer behavior. Although Word, Zanna, and Cooper (1974) did not specifically measure primed avoidance behavior, the results of their study are consistent with the notion that primed avoidance behavior can affect interviewer behavior and, in turn, affect interviewee performance. Due to the activation of stereotypes associated with the candidate’s group, the interviewer becomes primed to act in accordance with the stereotype (a phenomenon called “assimilation”).2 Bargh, Chen, and Burrows (1996) primed half of their participants with stereotypes related to the category “elderly,” then a hidden experimenter timed participants as they left the lab and walked to the elevator. Participants who were primed with the elderly stereotype walked significantly slower than participants who were not primed, demonstrating behavioral assimilation of the stereotype. In a follow-up study, participants were primed with either the concept of “rude” or “polite.” After the priming procedure, they were told to find the experimenter to receive further instructions. When they did so, they found the experimenter deep in conversation with a confederate who feigned confusion about the experimental instructions. Participants primed with the “rude” concept interrupted the 2
Although see Wheeler and Petty (2001) for a review of when contrast effects may instead occur.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
312 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY conversation significantly faster than participants primed with “polite.” This classic experiment and others demonstrate the effect of primed categories and concepts on perceiver behavior. The implication is that an individual who is interviewing an outgroup candidate may behave like the stereotype held of the outgroup, which may then create self-fulfilling prophecies and “confirm” the stereotype. Automatic stereotype assimilation and automatic avoidance behavior have been demonstrated to create self-fulfilling prophecies in stranger-to-stranger interactions. Chen and Bargh (1997) found that independent coders in the USA blind to experimental condition rated participants who were subliminally primed with African-American faces, and their nonprimed interaction partners, as significantly more hostile than participants who were primed with Caucasian faces and their interaction partners. Importantly, the primed participants themselves also rated their interaction partners as significantly more hostile in the African-American prime condition than in the Caucasian prime condition. Thus, the automatic activation of demographic-based stereotypes may adversely affect the interview performance of minority candidates. These self-fulfilling prophecies seem to occur outside of the interviewer’s awareness. This lack of awareness was demonstrated by Dovidio, Kawakami, and Gaertner (2002) using mock college student interviews in the USA between White interviewers and both White and Black confederate interviewees (who were blind to the experimental hypotheses). The dependent variables of interest were ratings of verbal and nonverbal friendliness toward the Black interviewees, relative to each participant’s baseline level of friendliness toward the White interviewees. The authors demonstrated that while Whites’ explicit racial attitudes predicted their verbal friendliness toward the Black interviewees, their implicit racial attitudes predicted their nonverbal friendliness as rated by independent coders. Importantly, a disconnect was shown between the perceptions of the interviewers and interviewees. While the interviewers believed that their friendliness was associated with their explicit prejudice scores and verbal behavior, the interviewees’ perceptions of interviewer friendliness were significantly associated only with their implicit prejudice and nonverbal behavior. Thus, their study provides an important demonstration of the effects of interviewer implicit attitudes on interviewee perceptions, and how the perceptions of an interviewer and interviewee might differ. Perceptual processes Initial stereotypes and expectations may also affect the interviewer’s readiness to perceive stereotypical information, such that the interviewer is primed to interpret ambiguous information as stereotype-consistent. For example, researchers assessed White participants’ implicit and explicit racial attitudes and then presented them with computerized White and Black male faces which
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
313
transitioned along a continuum from smiling to hostile, passing through ambiguously hostile facial expressions in the middle of the continuum. They found that participants’ implicit (but not explicit) racial attitudes predicted the onset and offset of perceived hostility, such that participants with a stronger implicit preference for Whites were quicker to perceive the Black ambiguous faces as hostile, relative to the White faces (Hugenberg & Bodenhausen, 2003). Similar perceptual processes have been empirically demonstrated in other contexts relevant to the interview, including perceiving anger in Arab male faces (Maner, Kenrick, Becker, et al., 2005), anger in male versus female faces (Becker, Kenrick, Neuberg, et al., 2007), and in readiness to perceive an ambiguous target as good or bad (Stapel & Koomen, 2000). Interviewers with more negative implicit attitudes or stronger implicit stereotypes of a group may be biased to perceive stereotypical information in interviews with applicants who are members of the stereotyped groups. In addition, if this process is driven by implicit, and not explicit, attitudes (e.g., as found by Hugenberg & Bodenhausen, 2003), then interventions for preventing this bias will need to utilize levers for adjusting the interviewer’s implicit attitudes. Compounding the issue of readiness to perceive stereotypical information is that perceivers also typically require few pieces of perceived stereotypeconsistent information to make stereotype-consistent trait inferences about an individual. Conversely, they require a large number of counter-stereotypic pieces of information in order to make counter-stereotypic trait inferences (Biernat & Ma, 2005). As the job interview takes place, interviewers are tasked with finding out “what kind of person” the applicant is, with the goal of determining what type of employee the applicant would be. Thus, drawing trait inferences about the applicant may be an important component of the interview process. If interviewers are both ready to perceive ambiguous information in stereotypic ways, and if few pieces of evidence are required to “confirm” the interviewers’ stereotypical expectations of the applicant, then automatic prejudice may produce expectations that are difficult for stereotyped applicants to overcome. Automatic activation of stereotypes and prejudice also directs attention. Landy (2008) argued that when counter-stereotypic individuating information is provided in the duration of an interview, the effects of stereotypes will be attenuated. While sometimes correct, research has shown that stereotypes affect how perceivers attend to, remember, and interpret counter-stereotypic information. The ways in which this happens appear complex and could be a fascinating avenue for future research on the job interview process. Evidence has suggested that when individuals (particularly those high in prejudice) are faced with counter-stereotypic individuating information about an outgroup member, they may either decrease their attention to that information to avoid disconfirming their expectations, or they may increase their attention to the counter-stereotypic information in order to attempt to discredit or discount it. Thus, interviewers faced with counter-stereotypic individuating
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
314 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY information may not adjust their impressions of the applicant in alignment with this information. Research related to these processes is discussed below. In impression formation tasks, people tend to allocate attention to expectancy-inconsistent information when they are motivated to form accurate impressions and to expectancy-consistent information when they are not (Fiske, Lin, & Neuberg, 1999). This finding suggests that when interviewers are not motivated to form accurate impressions, they will allocate more attention to stereotype-consistent information than to stereotype-inconsistent information and, thus, the stereotypes will be resistant to change. Furthermore, individuals higher in prejudice often allocate increased attention to stereotype-inconsistent information – but they do so in order to scrutinize it and discount it, thereby preserving their stereotypes. Sherman, Stroessner, Conrey, et al. (2005) found such effects for both implicit (Experiment 3) and explicit (Experiments 1 and 2) prejudice and for social groups including sexual orientation and race. Similar to research presented previously which suggests that people require fewer pieces of stereotype-consistent information to make stereotypical trait inferences, this research suggests that people high in prejudice have a higher threshold for accepting counter-stereotypic than stereotypic information and scrutinize such information carefully. They also found that rather than discounting stereotype-consistent information, low prejudice perceivers incorporated all information into their impressions and formed more accurate perceptions of the target. Thus, organizations may wish to find strategies for reducing prejudice among interviewers or consider taking steps to select interviewers with lower prejudice. However, other evidence suggests that low prejudice persons show information processing biases toward discounting stereotype-consistent information (Wyer, 2004). Research investigating these relationships and extending the findings to applied settings is necessary. Some research has found effects on memory for preferred and non-preferred candidates’ qualifications. In a mock college student selection scenario, raters evidenced memory distortion such that the differences in the preferred and non-preferred candidates’ qualifications were exaggerated when racial information was provided (Norton, Vandello, & Darley, 2004). Thus, attempts to ask interviewers to justify their decisions may fall victim to information processing biases. Individuals may also make differential attributions for stereotypeinconsistent behavior. Sekaquaptewa and colleagues examined individuals’ tendency to try to explain stereotype-inconsistent, more so than stereotype consistent, behavior – a phenomenon termed the Stereotype Explanatory Bias (SEB). They found that individuals try to explain counter-stereotypic behavior for members of low status groups more than members of high status groups (Sekaquaptewa & Espinoza, 2004). This tendency is strong enough that it has been successfully used as a measure of racial bias. Sekaquaptewa, Espinoza, Thompson, et al. (2003: Study 1) found that in mock interviews with
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
315
African-Americans, the interviewers’ SEB scores significantly predicted the extent to which they chose stereotypic question forms (e.g., “Some people think they can get away with stealing food, silverware, even cash. Have you ever had any trouble like this?”) as opposed to non-stereotypic forms (e.g., “Some people think they can get away with taking work supplies home. Have you ever experienced it, and what did you do about it?”). This study underscores the need for strict interview structure on the questions posed in the interview in order to prevent bias, as well as highlighting the effects of SEB on behavior. Furthermore, SEB has been found to have differential effects depending on whether the explanations provided for counter-stereotypic behavior made internal or external attributions for the behavior (Sekaquaptewa, Espinoza, Thompson, et al. 2003: Study 2). Participants who made a greater proportion of external attributions for stereotype-inconsistent behaviors (“Shaniqua scored high on the SAT . . . because she took preparation courses”) relative to internal attributions (“Shaniqua scored high on the SAT . . . because she is smart”) received lower social interaction scores from Black, but not White, interaction partners. These results suggest that the attributions an interviewer makes for stereotype-inconsistent individuating information may have an important role not only in information processing and discounting, but in the actual interaction. The above findings apply to situations in which interviewers are under low cognitive load (Sherman, Stroessner, Conrey, et al., 2005). However, in many cases, interviewers tasked with asking questions, listening to answers, forming impressions, and taking notes may experience significant cognitive load. While Sherman, Stroessner, Conrey, et al.’s high prejudice participants were not able to devote increased attention to stereotype-inconsistent information when under higher load, other studies have found that stereotypes can instead create memory bias under high cognitive load conditions. Several studies have found that stereotype-inconsistent individuating information is recalled less well than consistent information under high load (e.g., Bodenhausen & Lichtenstein, 1987; Macrae, Hewstone, & Griffiths, 1993; Stangor & Duan, 1991; Stangor & McMillan, 1992), suggesting that interviewers under high cognitive load simply may not remember individuating information when making ratings following the interview. This process does, however, differ according to whether “remembering” is defined as free recall or recognition. When under cognitive load and when “remembering” means free recall, memory is indeed stronger for stereotype-consistent information (Sherman & Frost, 2000). In other words, if interviewers under cognitive load are simply asked to recall the interviewees’ answers and behaviors, they may be more likely to remember stereotype-consistent information than stereotype-inconsistent information, thereby preserving stereotypes and possibly creating adverse impact in interview ratings. However, when “remembering” is defined as recognition of information that either was or was not present, memory favors stereotypeinconsistent individuating information (Sherman & Frost, 2000). Thus,
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
316 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY organizational researchers may wish to investigate various note-taking strategies, or memory aids using technology (such as recording the interview for later playback), and their effects on group differences in interview ratings. Standardizing the evaluation component of the interview by using behaviorally anchored rating scales or behavioral checklists may also prove beneficial.
Post-Interview Decision-Making Phase In the post-interview phase, hiring managers must make decisions about candidates based on interview performance and other information gained during the selection process. As discussed, candidates’ interview performance, and the interviewer’s attention to and memory of the interview can be influenced by group membership. Thus, by the time this stage in the process is reached, the decision-making process may have already been affected. During the postdecision-making stage there are additional biasing processes that may enter the interview: shifting standards, status characteristics, and constructed criteria. The shifting standards model describes a process by which stereotypical expectations may bias ratings. The model, proposed by Biernat, Manis, and Nelson (1991) suggests that descriptive scale anchors may acquire different meanings for members of different social groups. This phenomenon has been demonstrated in many domains, including comparisons of men and women on height, weight, income, verbal ability, athleticism, and competence; and comparisons of Blacks and Whites on verbal ability and athletic ability (Biernat, Kobrynowicz, & Weber, 2003; Biernat & Manis, 1994; Biernat, Manis, & Nelson, 1991). According to the shifting standards model, members of groups with lower expectations (i.e., more negative stereotypes) may receive higher performance ratings on subjective rating scales (e.g., “very good”) than objectively equivalent members of groups with higher expectations. This process may obscure discrimination when ratings on subjective response scales are examined. Two applicants of different sub-groups may receive ratings that, on the surface, look equivalent; however, when a decision has to be made between the two, the decision may favor the member of the group with higher expectations, as their latent response scale implies higher objective performance. One such result was found by Cunningham and Macan (2007) in their study of pregnant and non-pregnant applicants. They found no significant mean differences between equally qualified pregnant and nonpregnant applicants on a subjective scale assessing perceived qualifications. However, when the participants were required to decide which of the two applicants would be offered the job, significant hiring discrimination against pregnant applicants was found. Furthermore, it has been found that women, when compared with men, may be more likely to make the “shortlist” for a job opening but less likely to receive the job offer (Biernat & Fuegen, 2001).
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
317
Thus, the shifting standards approach may be one explanation for the small effect sizes found by Landy in his meta-analysis of subgroup differences in performance evaluations conducted in field settings (d = 0.05–0.10; as cited in Landy, 2008). It has been found that subjective rating scales (e.g., Likert-type scales) mask rating differences amongst ratees because, perhaps unconsciously, raters are placing members of different groups on different rating scales. Importantly, these effects only manifest on subjective (not objective) rating scales (Biernat & Kobrynowicz, 1997), and such effects have been found to intensify as cognitive load increases (Biernat, Kobrynowicz, & Weber, 2003). Thus, organizational researchers should be aware of the potential masking effects of subjective rating scales when examining discrimination effects. Discrimination may occur on both types of scales but may be more identifiable when objective rating scales are used. Second, the status characteristics model states that members of groups with lower expectations (negatively stereotyped groups) must perform above and beyond the levels required for groups with higher expectations, simply in order to be perceived as equivalent (Foschi, 1992). The notion is that, “unexpected performance elicits a stricter standard, because the judge requires stronger evidence that the performance was due to ability” (Foddy & Smithson, 1989: 76). For example, work products tend to be perceived as higher quality when attributed to men than to women (Swim, Borgida, Maruyama, et al., 1989), women tend to have less influence than men in mixed gender groups (Pugh & Wahrman, 1983), and people tend to believe that men will be more competent than women on novel tasks (Balkwell & Berger, 1996; Heilman & Guzzo, 1978). Thus, in contrast to the shifting standards model, status characteristics theory proposes that equivalently qualified members of low status groups will be rated lower relative to members of high status groups. Biernat and Kobrynowicz (1997) reconciled the seemingly contradictory shifting standards and status characteristics models. They proposed that perceivers set lower standards for low status groups when evaluating minimumcompetency levels, but they have higher standards for determining broad ability judgments of low status groups. In two studies using mock applicant screening, they demonstrated that raters required fewer pieces of evidence for minimum standards of competence for women (Study 1) and Blacks (Study 2) relative to men and Whites. However, perceivers required more pieces of evidence to determine high ability for members of these groups. These results suggest that different types of group-based biases may be found depending on whether the selection criterion is a minimum cut-off (shifting standards model) or an assessment of the applicant’s ability level (status characteristics model). Importantly, these effects were found only when objective rating scales were used (e.g., number of examples of skills), and not when subjective rating scales were used (e.g., few to many examples of skills). The third form of bias that may enter the post-interview stage relates to a tendency to construct criteria in a way that favors the preferred candidate.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
318 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY This bias may enter when several pieces of information must be combined to form a composite interview score. Despite evidence suggesting that actuarial/statistical models for weighting the various sources of information are more valid than clinical/expert judgment models, considerable resistance to the use of actuarial models persists (cf. Dawes, Faust, & Meehl, 1989). When the clinical model is used and individuals make unstructured decisions regarding how various scores should be weighted, the potential arises for constructed criteria bias to contribute to discrimination against non-preferred candidates, which may be members of stigmatized groups. Several empirical studies have documented constructed criteria effects. For example, Hodson, Dovidio, and Gaertner (2002) found that for college student admissions decisions, decision makers weighted either the applicant’s standardized test scores or high school GPA more heavily depending on which piece of information favored the preferred candidate. When a Black applicant was described as having strong standardized test scores but a moderate high school GPA, decision makers rated high school GPA as more important than standardized test scores. The reverse pattern was seen when the Black applicant was described as having a strong high school GPA but average standardized test scores. No such bias occurred for White applicants. In a different study (Uhlmann & Cohen, 2005: Experiment 1), a similar pattern was found for female applicants for the position of police chief. Furthermore, this process may produce reverse discrimination effects, where the historically disadvantaged applicant may be favored (Norton, Sommers, Vandello, et al., 2006; Norton, Vandello, Biga, et al., 2008; Uhlmann & Cohen, 2005: Experiment 2). This process seems to occur even more strongly when decision makers are held accountable for their decisions (Norton, Vandello, & Darley, 2004). Decision makers showed memory bias favoring the preferred applicant’s qualifications (Studies 3 and 4) and showed even stronger rates of choice of the preferred candidate and minimization of race in their explanation of their decision making strategies than decision makers who were not held accountable (Study 5). One potential remedy for the constructed criteria bias may be to develop a well-validated actuarial model for determining interview scores. In such a model, the weights for various pieces of interview information would be determined empirically, according to their relationships with performance outcomes. Although some organizations may resist such a model, and in this model bias may still enter into the scores for the individual components, this strategy should eliminate the constructed criteria bias in determining interview scores. A more controversial solution is to have decision makers commit to the relative importance of various pieces of information from the interview in advance. This strategy was found effective by Uhlmann and Cohen (2005: Experiment 3) but ineffective by Norton, Vandello, and Darley, (2004: Study 6). Thus, future research is needed to determine whether committing to the
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
319
weights of sources of interview information in advance effectively prevents constructed criterion bias.
STRATEGIES FOR PREVENTING IMPLICIT BIAS In this final section, several issues that apply broadly across the stages of the interview process are discussed. Key to this discussion is the distinction between stereotype activation and stereotype application (Gilbert & Hixon, 2001; Kunda & Spencer, 2003). Activation refers to the mental initiation of the stereotypical belief structure or evaluative valence in the perceiver’s memory. This activation process was previously thought to be inevitable upon encountering a member of a stigmatized group (however, recent research has called this assumption into question). Application refers to the perceiver’s use of the activated mental concepts in making judgments or in forming behavior. Several researchers have noted that application of stereotypes and prejudice may be prevented under the proper conditions – specifically when one is motivated to do so and when one has sufficient cognitive resources to accomplish that goal (e.g., Bodenhausen, Macrae, & Sherman, 1999; Fazio & TowlesSchwen, 1999; Fiske, Lin, & Neuberg, 1999; van Knippenberg, Dijksterhuis, & Vermeulen, 1999). We begin with a discussion of motivational effects on the application and activation of automatic stereotypes and prejudice, followed by a discussion of the effects of cognitive load. Finally, we discuss potential avenues for preventing bias by changing the implicit attitudes and stereotypes themselves. Motivational Effects on Stereotypes and Prejudice Several studies have documented the effects of various motivational states and goals on prevention of the application of automatic stereotypes and prejudice. For example, motivation to control prejudice moderates the association between implicit and explicit measures (e.g., Akrami & Ekehammar, 2005; Dunton & Fazio, 1997; Fazio, Jackson, Dunton, et al., 1995; Gawronski, Geschke, & Banse, 2003; Hofmann, Gschwendener, & Schmitt, 2005; Payne, Cheng, Govorun, et al., 2005), suggesting that when motivated, people can and do prevent the application of implicit bias to their explicit ratings. Perhaps even more exciting, some research suggests that temporary and chronic motivational states may actually prevent the initial activation of stereotypes and prejudice, therefore by definition preventing their application. Devine, Plant, Amodio, et al. (2002) found that stereotype activation and application were minimized when individuals were high in internal motivation to control prejudice and low in external motivation to control prejudice. Activation was measured both on a sequential priming task (Study 1) and IAT (Studies 2 and 3), and application was measured using explicit self-report measures. Particularly important is the notion that this study found that it is
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
320 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY internal (i.e., intrinsic) motivation to control prejudice, rather than external (i.e., socially imposed) motivation, that most effectively prevented activation and application. Kunda and Spencer (2003) have suggested that three types of motivations affect whether stereotypes will be applied and activated: comprehension, selfenhancement, and the desire to avoid prejudice. Stereotypes will be activated when they are believed to provide information supporting comprehension or understanding of a target, and not activated when they are believed to be irrelevant or distracting. For example, stereotypes may be activated on an impression-formation task (Hoshino-Browne & Kunda, 2000). It is interesting to note that individuals higher in self-perceived objectivity have been found to be particularly likely to act on their stereotypical beliefs in a hiring context (Uhlmann & Cohen, 2007). Self-enhancement motives relate to stereotypic activation in that people may be more likely to stereotype in order to derogate the outgroup when doing so increases their self-image, such as when they have just received negative feedback from an outgroup member (Fein & Spencer, 1997; Sinclair & Kunda, 1999). However, stereotypes will not be activated when doing so prevents the desired self-enhancement, such as when receiving praise from an outgroup member (Sinclair & Kunda, 1999). In addition, several forms of motivation to avoid prejudice have been found to prevent stereotype activation and application, including chronic egalitarian goals (Moskowitz, Gollwitzer, Wasel, et al., 1999), salience of egalitarian norms (Fein, Hoshino-Brown, Davies, et al., 2003), and intrinsic and extrinsic motivation to control prejudice at both the implicit and explicit levels (e.g., Devine, Plant, Amodio, et al., 2002; Gabriel, Banse, & Hug, 2007; Hausmann & Ryan, 2004). Some research suggests that an organizational culture that promotes a strong climate for diversity and social egalitarian norms can provide extrinsic motivation to control prejudice, whereas a culture of bias can allow it to occur (Ziegert & Hanges, 2005). However, a climate-based intervention may be effective primarily at the stereotype application phase, as other studies have found that external motivation to control prejudice is positively associated with stereotype activation (e.g., Devine, Plant, Amodio, et al., 2002; Hausmann & Ryan, 2004). Research should continue to investigate the effectiveness of interventions designed to increase external (e.g., rewards and punishments) and internal (e.g., persuasion) motivations to control prejudice in organizations. Cognitive Load and Thought Suppression The second prerequisite for being able to override automatic attitudes’ effects on evaluations and behavior is availability of cognitive resources (Fazio, 1990; Fazio & Towles-Schwen, 1999; Schuette & Fazio, 1995). When under heavy cognitive load, such as when required to perform under time pressure, high stress, or multitasking conditions, even motivated perceivers will likely be
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
321
unable to control the effects of automatic constructs. Of course, if the individual’s level of automatic bias is already low, the cognitive load will have few effects (Gonsalkorale, von Hippel, Sherman, et al., 2009). Although cognitive load is often thought to increase the effects of stereotypes on judgment and behavior, this is not always the case. In fact, Gilbert and Hixon (1991) found that perceivers were unable to activate stereotypes when under sufficiently high cognitive load. Even if stereotypes do become activated, motivated perceivers with available cognitive resources attempt to correct their judgments for the degree and direction of bias they believe occurred – however, this process depends on both motivation and availability of resources (Wegener & Petty, 1995). Furthermore, it is important for research to examine whether these corrections are accurate, or whether they might result in undercorrection or overcorrection for perceived bias. Incorporating elements of diversity issues in interviewer training programs to raise awareness of automatic stereotypes and prejudice may be beneficial. If interviewers become aware of the possibility that automatic processes might bias their judgment, they would be more likely to suppress stereotypes and prejudice and arrive at unbiased conclusions. However, under cognitive load, such attempts to suppress prejudice have been shown to have ironic effects (for a review see Wenzlaff & Wegner, 2000). Following suppression attempts, stereotypes have been shown to “rebound,” arising with even greater insistence than if no suppression attempt had been made (Macrae, Bodenhausen, & Milne, 1998; Macrae, Bodenhausen, Milne, et al., 1994). This suggests that in a job interview context, a motivated interviewer with available cognitive resources may be able to suppress stereotypical thinking about an interviewee. However, as resources fade or after the interview concludes, these stereotypes may rebound, affecting subsequent interviews or interactions with other employees or how the interviewee is later viewed. This is particularly likely to occur when the perceiver is under cognitive load (Wegner, 1994, 1997). Further compounding the problem of stereotype suppression, research has shown that intergroup interactions are cognitively demanding, with cognitive resources and self-regulatory capacity being depleted following the interaction, particularly for those higher in implicit prejudice (Richeson & Shelton, 2003). Corroborating this result, neuroimaging research has found that more implicitly prejudiced people (as measured by an IAT) show more active executive function during interracial interaction and greater cognitive resource depletion following that interaction than do less implicitly prejudiced people (Richeson, Baird, Gordon, et al., 2003). Therefore, those individuals who are most prejudiced and most in need of ability to suppress stereotypical and prejudiced thoughts during intergroup interactions may be those for whom suppression is most difficult. Future research should examine both person and situation effects on interviewer cognitive load, suppression, and interview outcomes in order to determine effective organizational interventions.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
322 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Changing Implicit Attitudes and Stereotypes Although several researchers have contended that implicit constructs are old, highly stable, and resistant to change (e.g., Connor & Feldman-Barrett, 2005; DeHart, Pelham, & Tennen, 2006; Gregg, Seibt, & Banaji, 2006; Jordan, ˜ Spencer, Zanna, et al., 2003; Petty, Tormala, Brinol, et al., 2006; Rydell & McConnell, 2006; Sinclair, Dunn, & Lowery, 2005; Wilson, Lindsey, & Schooler, 2000), these assumptions have been recently questioned (e.g., Blair, 2002; Gawronski, Deutsch, Mbirkou, et al., 2008; Gawronski, LeBel, & Peters, 2007). Empirical research has shown that implicit attitudes can be flexible, context-dependent, and relatively easily formed and changed through conditioning processes. Several studies have found that implicit attitudes and stereotypes are contextdependent and can therefore potentially be affected by organizational contextbased interventions. For example, the presence of a Black experimenter was found to decrease prejudice as measured by an IAT (Lowery, Hardin, & Sinclair, 2001). Reminding respondents of well-liked Black exemplars (e.g., Denzel Washington) and disliked White exemplars (e.g., Jeffrey Dahmer) has also been found to reduce implicit prejudice on a race IAT, with similar results found for a manipulation involving age (Dasgupta & Greenwald, 2001). The relative proportion of women in leadership roles in one’s environment has also been associated with levels of implicit gender stereotypes (Dasgupta & Asgari, 2004). Using images embedded in various backgrounds as stimuli for an IAT, Wittenbrink, Judd, and Park (2001) found that context mattered, with implicit prejudice lower for images of African-Americans embedded in positive (e.g., family barbeque, church) as opposed to negative (e.g., gang scene, street corner) contexts. In a follow-up study, it was found that not only background context matters, but also the target’s social role. A Black individual in a prison context, wearing a prisoner’s uniform, was evaluated negatively, but the same individual in the same context but wearing a suit (implying an attorney role) was evaluated positively (Barden, Maddux, Petty, et al., 2004). Thus, several contextual variations may affect automatic evaluations of individuals in organizations. Similarly, it has been found that the relative salience of different ¨ social categories affects prejudice (e.g., Kuhnen, Schiessl, Bauer, et al., 2001; Mitchell, Nosek, & Banaji, 2003; Pratto & Shih, 2000; Steele & Ambady, 2006). Even inanimate objects (e.g., a briefcase) present in the environment may prime concepts or associations (“material priming”; e.g., Kay, Wheeler, Bargh, et al., 2004). Another contextual variable that has been found to affect stereotyping and prejudice is one’s anticipated role in an upcoming interaction. When expecting to occupy the high status role in an interaction, Whites had more negative IAT scores than when expecting to occupy the low status role (Richeson & Ambady, 2003). However, the pattern for job status with gender was different, such that men had more negative gender attitudes when expecting to occupy
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
323
the low status role (Richeson & Ambady, 2001). Increased stereotyping has also been observed under conditions of mortality salience – when the perceiver is thinking about death (Schimel, Simon, Greenberg, et al., 1999), when the perceiver is angry (DeSteno, Dasgupta, Bartlett, et al., 2004), and, in some cases, when the perceiver is in a darker room (Schaller, Park, & Mueller, 2003). An interesting avenue for organizational researchers is to investigate how the interview context may affect primed constructs, category salience, and evaluations of interviewees from various subgroups. Consistent with the notion that automatic constructs are associative in nature (e.g., Gawronski & Bodenhausen, 2006), several studies have found that such attitudes can be changed using conditioning techniques. Implicit attitudes have been conditioned for previously novel stimuli including Chinese characters (Murphy, Monahan, & Zajonc, 1995) and Pokemon characters (Olson & Fazio, 2001). Researchers have also created a subliminal “mere exposure” effect in which objects that were repeatedly presented subliminally at exposures of only 5 ms were significantly better-liked than other objects (Monahan, Murphy, & Zajonc, 2000). In one study, the evaluative valence of neutral words was significantly changed by repeatedly pairing the words with positive or negative stimuli (De Houwer, Baeyens, & Eelen, 1994). Conditioning manipulations have also been found to produce significant changes in implicit attitudes toward social groups. Karpinski and Hilton (2001) manipulated implicit attitudes toward the elderly by pairing the concept of elderly with either positive or negative stimuli (and the concept of youth with oppositely valenced stimuli) 200 times. Olson and Fazio (2006) used a similar procedure to successfully manipulate implicit attitudes toward Blacks and found that the effects persisted on a retest 2 days later. Work on supraliminal evaluative conditioning of stimuli suggests that these effects may be quite longlived. Effects have been found to persist for 2 months (Baeyens, Crombez, van den Bergh, et al., 1988) and even 18 months (Levey & Martin, 1975). Although the longevity of manipulations for social groups (which may be affected by outside experiences such as media portrayals) is unknown, conditioning-based interventions may have potential for inspiring training programs designed to combat bias at its source, and we believe that organizational researchers should investigate the potential effectiveness of such interventions. One such training program is described by Kawakami, Dovidio, Moll, et al. (2000). In this technique, pictures of Black and White individuals were displayed on a computer screen along with either stereotypical or nonstereotypical labels. Participants were instructed to respond “no” to stereotypical labels and “yes” to non-stereotypical labels. The notion is that participants were practicing “unlearning” stereotypical associations. After the training, participants performed better (less stereotypic associations) on subsequent categorization tasks. Control participants showed no change. Importantly, the training effects persisted upon a retest 24 hours later. Subsequent research has identified affirmation of the counter-stereotypic associations, rather than the
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
324 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY negation of stereotypic associations, as the key driver of these effects (Gawronski, Deutsch, Mbirkou, et al., 2007). In a follow-up study, similar effects were found for the conditioning-type training regarding gender stereotypes in a simulated applicant selection context (Kawakami, Dovidio, & van Kamp, 2005). Participants were significantly more likely to recommend hiring the male than the female across conditions; however, this tendency was significantly reduced following the counter-stereotype conditioning. The authors found that in order for the conditioning manipulation to be effective, participants needed to be unaware that the conditioning training and the selection task were linked. When participants were aware of the link, they identified the potential effects of the training and (when they possessed sufficient cognitive resources to do so), they corrected for those effects in their ratings of the applicants. Thus, while conditioning-based interviewer training may have potential benefits in removing bias from selection systems, it will be important for researchers to examine the reactions of trainees to such manipulations in order to assess the potential for transfer of training to occur.
CONCLUSIONS The employment interview is very popular, not only within organizations as a frequently used selection device, but also among scholars as a topic of research for nearly 100 years. While we have learned much about the interview over the years, the nature of discrimination in the interview process has changed, becoming more subtle and interpersonal. While discrimination based on explicit prejudice persists, the legal system has also begun to note discrimination based on implicit processes and to hold organizations accountable for such discrimination. Moreover, we believe that it is important for organizations to examine any factors that might decrease interview validity and fairness and to strive to be inclusive of all individuals. As such, awareness of the potential biasing role of implicit cognition is likely to spread among business organizations, and those organizations may turn to us, organizational psychologists, for information and potential solutions. Therefore, it is incumbent on us to inform ourselves on the nature, measurement, and operation of implicit cognition in organizations and to conduct applied research focusing on potential interventions. Several studies indicate that implicit attitudes and stereotypes may be quite malleable and subject to intervention-based change. The implication of these findings is that we may be able to identify interventions to reduce or eliminate implicit discrimination at its source, rather than relying solely on the motivation and cognitive capacity of individuals to attempt to suppress the application of automatic cognition. We believe that discovering ways to eliminate implicit bias in the employment interview process is a pressing, timely, and advantageous direction for future research.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
325
Our colleague, Kandola (2009) communicates our sentiments best in his book, The Value of Difference: It seems to me that we’re acting like the drunk who lost his house keys in the road, but searched for them only under the streetlamp because that is where the light was. We’re looking for fixes for diversity in the places where we happen to have some solutions, rather than venturing into the unlit areas where the problems really lie. To achieve true diversity, we need to look in the dark places: within our own prejudices and habits. We have to face the forces of discrimination which have been driven underground by the early progress of diversity campaigners – and which exist in us all . . . diversity is being held back by unconscious bias. (Kandola, 2009: 2–3)
We hope that our review inspires others to explore these “unlit” places (and continue work in lit areas) within the employment interview process.
REFERENCES Akrami, N., & Ekehammar, B. (2005). The association between implicit and explicit prejudice: The moderating role of motivation to control prejudiced reactions. Scandinavian Journal of Psychology, 46(4), 361–6. Arvey, R.D., & Campion, J.E. (1982). The employment interview: A summary and review of recent research. Personnel Psychology, 35, 281–322. ¨ Asendorpf, J.B., Banse, R., & Mucke, D. (2002). Double dissociation between implicit and explicit personality self-concept: The case of shy behavior. Journal of Personality and Social Psychology, 83, 380–93. Baeyens, F., Crombez, G., van den Bergh, O., & Eelen, P. (1988). Once in contact always in contact: Evaluative conditioning is resistant to extinction. Advances in Behavior Research and Therapy, 10, 179–99. Balkwell, J.W., & Berger, J. (1996). Gender, status, and behavior in task situations. Social Psychology Quarterly, 59, 273–83. Banks, R.R., & Ford, R.T. (2009). (How) does unconscious bias matter? Law, politics, and racial inequality. Emory Law Journal, 58, 1053–122. Banerjee, A., Bertrand, M., Datta, S., & Mullainathan, S. (2009). Labor market discrimination in Delhi: Evidence from a field experiment. Journal of Comparative Economics, 37, 14–27. Barden, J., Maddux, W.W., Petty, R.E., & Brewer, M.B. (2004). Contextual moderation of racial bias: The impact of social roles on controlled and automatically activated attitudes. Journal of Personality and Social Psychology, 87, 5–22. Bargh, J.A. (1994). The four horsemen of automaticity: Awareness, intention, efficiency, and control in social cognition. In R.S. Wyer Jr. & T.K. Srull (Eds), Handbook of Social Cognition (volume 1–2), (pp. 1–40). Hillsdale, NJ: Lawrence Erlbaum. Bargh, J.A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71, 230–44. Bartlett, K.T. (2009). Making good on good intentions: The critical role of motivation in reducing implicit workplace discrimination. Virginia Law Review, 95, 1893– 971. BBC News (2007). Mother wins discrimination case. http://news.bbc.co.uk/2/hi/uk news/scotland/glasgow and west/7078880.stm (accessed 21 June 2010).
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
326 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Becker, D.V., Kenrick, D.T., Neuberg, S.L., Blackwell, K.C., & Smith, D.M. (2007). The confounded nature of angry men and happy women. Journal of Personality and Social Psychology, 92, 179–90. Bendick, M., Jackson, C.W., & Romero, J.H. (1997). Employment discrimination against older workers: An experimental study of hiring practices. Journal of Aging and Social Policy, 8(4), 25–46. Bertrand, M., & Mullainathan, S. (2004). Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination. American Economic Review, 94(4), 991–1013. Biernat, M., & Fuegen, K. (2001). Shifting standards and the evaluation of competence: Complexity in gender-based judgment and decision making. Journal of Social Issues, 57, 707–24. Biernat, M., & Kobrynowicz, D. (1997). Gender- and race-based standards of competence: Lower minimum standards but higher ability standards for devalued groups. Journal of Personality and Social Psychology, 72, 544–57. Biernat, M., Kobrynowicz, D., & Weber, D.L. (2003). Stereotypes and shifting standards: Some paradoxical effects of cognitive load. Journal of Applied Social Psychology, 33, 2060–79. Biernat, M., & Ma, J.E. (2005). Stereotypes and the confirmability of trait concepts. Personality and Social Psychology Bulletin, 31, 483–95. Biernat, M., & Manis, M. (1994). Shifting standards and stereotype-based judgments. Journal of Personality and Social Psychology, 66, 5–20. Biernat, M., Manis, M., & Nelson, T.E. (1991). Stereotypes and standards of judgment. Journal of Personality and Social Psychology, 60, 485–99. Blair, I.V. (2002). The malleability of automatic stereotypes and prejudice. Personality and Social Psychology Review, 6, 242–60. Blair, I.V., Ma, J.E., & Lenton, A.P. (2001). Imagining stereotypes away, the moderation of implicit stereotypes through mental imagery. Journal of Personality and Social Psychology, 81, 828–41. Bodenhausen, G.V., & Lichtenstein, M. (1987). Social stereotypes and informationprocessing strategies: The impact of task complexity. Journal of Personality and Social Psychology, 52, 871–80. Bodenhausen, G.V., Macrae, C.N., & Sherman, J.S. (1999). On the dialectics of discrimination: Dual processes in social stereotyping. In S. Chaiken & Y. Trope (Eds), Dual-Process Theories in Social Psychology (pp. 271–90). New York, NY: Guilford Press. Bragger, J.D., Kutcher, E., Morgan, J., & Firth, P. (2002). The effects of the structured interview on reducing biases against pregnant job applicants. Sex Roles, 46, 215–26. Brecher, E.D., Bragger, J.D., & Kutcher, E.J. III. (2006). The structured interview: Reducing bias toward job applicants with physical disabilities. Employee Responsibilities and Rights Journal, 18, 155–70. Chapman, D.S., Uggerslev, K.L., & Webster, J. (2003). Applicant reactions to faceto-face and technology-mediated interviews: A field investigation. Journal of Applied Psychology, 88, 944–53. Chen, M., & Bargh, J.A. (1997). Nonconscious behavior confirmation processes: The self-fulfilling consequences of automatic stereotype activation. Journal of Experimental Social Psychology, 33, 541–60. Cleveland, J.N., Festa, R.M., & Montgomery, L. (1988). Applicant pool composition and job perceptions: Impact on decisions regarding an older applicant. Journal of Vocational Behavior, 32, 112–25. Conner, T., & Feldman-Barrett, L. (2005). Implicit self-attitudes predict spontaneous affect in daily life. Emotion, 5, 476–88.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
327
Conway, J.M., Jako, R.A., & Goodman, D.F. (1995). A meta-analysis of interrater and internal consistency reliability of selection interviews. Journal of Applied Psychology, 80, 565–79. Correll, S.J., Benard, S., & Paik, I. (2007). Getting a job: Is there a motherhood penalty? American Journal of Sociology, 112, 1297–338. Crisp, R.J., & Turner, R.N. (2009). Can imagined interactions produce positive perceptions? Reducing prejudice through simulated social contact. American Psychologist, 64, 231–40. Cunningham, J., & Macan, T. (2007). Effects of applicant pregnancy on hiring decisions and interview ratings. Sex Roles, 57, 497–508. Cunningham, W.A., Johnson, M.K., Raye, C.L., Gatenby, J.C., Gore, J.C., & Banaji, M.R. (2004). Separable neural components in the processing of black and white faces. Psychological Science, 15, 806–13. Dalgin, R.S., & Bellini, J. (2008). Invisible disability disclosure in an employment interview: Impact on employers’ hiring decisions and views of employability. Rehabilitation Counseling Bulletin, 52, 6–15. Dasgupta, N., & Asgari, S. (2004). Seeing is believing: Exposure to counterstereotypic women leaders and its effect on the malleability of automatic gender stereotyping. Journal of Experimental Social Psychology, 40, 642–58. Dasgupta, N., & Greenwald, A.G. (2001). On the malleability of automatic attitudes: Combating automatic prejudice with images of admired and dislikes individuals. Journal of Personality and Social Psychology, 81, 800–14. Dawes, R.M., Faust, D., & Meehl, P.E. (1989). Clinical versus actuarial judgment. Science, 243, 1668–74. DeHart, T., Pelham, B.W., & Tennen, H. (2006). What lies beneath: Parenting style and implicit self-esteem. Journal of Experimental Social Psychology, 42, 1–17. DeHouwer, J. (2003). The extrinsic affective Simon task. Experimental Psychology, 50, 77–85. De Houwer, J., Baeyens, F., & Eelen, P. (1994). Verbal evaluative conditioning with undetected US presentations. Behaviour Research and Therapy, 32, 629–33. DeSteno, D., Dasgupta, N., Bartlett, M.Y., & Cajdrie, A. (2004). Prejudice from thin air: The effect of emotion on automatic intergroup attitudes. Psychological Science, 15, 319–24. Devine, P.G., Plant, E.A., Amodio, D.M., Harmon-Jones, E., & Vance, S.L. (2002). The regulation of explicit and implicit race bias: The role of motivations to respond without prejudice. Journal of Personality and Social Psychology, 85, 835–48. Ding, V.J., & Stillman, J.A. (2005). An empirical investigation of discrimination against overweight female job applicants in New Zealand. New Zealand Journal of Psychology, 34(3), 139–48. Dipboye, R.L. (1994). Structured and unstructured interviews: Beyond the job-fit model. In G. Ferris (Ed.), Frontiers of Industrial and Organizational Psychology: Personnel Selection. San Francisco, CA: Jossey-Bass. Dipboye, R.L. (1997). Structured selection interviews: Why do they work? Why are the underutilized? In N. Anderson & P.H. Lodsl (Eds), International Handbook of Selection and Assessment (pp. 455–75). New York: John Wiley & Sons. Dipboye, R.L., & Colella, A. (Eds). (2005). Discrimination at Work: The Psychological and Organizational Bases. Mahwah, NJ: Lawrence Erlbaum. Dipboye, R.L., & Macan, T.M. (1988). A process view of the selection/recruitment interview. In R.S. Schuler, S.A. Youngblood, & V.L. Huber (Eds), Readings in Personnel and Human Resource Management. MN: West Publishing. Dovidio, J.F., & Gaertner, S.L. (2000). Aversive racism and selection decisions: 1989 and 1999. Psychological Science, 11, 315–9.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
328 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Dovidio, J.F., Kawakami, K., & Gaertner, S.L. (2002). Implicit and explicit prejudice and interracial interaction. Journal of Personality and Social Psychology, 82, 62–8. Dovidio, J.F., Kawakami, K., Johnson, C., Johnson, B., & Howard, A. (1997). On the nature of prejudice: Automatic and controlled processes. Journal of Experimental Social Psychology, 33, 510–40. Drydakis, N. (2009). Sexual orientation discrimination in the labour market. Labour Economics, 16, 364–72. Dunton, B.C., & Fazio, R.H. (1997). An individual difference measure of motivation to control prejudiced reactions. Personality and Social Psychology Bulletin, 23, 316–26. Equal Employment Opportunity Commission (2010). Pregnancy Discrimination Charges EEOC and FEPAs Combined: FY 1997-FY 2009. http://www.eeoc.gov/eeoc/statistics/ enforcement/pregnancy.cf (accessed 2 July 2010). Equal Employment Opportunity Commission. (2010 Press release) 20 May 2010. Orkin Pest Control sued by EEOC for age and religious discrimination in hiring and advertising. http://www.eeoc.gov/eeoc/newsroom/release/5-20-10a.cfm (accessed 2 July 2010). Fazio, R.H. (1990). A practical guide to the use of response latency in social psychological research. In C. Hendrick, & M.S. Clark (Eds), Research Methods in Personality and Social Psychology (pp. 74–97). Thousand Oaks, CA: Sage Publications. Fazio, R.H., Jackson, J.R., Dunton, B.C., & Williams, C.J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Journal of Personality and Social Psychology, 69, 1013–27. Fazio, R.H., Sanbonmatsu, D.M., Powell, M.C., & Kardes, F.R. (1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229–38. Fazio, R.H., & Towles-Schwen, T. (1999). The MODE model of attitude-behavior processes. In S. Chaiken, & Y. Trope (Eds), Dual-Process Theories in Social Psychology (pp. 97–116). NY: Guilford Press. Fein, S., Hoshino-Browne, E., Davies, P.G., & Spencer, S.J. (2003). Self-image maintenance goals and sociocultural norms in motivated social perception. In S.J. Spencer, S. Fein, M.P. Zana, & J.M. Olson (Eds), Motivated Social Perception: The Ontario Symposium (volume 9), (pp. 21–44). Mahwah, NJ: Lawrence Erlbaum. Fein, S., & Spencer, S. (1997). Prejudice as self-image maintenance: Affirming the self through derogating others. Journal of Personality and Social Psychology, 73, 31–44. Finkelstein, L.M., Demuth, R.L.F., & Sweeney, D.L. (2007). Bias against overweight job applicants: Further explorations of when and why. Human Resource Management, 46, 203–22. Fiske, S.T., Lin, M., & Neuberg, S.L. (1999). The continuum model: Ten years later. In S. Chaiken & Y. Trope (Eds), Dual-Process Theories in Social Psychology (pp. 231–54). New York, NY: Guilford Press. Foddy, M., & Smithson, M. (1989). Fuzzy sets and double standards: Modeling the process of ability inference. In J. Berger, M. Zelditch, & B. Anderson (Eds), Sociological Theories in Progress: New Formulations (pp. 73–99). Newbury Park, CA: Sage. Ford, T.E., Gambino, F., Lee, H., Mayo, E., & Ferguson, M.A. (2005). The role of accountability in suppressing managers’ pre-interview biases against African-American sales job applicants. Journal of Personal Selling and Sales Management, 24(2), 113– 24. Foschi, M. (1992). Gender and double standards for competence. In C.L. Ridgeway (Ed.), Gender, Interaction, and Inequality (pp. 181–207). New York: Springer-Verlag. Frantz, C.M., Cuddy, A.J.C., Burnett, M., Ray, H., & Hart, A. (2004). A threat in the computer: The race Implicit Association Test as a stereotype threat experience. Personality and Social Psychology Bulletin, 30, 1611–24.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
329
Gabriel, U., Banse, R., & Hug, F. (2007). Predicting private and public helping behavior by implicit attitudes and the motivation to control prejudiced reactions. British Journal of Social Psychology, 46, 365–82. Galgano, S.W. (2009). Barriers to reintegration: An audit study of the impact of race and offender status on employment opportunities for women. Social Thought and Research, 30, 21–37. Garcia, M.F., Posthuma, R.A., & Colella, A. (2008). Fit perceptions in the employment interview: The role of similarity, liking, and expectations. Journal of Occupational and Organizational Psychology, 81, 173–89. Gawronski, B., & Bodenhausen, G.V. (2006). Associative and propositional processes in evaluation: An integrative review of implicit and explicit attitude change. Psychological Bulletin, 132, 692–731. Gawronski, B., Deutsch, R., Mbirkou, S., Seibt, B., & Strack, F. (2008). When “Just Say No” is not enough: Affirmation versus negation training and the reduction of automatic stereotype activation. Journal of Experimental Social Psychology, 44, 370–7. Gawronski, B., Geschke, D., & Banse, R. (2003). Implicit bias in impression formation: Associations influence the construal of individuating information. European Journal of Social Psychology, 33, 573–89. Gawronski, B., Hofmann, W., & Wilbur, C.J. (2006). Are “implicit” attitudes unconscious? Consciousness and Cognition: An International Journal, 15, 485–99. Gawronski, B., LeBel, E.P., & Peters, K.R. (2007). What do implicit measures tell us?: Scrutinizing the validity of three common assumptions. Perspectives on Psychological Science, 2, 181–93. Gawronski, B., & Strack, F. (2004). On the propositional nature of cognitive consistency: Dissonance changes explicit, but not implicit attitudes. Journal of Experimental Social Psychology, 40, 535–42. Gilbert, D.T., & Hixon, J.G. (1991). The trouble of thinking: Activation and application of stereotypic beliefs. Journal of Personality and Social Psychology, 60, 509–17. Gilbert, D.T., & Hixon, J.G. (2001). The trouble of thinking: Activation and application of stereotypic beliefs. Journal of Personality and Social Psychology, 60, 509–17. Goff, P.A., Steele, C.M., & Davies, P.G. (2008). The space between us: Stereotype threat and distance in interracial contexts. Journal of Personality and Social Psychology, 94, 91–107. Goffman, E. (1963). Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs, NJ: Prentice Hall. Gonsalkorale, K., von Hippel, W., Sherman, J.W., & Klauer, K.C. (2009). Bias and regulation of bias in intergroup interactions: Implicit attitudes toward Muslims and interaction quality. Journal of Experimental Social Psychology, 45, 161–6. Greenwald, A.G., McGhee, D.E., & Schwartz, J.L.K. (1998). Measuring individuals differences in implicit cognition: The Implicit Association Test. Journal of Personality and Social Psychology, 74, 1464–80. Greenwald, A.G., Poehlman, T.A., Uhlmann, E.L., & Banaji, M.R. (2009). Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. Journal of Personality and Social Psychology, 97, 17–41. Gregg, A.P., Seibt, B., & Banaji, M.R. (2006). Easier done than undone: Asymmetry in the malleability of implicit preferences. Journal of Personality and Social Psychology, 90, 1–20. Haines, E.L., & Sumner, K.E. (2006). Implicit measurement of attitudes, stereotypes, and self-concepts in organizations: Teaching old dogmas new tricks. Organizational Research Methods, 9, 536–53. Harris, M. (1989). Reconsidering the employment interview: A review of recent literature and suggestions for future research. Personnel Psychology, 42, 691–726.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
330 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Hausmann, L.R.M., & Ryan, C.S. (2004). Effects of external and internal motivation to control prejudice on implicit prejudice: The mediating role of efforts to control prejudiced responses. Basic and Applied Social Psychology, 26, 215–25. Hayes, T.L., & Macan, T.H. (1997). Comparison of the factors influencing interviewer hiring decisions for applicants with and those without disabilities. Journal of Business and Psychology, 11, 357–71. Hebl, M.R., Foster, J.B., Mannix, L.M., & Dovidio, J.F. (2002). Formal and interpersonal discrimination: A field study of bias toward homosexual applicants. Personality and Social Psychology Bulletin, 28, 815–25. Hebl, M.R., & Heatherton, T.F. (1998). The stigma of obesity in women: The difference is black and white. Personality and Social Psychology Bulletin, 24, 417– 26. Hebl, M.R., King, E.B., Glick, P., Singletary, S.L., & Kazama, S. (2007). Hostile and benevolent reactions toward pregnant women: Complementary interpersonal punishments and rewards that maintain traditional roles. Journal of Applied Psychology, 92, 1499–511. Hebl, M.R., & Kleck, R.E. (2002). Acknowledging one’s stigma in the interview setting: Effective strategy or liability? Journal of Applied Social Psychology, 32, 223–49. Hebl, M.R., & Skorinko, J.L. (2005). Acknowledging one’s physical disability in the interview: Does “when” make a difference? Journal of Applied Social Psychology, 35, 2477–92. Hebl, M.R., & Turchin, J.M. (2005). The stigma of obesity: What about men? Basic and Applied Social Psychology, 27, 267–75. Heilman, M.E., & Guzzo, R.A. (1978). The perceived cause of work success as a mediator of sex discrimination in organizations. Organizational Behavior and Human Performance, 21, 346–57. Heilman, M.E., & Okimoto, T.G. (2008). Motherhood: A potential source of bias in employment decisions. Journal of Applied Psychology, 93, 189–98. Herold, K.P. (2000). Communication strategies in employment interviews for applicants with disabilities. In D.O. Braithwaite, & T.L. Thompson (Eds), Handbook of Communication and People with Disabilities: Research and Application (pp. 159–75). Mahwah, NJ: Lawrence Erlbaum. Hodson, G., Dovidio, J.F., & Gaertner, S.L. (2002). Processes in racial discrimination: Differential weighting of conflicting information. Personality and Social Psychology Bulletin, 28, 460–71. Hofmann, W., Gawronski, B., Gschwendner, T., Le, H., & Schmitt, M. (2005). A meta-analysis on the correlation between the Implicit Association Test and explicit self-report measures. Personality and Social Psychology Bulletin, 31, 1369–85. Hofmann, W., Gschwendner, T., & Schmitt, M. (2005). On implicit-explicit consistency: The moderating role of individual differences in awareness and adjustment. European Journal of Personality, 19, 25–49. Hofmann, W., Rauch, W., & Gawronski, B. (2007). And deplete us not into temptation: Automatic attitudes, dietary restraint, and self-regulatory resources as determinants of eating behavior. Journal of Experimental Social Psychology, 43, 497–504. Hoshino-Browne, E., & Kunda, Z. (2000). Activation and application processes of Asian stereotypes in realistic interaction. Paper presented at the 108th Annual Convention of the American Psychological Association, Washington, DC. Huffcutt, A.I., & Arthur, W. Jr. (1994). Hunter and Hunter (1984) revisited: Interview validity for entry-level jobs. Journal of Applied Psychology, 79, 184–90. Huffcutt, A.I., & Culbertson, S.S. (2010). Interviews. In S. Zedeck (Ed.), APA Handbook of Industrial and Organizational Psychology (volume 2), (pp. 185–203). Washington, DC: American Psychological Association.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
331
Huffcutt, A.I., & Woehr, D.J. (1999). Further analysis of employment interview validity: A quantitative evaluation of interviewer-related structuring methods. Journal of Organizational Behavior, 20, 549–60. Hugenberg, K., & Bodenhausen, G.V. (2003). Facing prejudice: Implicit prejudice and the perception of facial threat. Psychological Science, 14, 640–3. Hugenberg, K., & Bodenhausen, G.V. (2004). Ambiguity in social categorization: The role of prejudice and facial affect in race categorization. Psychological Science, 15, 342–5. Jordan, C.H., Spencer, S.J., Zanna, M.P., Hoshino-Browne, E., & Correll, J. (2003). Secure and defensive high self-esteem. Journal of Personality and Social Psychology, 85, 969–78. Kandola, B. (2009). The Value of Difference: Eliminating Bias in Organizations. Oxford, UK: Pearn Kandola. Karpinski, A., & Hilton, J.L. (2001). Attitudes and the Implicit Association Test. Journal of Personality and Social Psychology, 81, 774–88. Kawakami, K., Dovidio, J.F., Moll, J., Hermsen, S., & Russin, A. (2000). Just say no (to stereotyping): Effects of training in the negation of stereotypic associations on stereotype activation. Journal of Personality and Social Psychology, 78, 871–88. Kawakami, K., Dovidio, J.F., & van Kamp, S. (2005). Kicking the habit: Effects of nonstereotypic association training and correction processes on hiring decisions. Journal of Experimental Social Psychology, 41, 68–75. Kay, A.C., Wheeler, S.C., Bargh, J.A., & Ross, L. (2004). Material priming: The influence of mundane physical objects on situational construal and competitive behavioral choice. Organizational Behavior and Human Decision Processes, 95, 83–96. King, E.B. (2008). The effect of bias on the advancement of working mothers: Disentangling legitimate concerns from inaccurate stereotypes as predictors of advancement in academe. Human Relations, 61, 1677–711. King, E.B., & Cortina, J.M. (2010). The social and economic imperative of lesbian, gay, bisexual, and transgendered supportive organizational policies. Industrial and Organizational Psychology, 3, 69–78. King, E.B., Madera, J.M., Hebl, M.R., Knight, J.L., & Mendoza, S.A. (2006). What’s in a name? A multiracial investigation of the role of occupational stereotypes in selection decisions. Journal of Applied Social Psychology, 36, 1145–59. Klehe, U. (2004). Choosing how to choose: Institutional pressures affecting the adoption of personnel selection procedures. International Journal of Selection and Assessment, 12, 327–42. ¨ ¨ Kuhnen, U., Schiessl, M., Bauer, N., Paulig, N., Pohlmann, C., & Schmidthals, K. (2001). How robust is the IAT? Measuring and manipulating implicit attitudes of East- and West-Germans. Experimental Psychology, 48, 135–44. Kunda, Z., & Spencer, S.J. (2003). When do stereotypes come to mind and when do they color judgment? A goal-based theoretical framework for stereotype activation and application. Psychological Bulletin, 129, 522–44. Kutcher, E.J., & Bragger, J.D. (2004). Selection interviews of overweight job applicants: Can structure reduce the bias? Journal of Applied Social Psychology, 34, 1993– 2022. Landrine, H., Klonoff, E.A., Alcaraz, R., Scott, J., & Wilkins, P. (1995). Multiple variables in discrimination. In B. Lott, & D. Maluso (Eds), The Social Psychology of Interpersonal Discrimination (pp. 183–224). New York: Guilford Press. Landy, F.J. (2008). Stereotypes, bias, and personnel decisions: Strange and stranger. Industrial and Organizational Psychology, 1, 379–92. Lee, A.J. (2005). Unconscious bias theory in employment discrimination litigation. Harvard Law Review, 40, 481–6.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
332 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Levey, A.B., & Martin, I. (1975). Classical conditioning of human “evaluative” responses. Behavior Research and Therapy, 13, 221–6. Levin, S., Sinclair, S., Veniegas, R.C., & Taylor, P.L. (2002). Perceived discrimination in the context of multiple group membership. Psychological Science, 13, 557–60. Lievens, F., & De Paepe, A. (2004). An empirical investigation of interviewer-related factors that discourage the use of high structure interviews. Journal of Organizational Behavior, 25, 29–46. Lowery, B.S., Hardin, C.D., & Sinclair, S. (2001). Social influence effects on automatic racial prejudice. Journal of Personality and Social Psychology, 81, 842–55. Macan, T. (2009). The employment interview: A review of current studies and directions for future research. Human Resource Management Review, 19, 203–18. Macan, T.H., & Dipboye, R.L. (1988). The effects of interviewers’ initial impressions on information gathering. Organizational Behavior and Human Decision Processes, 42, 364–87. Macan, T.H., & Dipboye, R.L. (1994). The effects of the application on processing of information from the employment interview. Journal of Applied Social Psychology, 24, 1291–314. Macan, T.H., & Hayes, T.L. (1995). Both sides of the employment interview interaction: Perceptions of interviews and applicants with disabilities. Rehabilitation Psychology, 40, 261–78. Macrae, C.N., Bodenhausen, G.V., & Milne, A.B. (1998). Saying no to unwanted thoughts: Self-focus and the regulation of mental life. Journal of Personality and Social Psychology, 74, 578–89. Macrae, C.N., Bodenhausen, G.V., Milne, A.B., & Jetten, J. (1994). Out of mind but back in sight: Stereotypes on the rebound. Journal of Personality and Social Psychology, 67, 808–17. Macrae, C.N., Hewstone, M., & Griffiths, R.J. (1993). Processing load and memory for stereotype-based information. European Journal of Social Psychology, 23, 77–87. Mallett, R.K., Wilson, T.D., & Gilbert, D.T. (2008). Expect the unexpected: Failure to anticipate similarities leads to an intergroup forecasting error. Journal of Personality and Social Psychology, 94, 265–77. Maner, J.K., Kenrick, D.T., Becker, D.V., Robertson, T.E., Hofer, B., Neuberg, S.L, Schaller, M. (2005). Functional Projection: How fundamental social motives can bias interpersonal perception. Journal of Personality and Social Psychology, 88, 63– 78. McCarthy, J., & Goffin, R. (2004). Measuring job interview anxiety: Beyond weak knees and sweaty palms. Personnel Psychology, 57, 607–37. McCarthy, J.M., Van Iddekinge, C.H., & Campion, M.A. (2010). Are highly structured job interviews resistant to demographic similarity effects? Personnel Psychology, 63, 325–59. McConahay, J.B. (1986). Modern racism, ambivalence, and the modern racism scale. In J.F. Dovidio, & S.L. Gaertner (Eds), Prejudice, Discrimination, and Racism (pp. 91–125). Orlando, FL: Academic Press. McKay, P.F., & Davis, J. (2008). Traditional selection methods as resistance to diversity in organizations. In K. Thomas (Ed.), Diversity Resistance in Organizations (pp. 151–74). New York: Taylor Francis Group/Lawrence Erlbaum. Mitchell, J.P., Nosek, B.A., & Banaji, M.R. (2003). Contextual variations in implicit evaluation. Journal of Experimental Psychology: General, 132, 455–69. Mitchell, G., & Tetlock, P.E. (2006). Antidiscrimination law and the perils of mindreading. Ohio State Law Journal, 67, 1023–121. Monahan, J.L., Murphy, S.T., & Zajonc, R.B. (2000). Subliminal mere exposure: Specific, general, and diffuse effects. Psychological Science, 11, 462–6.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
333
Morgeson, F.P., Reider, M.H., Campion, M.A., & Bull, R.A. (2008). Review of research on age discrimination in the employment interview. Journal of Business and Psychology, 22, 223–32. Moskowitz, G.B., Gollwitzer, P.M., Wasel, W., & Schaal, B. (1999). Preconscious control of stereotype activation through chronic egalitarian goals. Journal of Personality and Social Psychology, 77, 167–84. Motowidlo, S.J., & Van Scotter, J.R. (1994). Evidence that task performance should be distinguished from contextual performance. Journal of Applied Psychology, 79, 475–80. Murphy, S.T., Monahan, J.L., & Zajonc, R.B. (1995). Additivity of nonconscious affect: Combined effects of priming and exposure. Journal of Personality and Social Psychology, 77, 167–84. Myors, B., Lievens, F., Schollaert, E., Van Hoye, G., Cronshaw, S.F., Mladinic, A., et al. (2008). International perspectives on the legal environment for selection. Industrial and Organizational Psychology: Perspectives on Science and Practice, 1, 206–46. Neumann, R., Hulsenbeck, K., & Seibt, B. (2004). Attitudes toward people with AIDS and avoidance behavior: Automatic and reflective bases of behavior. Journal of Experimental Social Psychology, 40, 543–50. Neumark, D., Bank, R.J., & Van Nort, K.D. (1996). Sex discrimination in restaurant hiring: An audit study. Quarterly Journal of Economics, 111, 915–42. Newman, M. (2009). Man wins sex discrimination case in tribunal. Times Higher Education. http://www.timeshighereducation.co.uk/story.asp?storycode=409516 (accessed 21 June 2010). Norton, M.I., Sommers, S.R., Vandello, J.A., & Darley, J.M. (2006). Mixed motives and racial bias: The impact of legitimate and illegitimate criteria on decision making. Psychology, Public Policy, and Law, 12, 36–55. Norton, M.I., Vandello, J.A., Biga, A., & Darley, J.M. (2008). Colorblindness and diversity: Conflicting goals in decisions influenced by race. Social Cognition, 26, 102– 11. Norton, M.I., Vandello, J.A., & Darley, J.M. (2004). Casuistry and social category bias. Journal of Personality and Social Psychology, 87, 817–31. Nosek, B.A. (2005). Moderators of the relationship between implicit and explicit evaluation. Journal of Experimental Psychology: General, 134, 565–84. Nosek, B.A., & Banaji, M.R. (2001). The go/no-go association task. Social Cognition, 19, 625–66. Nosek, B.A., Greenwald, A.G., & Banaji, M.R. (2007). The Implicit Association Test at age 7: A methodological and conceptual review. In J.A. Bargh (Ed.), Automatic Processes in Social Thinking and Behavior (pp. 265-292). Psychology Press. Olson, M.A., & Fazio, R.H. (2001). Implicit attitude formation through classical conditioning. Psychological Science, 12, 413–7. Olson, M.A., & Fazio, R.H. (2006). Reducing automatically activated racial prejudice through implicit evaluative conditioning. Personality and Social Psychology Bulletin, 32, 421–33. Paladino, M.P., & Castelli, L. (2008). On the immediate consequences of intergroup categorization: Activation of approach and avoidance motor behavior toward ingroup and outgroup members. Personality and Social Psychology Bulletin, 34, 755–68. Payne, J. (2006). Pregnancy Discrimination in the Workplace. European Working Conditions Observatory. http://www.eurofound.europa.eu/ewco/2006/06/UK0606019I. htm (accessed 21 June 2010). Payne, B.K., Cheng, C.M., Govorun, O., & Stewart, B.D. (2005). An inkblot for attitudes: Affect misattribution as implicit measurement. Journal of Personality and Social Psychology, 89, 277–93.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
334 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Perry, E.L., & Bourhis, A.C. (1998). A closer look at the role of applicant age in selection decisions. Journal of Applied Social Psychology, 28, 1670–97. ˜ Petty, R.E., Fazio, R.H., & Brinol, P. (Eds). (2008). Attitudes: Insights from the New Implicit Measures. New York, NY: Psychology Press. ˜ Petty, R.E., Tormala, Z.L., Brinol, P., & Jarvis, W.B.G. (2006). Implicit ambivalence from attitude change: An exploration of the PAST model. Journal of Personality and Social Psychology, 90, 21–41. Pingitore, R., Dugoni, B.L., Tindale, R.S., & Spring, B. (1994). Bias against overweight job applicants in a simulated employment interview. Journal of Applied Psychology, 79, 909–17. Posthuma, R.A., Morgeson, F.P., & Campion, M.A. (2002). Beyond employment interview validity: A comprehensive narrative review of recent research and trends over time. Personnel Psychology, 55, 1–81. Pratto, F., & Shih, M. (2000). Social dominance orientation and group context in implicit group prejudice. Psychological Science, 11, 515–8. Pugh, M.D., & Wahrman, R. (1983). Neutralizing sexism in mixed-sex groups. Do women have to be better than men? American Journal of Sociology, 88, 746–62. Purkiss, S.L.S., Perrewe, P.L., Gillespie, T.L., Mayes, B.T., & Ferris, G.R. (2006). Implicit sources of bias in employment interview judgments and decisions. Organizational Behavior and Human Decision Processes, 101, 152–67. Raza, S.M., & Carpenter, B.N. (1987). A model of hiring decisions in real employment interviews. Journal of Applied Psychology, 72, 596–603. Reilly, N.P., Bocketti, S.P., Maser, S.A., & Wennet, C.L. (2006). Benchmarks affect perceptions of prior disability in a structured interview. Journal of Business and Psychology, 20, 489–500. Richeson, J.A., & Ambady, N. (2001). Who’s in charge? Effects of situational roles on automatic gender bias. Sex Roles, 44, 493–512. Richeson, J.A., & Ambady, N. (2003). Effects of situational power on automatic racial prejudice. Journal of Experimental Social Psychology, 39, 177–83. Richeson, J.A., Baird, A.A., Gordon, H.L., Heatherton, T.F., Wyland, C.L., Trawalter, S., & Shelton, J.N. (2003). An fMRI investigation of the impact of interracial contact on executive function. Nature Neuroscience, 6, 1323–8. Richeson, J.A., & Shelton, J.N. (2003). When prejudice does not pay: Effects of interracial contact on executive function. Psychological Science, 14, 287–90. Roberts, L.L., & Macan, T.H. (2006). Disability disclosure effects on employment interview ratings of applicants with nonvisible disabilities. Rehabilitation Psychology, 51, 239–46. Roehling, M.V. (1999). Weight-based discrimination in employment: Psychological and legal aspects. Psychological and legal aspects, 52, 969–1016. Roehling, M.V., Roehling, P.V., & Odland, L.M. (2008). Investigating the validity of stereotype about overweight employees: The relationship between body weight and normal personality traits. Group and Organization Management, 33, 392– 424. Roehling, M.V., Roehling, P.V., & Pichler, S. (2007). The relationship between body weight and perceived weight-related employment discrimination: The role of sex and race. Journal of Vocational Behavior, 71, 300–18. Rooth, D.O. (2007). Implicit discrimination in hiring. IZA Discussion Papers. http://ssrn.com/abstract=984432 Roth, P.L., Van Iddekinge, C.H., Huffcutt, A.I., Eidson, C.E. Jr., & Bobko, P. (2002). Corrections for range restrictions in structured interview ethnic group differences: The values may be larger than researchers thought. Journal of Applied Psychology, 87, 369–76.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
335
Rothblum, E.D., Miller, C.T., & Garbutt, B. (1988). Stereotypes of obese female job applicants. International Journal of Eating Disorders, 7, 277–83. Rudolph, C.W., Wells, C.A., Weller, M.D., & Baltes, B.B. (2009). A meta-analysis of empirical studies of weight-based bias in the workplace. Journal of Vocational Behavior, 74, 1–10. Rydell, R.J., & McConnell, A.R. (2006). Understanding implicit and explicit attitude change: A systems of reasoning analysis. Journal of Personality of Social Psychology, 91, 995–1008. Sacco, J.M., Scheu, C.R., Ryan, A.M., & Schmitt, N. (2003). An investigation of race and sex similarity effects in interviews: A multilevel approach to relational demography. Journal of Applied Psychology, 88, 852–65. Schaller, M., Park, J.H., & Mueller, A. (2003). Fear of the dark: Interactive effects of beliefs about danger and ambient darkness on ethnic stereotypes. Personality and Social Psychology Bulletin, 29, 637–49. Schimel, J., Simon, L., Greenberg, J., Pyszczynski, T., Solomon, S., Waxmonsky, J., & Arndt, J. (1999). Stereotypes and terror management: Evidence that mortality salience enhances stereotypic thinking and preferences. Journal of Personality and Social Psychology, 77, 905–26. Schmitt, N. (1976). Social and situational determinants of interview decisions: Implications for the employment interview. Personnel Psychology, 29, 79–101. Schuette, R.A., & Fazio, R.H. (1995). Attitude accessibility and motivation as determinants of biased processing: A test of the MODE model. Personality and Social Psychology Bulletin, 21, 704–10. Sekaquaptewa, D., & Espinoza, P. (2004). Biased processing of stereotypeincongruency is greater for low than high status groups. Journal of Experimental Social Psychology, 40, 128–35. Sekaquaptewa, D., Espinoza, P., Thompson, M., Vargas, P., & von Hippel, W. (2003). Stereotypic explanatory bias: Implicit stereotyping as a predictor of discrimination. Journal of Experimental Psychology, 39, 75–82. Shelton, J.N., & Richeson, J.A. (2005). Intergroup contact and pluralistic ignorance. Journal of Personality and Social Psychology, 88, 91–107. Sherman, J.W., & Frost, L.A. (2000). On the encoding of stereotype-relevant information under cognitive load. Personality and Social Psychology Bulletin, 26, 26– 34. Sherman, J.W., Stroessner, S.J., Conrey, F.R., & Azam, O.A. (2005). Prejudice and stereotype maintenance process: Attention, attribution, and individuation. Journal of Personality and Social Psychology, 89, 607–22. Sieverding, M. (2009). ‘Be cool!’: Emotional costs of hiding feelings in a job interview. International Journal of Selection and Assessment, 17, 391–401. Simola, S.K., Taggar, S., & Smith, G.W. (2007). The employment selection interview: Disparity among research-based recommendations, current practices and what matters to Human Rights Tribunals. Canadian Journal of Administrative Sciences, 24, 30–44. Sinclair, S., Dunn, E., & Lowery, B.S. (2005). The relationship between parental racial attitudes and children’s implicit prejudice. Journal of Experimental Social Psychology, 41, 283–9. Sinclair, L., & Kunda, Z. (1999). Reactions to a Black professional: Motivated inhibition and activation of conflicting stereotypes. Journal of Personality and Social Psychology, 77, 885–904. Singer, M.S., & Sewell, C. (1989). Applicant age and selection interview decisions: Effect of information exposure on age discrimination in personnel selection. Personnel Psychology, 42, 135–54.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
336 I NTERNATIONAL R EVIEW OF I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY Singletary, S.L., & Hebl, M.R. (2009). Compensatory strategies for reducing interpersonal discrimination: The effectiveness of acknowledgments, increased positivity, and individuating information. Journal of Applied Psychology, 94, 797–805. Stangor, C., & Duan, C. (1991). Effects of multiple task demands upon memory for information about social groups. Journal of Experimental Social Psychology, 27, 357–78. Stangor, C., & McMillan, D. (1992). Memory for expectancy-congruent and expectancy-incongruent information: A review of the social and social developmental literatures. Psychological Bulletin, 111, 42–61. Stapel, D.A., & Koomen, W. (2000). How far do we go beyond the information given? The impact of knowledge activation on interpretation and inference. Journal of Personality and Social Psychology, 78, 19–37. Steele, J.R., & Ambady, N. (2006). “Math is hard!” The effect of gender priming on women’s attitudes. Journal of Experimental Social Psychology, 42, 428–36. Stone, D.L., & Colella, A. (1996). A model of factors affecting the treatment of disabled individuals in organizations. Academy of Management Review, 21, 352–401. Straus, S.G., Miles, J.A., & Levesque, L.L. (2001). The effects of videoconference, telephone, and face-to-face media on interviewer and applicant judgments in employment interviews. Journal of Management, 27, 363–81. Swami, V., Chan, F., Wong, V., Furnham, A., & Tovee, M.J. (2008). Weight-based discrimination in occupational hiring and helping behavior. Journal of Applied Social Psychology, 38, 968–81. Swim, J., Borgida, E., Maruyama, G., & Myers, D.G. (1989). Joan McKay versus John McKay: Do gender stereotypes bias evaluations? Psychological Bulletin, 105, 409– 29. Thompson, T.L. (1982). Disclosure as a disability-management strategy: A review and conclusions. Communication Quarterly, 30, 196–203. Uhlmann, E.L., & Cohen, G.L. (2005). Constructed criteria: Redefining merit to justify discrimination. Psychological Science, 16(6), 474–80. Uhlmann, E.L., & Cohen, G.L. (2007). “I think it, therefore it’s true”: Effects of selfperceived objectivity on hiring discrimination. Organizational Behavior and Human Decision Processes, 104, 207–23. van der Zee, K.I., Bakker, A.B., & Bakker, P. (2002). Why are structured interviews so rarely used in personnel selection? Journal of Applied Psychology, 87, 176–84. van Knippenberg, A.V., Dijksterhuis, A., & Vermeulen, D. (1999). Judgement and memory of a criminal act: The effects of stereotypes and cognitive load. European Journal of Social Psychology, 29, 191–201. Vorauer, J.D., Hunter, A.J., Main, K.J., & Roy, S.A. (2000). Meta-stereotype activation: Evidence from indirect measures for specific evaluative concerns experienced by members of dominant groups in intergroup interaction. Journal of Personality and Social Psychology, 78, 690–707. Vorauer, J.D., Main, K.J., & O’Connell, G.B. (1998). How do individuals expect to be viewed by members of lower status groups? Content and implications of metastereotypes. Journal of Personality and Social Psychology, 75, 917–37. Ware, L. (2007–2008). A comparative analysis of unconscious and institutional discrimination in the United States and Britain. Georgia Journal of International and Compensation Law, 36, 89–138. Wegener, D.T., & Petty, R.E. (1995). Flexible correction processes in social judgment: The role of na¨ıve theories in corrections for perceived bias. Journal of Personality and Social Psychology, 68, 36–51. Wegner, D.M. (1994). Ironic processes of mental control. Psychological Review, 101, 34–52.
P1: OTA/XYZ P2: ABC c08 JWST046-Hodgkinson
January 21, 2011
13:5
Printer Name: Yet to Come
DISCRIMINATION IN THE EMPLOYMENT INTERVIEW PROCESS
337
Wegner, D.M. (1997). When the antidote is the poison: Ironic mental control processes. Psychological Science, 8, 148–50. Wenzlaff, R.M., & Wegner, D.M. (2000). Thought suppression. Annual Review of Psychology, 51, 59–61. Wheeler, S.C., & Petty, R.E. (2001). The effects of stereotype activation on behavior: A review of possible mechanisms. Psychological Bulletin, 127, 797–826. Wilson, T.D., Lindsey, S., & Schooler, T.Y. (2000). A model of dual attitudes. Psychological Review, 107, 101–26. Wittenbrink, B., Judd, C.M., & Park, B. (2001). Spontaneous prejudice in context: Variability in automatically activated attitudes. Journal of Personality and Social Psychology, 81, 815–27. Wittenbrink, B., & Schwarz, N. (Eds). (2007). Implicit Measures of Attitudes. New York, NY: Guilford Press. Wright, O.R. Jr. (1969). Summary of research on the selection interview since 1964. Personnel Psychology, 22, 391–413. Word, C.O., Zanna, M.P., & Cooper, J. (1974). The nonverbal mediation of selffulfilling prophecies in interracial interaction. Journal of Experimental Social Psychology, 10, 109–20. Wyer, N.A. (2004) Not all stereotypic biases are created equal: Evidence for a stereotype-disconfirming bias. Personality and Social Psychology Bulletin, 30, 706–20. Ziegert, J.C., & Hanges, P.J. (2005). Employment discrimination: The role of implicit attitudes, motivation, and a climate for racial bias. Journal of Applied Psychology, 90, 553–62.
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
INDEX affect affective states/reactions 4–5, 15, 17–18 affective well-being 4–5, 9 appraisals and 21–2 cognitive theories 15 goal based theories 15 suppression of unpleasant affect 26, 28 trait affect 7, 11 trust and 172 Affective Events Theory (AET) 17–18 affective expression 26–7, 28 affective suppression 26, 28 age discrimination 294, 298 agency well-being and see stress and well-being applicant faking see faking appraisal theories see goal-based appraisal theories arousal consumer behaviour and 69–70, 72–4 informational reinforcement and 72, 79 see also emotion associative model 106, 107–8, 109, 110 Associative-Propositional Evaluation (APE) model 304 attitudes attitude-intention-behaviour 48 behavioural consistency 48 Behavioural Perspective Model (BPM) 48, 55 consumer behaviour 48 emotional attitudes 65 faking and 246 longitudinal assessment of change 93–116 propositional attitudes 55, 57, 67, 68, 84 attribution theory trust and 160–1, 165 autonomy see job autonomy Behavioural Perspective Model (BPM) 47–9 attitudes 48, 55
behavioural consistency 48 emotional attitudes 65 propositional attitudes 55, 57, 67, 68, 84 behavioural continuity 47, 49, 58–61, 80–1 behavioural economics research 45 consumer behaviour setting (CBS) 50, 56, 58–9, 64, 69, 74, 80, 83 closed 50, 51, 53, 76, 78–9 open 50–1, 53, 75, 76, 78–9 PAD and 78–9 consumer demand 54 consumer situation 48, 50, 51–2, 56, 64, 78, 79–80, 83 contingency matrix 49, 52, 53 discriminative stimulus 57 economic choice 54 emotion-contingency links 48–9, 79–80, 83 empirical research 53–4 extensional construal 47, 54–8 extensional language 54–6 generic model 49–53 informational reinforcement 49, 52, 54, 57, 64, 79, 80 arousal and 72, 79 learning history 51–2, 56, 59, 63–4, 65, 66, 68, 72, 80–1, 82–4 operant consumption 49 operant response 57 propositional attitudes 55, 57, 67, 68, 84 stimuli 50, 51–2 utilitarian reinforcement 49, 52, 54, 56–7, 64, 79, 80 pleasure and 71, 79 see also consumer behaviour breaks from work 16, 25, 28, 30 catharsis 26 cheating 250–1 see also faking cognitive overload/over-stimulation 199–200, 202
International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
340
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
INDEX
composition models 93, 96 process composition models 96–9 conditional reasoning tests 255 consumer behaviour 47–84 arousal and 69–70, 72–4 attitude-intention-behaviour 48 BPM see Behavioural Perspective Model contingencies of reward and punishment 48 economic choice 54 emotion and 47, 48, 49, 64–80 pleasure, arousal and dominance (PAD) and 78–9 see also emotion environmental contingency 48, 49, 55, 57, 58 emotion and 67, 68, 69–70, 72, 73, 74, 77, 78, 79 intentional behaviourism and 61, 62, 64, 80, 81 extensional language 54–6 intentional behaviourism see intentional behaviourism reinforcement 49–50 control job control see job control office environment and 200, 209, 210, 212–13, 222–3 trust and 152, 159–61 coping 18, 21, 22–4 choice 24 goal-based appraisal theories 18 problem-focused 24–5, 28 coping behaviour 22, 23–4, 28 coping function and 23 coping resources and 23–4 job characteristics and 23–4, 28 job control and 23, 24, 28 social support and 23, 24 coping behaviours job design and 23–4 coping function 2, 23, 28 coping resources 22, 23–4 curve-of-factors model 110–11 Daniels, K. 13–16 Dennett, D.C. 47–8, 61–4, 67, 68, 81 detachment from work job control and 26 recovery and 25, 26 disabilities job applicants with 298–9 discrimination in the employment interview process 293–6 accountability 295, 309
applicant demographics 297–8 applicants with disabilities 298–9 future research suggestions 295 implicit bias 297, 302–3 attitudinal scales 303–4 changing 322–4 cognitive load 304–5, 315, 317, 320–1 conditioning techniques 323–4 context-based interventions 322–3 empirical evidence 304–5 implicit/automatic cognition 303–4 interview phase 310–16 motivational effects 319–20 post interview decision-making phase 316–19 pre-interview phase 307–10 prevention strategies 319–24 social desirability 304 stereotype activation 319, 320 stereotype application 319, 320 thought suppression 321 interview phase 310 assimilation 311, 312 avoidance behaviors 310–11–2 behavioral processes 295, 310–12 counter-stereotypic individuating information 313–14 future research suggestions 295 memory 314, 315–16 perceptual processes 295, 312–16 self-fulfilling prophecies 311–12 stereo-type inconsistent information/behavior 314–15 interview process model 305–19 Lesbian Gay Bisexual Transgender (LGBT) applicants 301–2 nonverbal behaviors 293, 304, 306, 307, 312 overweight and obese applicants 300–1 paralinguistic behaviors 293, 306 post interview decision-making phase 316–19 constructed criteria effects 295, 317–19 shifting standards model 295, 316–17 status characteristics model 295, 317 post-interview phase, future research suggestions 295 pre-interview phase 305 accountability 309 future research suggestions 295 intergroup forecasting error 295, 309–10 mental imagery, using 309–10 preparing to interview 309–10
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
INDEX
r´esum´e screening 295, 307–9 stereotype threat 295, 310 pregnant applicants 294, 299–300 rating scales behaviorally anchored 299, 316 descriptive anchors 316 latent response scales 316 objective scales 317 subjective scales 316, 317 research directions examining discrimination issues 302 review of research 296–302 social desirability 304 structured interviews 296, 297, 298–9, 302, 305, 306, 315 verbal behaviors 293, 304, 307, 312 working mother applicants 299–300 distrust 162, 163–4 see also trust dominance consumer behaviour and 70, 73, 74–5 see also emotion dopamine 71, 73, 74–5 downsizing 10 emotion ascription of 49, 58, 64–80, 81, 82 behavioural expression 69–70, 81, 84 biological naturalism 65–6 brain and 70–5 consumer behaviour and 47, 48, 49, 64–80 PAD and 78–9 emotion-contingency links 83 BPM basis 48–9, 79–80, 83 genetic factors 75–7 importance 75–8 PAD and consumer behaviour 78–9 emotional attitudes 65 environmental contingency and 67, 68, 69–70, 72, 73, 74, 77, 78, 79 explanation and 68–9 genetic factors 75–7 intentional behaviourism 66, 68–9 intentionality and 67–8 pleasure, arousal and dominance (PAD) 69 arousal 69–70, 72–4 biochemical basis 73 consumer behaviour and 78–9 dominance 70, 73, 74–5 dopamine 71, 73, 74–5 neurobiological basis 70–5 opioids 71, 74 pleasure 69, 71–2
341
salient facets 64–9 as subjective experience 65–7 verbal behaviour 48, 66, 67, 69, 70, 81 see also consumer behaviour employment interviews see discrimination in the employment interview process European Union (EU) 1, 6, 7 expectancy theories 244 experience sampling methods 1, 31–3 extensional construal 47, 54–8 extensional language 54–6 factor-of-curves model 108–10 faking 239–40 academic domains 248 acting 248 behaviours 249–53 capability 247–9 cheating 250–1 coaching and 249, 251 cognitions 249–53 conditional guessing 253 conditional reasoning tests 255 content tests 240, 258 correcting scores 263–4 contextual factors 247–9 countermeasures 244–7, 253–6 criterion-related validity and 259–62 cultural concerns 268 employee deception 269–70 fairness issues 268–70 faking as privacy 270 similar conceptions of faking 268 socially desirable responding (SDR) 268, 269 deception 240–2 cultural concerns 269–70 definition 240–1 detection 243–4 integrating deception theories with faking models 242–4, 245 maintaining 265–6 omissive deception 270 deterrent-based countermeasures 244–7 dishonesty and 241, 263, 265–6 distal outcomes 265–7 effective selection decisions 264–5 extrapolation 252 face-to-face interactions 247–8, 253 facial expressions 242, 248, 249, 272 factor structures and 258 forced-choice response formats 254–5 “getting away with it” 265–6 implicit associations tests (IATs) 255 indirect scoring methods 253–4
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
342
T1: IML 13:6
INDEX
faking (cont.) ipsative tests 254–5 item response theory (IRT) 272 job desirable responding (JDR) 252, 258 lie detection 243–4, 248, 255, 262–3, 264 measures 271–2 models 241–2 integrating deception theories with 242–4, 245 motivation 240, 241, 242, 244–7, 248, 254, 257, 265 attitude of peers 246 expectancy theories 244 “honesty contracts” 246 internal factors 246–7 Machiavellianism 246–7 personality 247 warnings and 246 multiple-choice tests 242, 250, 253, 254, 255 omissive deception 270 opportunities to fake 253–6 organizational reactions 262–3 organized groups 253 outcomes distal outcomes 265–7 proximal outcomes 256–65 personality and 247 practice 248 as privacy 270 process 244–56 randomization of tests 253 range-restriction 260–1 rank-ordering and 258–9 reproduction 250–1 response format manipulation 255 retests 248 rule application 251–2 selection test score shifts and 256–7 situational constraints 253–6 social domains 248 socially desirable responding (SDR) 251–2, 254, 257, 258, 264, 265, 271 cultural concerns 268, 269 statistical simulations 259, 260, 261, 264 strategies 243, 244, 249–53 technology and 255–6 trust and 266–7 variations in research methodology 270–3 warnings 246, 249, 257, 265, 273 focus groups 30 goal-based appraisal theories 16–17 acceptable rate of progress 19 actions following appraisal 21–2
Printer Name: Yet to Come
affective consequences of appraisals 21–2 Affective Events Theory (AET) 17–18 automatic processing 18, 20, 21 basis of appraisals 17–18 beliefs and 18, 19, 20, 21, 22 categorization of work events 19, 29 controlled processing 18–20, 21 coping 18 criticism 17 dual-process models 18 enacted job characteristics 17 inferences concerning goal progress 19 influence of appraisals on well-being 18–20 job characteristics and 17, 19–20, 21 job demand and 20, 21 job design and 18, 21 long-term memory and 18–19, 20, 21, 22 mental models 18–9 see also stress and well-being hot-desking 206, 209, 210, 212, 221 human resource management policies and practices trust and 155–7 implicit associations tests (IATs) 255 informational reinforcement see Behavioural Perspective Model intentional behaviourism 47–8, 49 afferent-efferent links 61, 62–3, 64, 81, 82 afferent neurons 61 ascription of emotion 49, 58, 64–80, 81, 82 see also emotion ascription of intentionality 58, 59, 62, 63, 64, 81 efferent neurons 61 emotion and 66, 68–9 environmental contingency and 61, 62, 64, 80, 81 intentional language/intentionality 47, 55, 59 intentionality and contingency 63–4 molar operant behaviour 62, 63–4 natural selection 61, 62, 64, 82 neural intentionality 61–3 personal level 61, 62, 63 see also consumer behaviour intentionality ascription of 58, 59, 62, 63, 64, 81 contingency and 63–4 emotion and 67–8 neural 61–3
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
INDEX
International Labour Organization (ILO) 1, 7 interviews see discrimination in the employment interview process item response theory (IRT) 272 job applicant faking see faking job autonomy 10, 29 affective expression and 26–7 self-reports 13 job characteristics coping behaviour and 23–4, 28 coping resources and 23 differences between individuals 10–12 enacted job characteristics 13, 14–16, 17, 23, 28, 29 goal-based appraisal theories and 17, 19–20, 21 health-related work limitation and 11–12 job autonomy 10, 13 job control 8, 15–16, 23 latent job characteristics 13, 15, 23, 24 Management Standards approach 6–7, 11, 28–9 meta-regression analyses 11 negatively oriented personality traits 11 objectivity 7–8, 12, 13 organizational interventions 9–10 perceived job characteristics 13–14 problem-focused coping and 24 questionnaires 7, 13, 31 randomized control trials 10 relationship with well-being and health 9, 12 revisioning 12–16 self-reports 7, 8, 10–11, 13, 14 social support 23, 24 stability 7, 8–9, 12 task restructuring 9–10 triangulation of results 8 see also psychosocial hazards job control 8, 15–16 affective expression and 27 coping behaviours and 23, 24, 28 psychological detachment from work and 26 recovery and 25–6, 28 work breaks 16, 25, 28, 30 job crafting 3, 14 job design 3 cognitive aspects 3 coping behaviours and 23–4 coping resources and 23 goal-based appraisal theories and 18, 21 see also job redesign
343
job interviews see discrimination in the employment interview process job performance longitudinal assessment of change 93–116 well-being and 2, 4, 25, 26 job redesign 9, 21, 29–30, 33 see also job design job satisfaction trust and 171–2 latent growth modeling (LGM) see longitudinal assessment of change leadership trust and 152–5 learning history see Behavioural Perspective Model Lesbian Gay Bisexual Transgender (LGBT) job applicants 301–2 lie detection 243–4, 248, 255, 262–3, 264 see also faking longitudinal assessment of change 93–4 alpha change 105, 111–12 beta change 111–12 between-group differences 94 composition models 93, 96 process composition models 96–9 conceptual and methodological issues 93–116 construct validity issues 93 cross-level situations 93–4 composite construct 94, 102 dynamic cross-level constructs 102–4 hierarchical linear modeling (HLM) 100 latent growth modeling 101–2 modeling cross-level effects 99–102 newcomer adaptation 100–2 person-environment (P-E) fit constructs 102–4 person-group fit 94, 102 gamma change 105, 111–12, 113 job performance and work attitudes 93–116 latent growth modeling (LGM) 95, 105, 106 associative model 106, 107–8, 109, 110 children’s social skills 107–8 cross-level situations 101–2 curve-of-factors model 110–11 factor-of-curves model 108–10 measurement invariance 97, 104, 111–13 multiple group issues 113–15 multiple indicator LGMs 106, 110–13
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
344
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
INDEX
latent growth modeling (LGM) (cont.) multivariate LGMs 106, 107–13 newcomer adaptation 101–2, 106 person-environment (P-E) fit constructs 104 levels of analysis 93 measurement invariance 97, 104, 111–13 multilevel issues 94–104 multiple group issues 113–15 multiple levels of analysis 95–9 composition models 93, 96 nested model comparisons 99 process composition models 96–9 team efficacy perceptions 97–9 multivariate issues 94, 104–13 process composition models 96–9 random fluctuations vs. systematic changes 94 reversible change 104–5 systematic change 94 Machiavellianism faking and 246–7 Management Standards for Work-Related Stress 2, 6–7, 11, 28–9, 30, 33 Mehrabian, A. 69–70, 72, 73, 74, 78, 79 multiple regression models explanatory purposes 121 job characteristics 11 multilevel models 94–5, 138 predictive purposes 120–1 relative importance of variables 119–20 beta-weights 123–4, 125, 129 bivariate correlations 123 complete dominance 127 conditional dominance 127 convergence of dominance weights and relative weights 129–30 definition of “relative importance” 122 dominance weights 126–7, 129–30, 133–4, 136 estimating 132–5 future research 137–9 general dominance 126 general dominance weights 126, 130 historical measures 123–5 incremental importance 121–2, 123 logistic dominance analysis 136 logistic relative weight analysis 136 in logistical regression 135–6 MANOVA designs 138 meaning of variable importance 121–3 models with interactions and other higher order effects 137–8 modern statistics 126–30
multilevel models 138 multivariate dominance analysis 133–4 multivariate relative weight analysis 134–5 patterns of dominance 126–7 predictor importance, estimating 132–6 product measure weights 123, 125–6 recent developments 130–6 relative weights 127–9, 130, 134–5, 136 rescaled general dominance weights 126–7 rescaled relative weights 128–9 squared beta-weights 123–4, 125, 130 squared bivariate correlations 123, 130 natural selection 61, 62, 64, 82 negatively oriented personality traits 11 neighbourhood work centers 206–7 neuroscience 63 newcomer adaptation cross-level situations 100–2 latent growth modeling (LGM) 101–2, 106 noise in offices 195, 200–1, 203, 211–12 office environment 193–4 activity magnet areas 207 break-out areas 211–12, 214–15, 218 cell offices 193, 207, 208, 209, 210, 228 combi-office 209–10 control issues 200, 209, 210, 212–13, 222–3 cost saving 197, 205, 210, 215, 226 desk space 208, 209, 214 Duffy’s hive/cell/den/club distinction 207–8 environmental satisfaction 200 evolution 203–4 drivers of change 204–6 effects 209–11 form of office 206–9 flex-office 209, 210 flexibility 197, 203, 209, 210, 211, 221 green issues 221–2, 223–4 hot-desking 206, 209, 210, 212, 221 informal meeting spaces 207 knowledge work 194, 199–200, 205, 208, 210, 223, 226 management of change process 211, 221 case studies 214–15, 218–19 new or redesigned workspace 211–12
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
INDEX
organizational change management compared 212–13 socio-technical principles 215–19, 223 user involvement in workspace design 212, 213–15, 216 meeting spaces 207 neighbourhood work centers 206–7 new opportunities 220–2 new technology and 204–5 noise 195, 200–1, 203, 211–12 open-plan see open-plan offices practical and methodological considerations 224 cross-disciplinary collaboration 227–8 moving beyond basic outcomes 226–7 physiological data 226 quasi-experimental approaches 224–5 temporal/real-time data collection 225–6 tipping point analysis 224 productivity and 193, 195 psychological needs-based approaches 208 satellite offices 206 shared offices 206–7, 209, 210, 228 social meeting spaces 207 socio-technical principles 215–19, 223 sustainable buildings 221–2, 223–4 theory development, and extensions 222–4 traditional offices 193 open-plan offices 193, 194, 195, 209, 228 benefits 196–8 cognitive overload/over-stimulation and 199–200, 202 communication and 197, 198, 207 contextual factors 201–2 cost savings 196–7 environmental satisfaction 200 financial benefits 196–7 flexibility 197, 203 historical overview 195–6 individual factors 201–2 knowledge work 194, 199–200, 205, 208, 210, 223, 226 noise 200–1, 203 privacy loss 199, 200, 201 productivity and 195 redesign, case study 214–15 refurbishment 198 risks 198–201 task complexity and 202 trade-offs perspective 202–3 working practices and 197–8
345
see also office environment opioids 71, 74 organizational downsizing 10 out-of-work activities recovery and 25 overweight and obese job applicants 300–1 performance see job performance person-environment (P-E) studies 102–4 pleasure consumer behaviour and 69, 71–2 utilitarian reinforcement and 71, 79 see also emotion pregnant job applicants 294, 299–300 presenteeism 12 problem-solving cognitive performance and 24–5 coping and 24–5, 28 well-being and 24–5, 28, 29 process composition models 96–9 team efficacy perceptions 97–9 psychological detachment from work job control and 26 recovery and 25 psychosocial hazards controlling 6 examples 5 health-related work limitation 11–12 job characteristics see job characteristics Management Standards approach 6–7, 11, 28–9 managing 5–12 presenteeism 12 questionnaires 6, 7, 31 risk assessments 6, 28 self-reports 7, 8, 10–11, 13, 14 see also stress and well-being Quine, W.V.O. 56 recovery 25–6 detachment from work and 25, 26 job control and 25–6, 28 out-of-work activities and 25 regression models see multiple regression models risk assessments 6, 28–9, 31 Rolls, E.T. 75–8, 83 Russell, J.A. 69–70, 72, 73, 74, 78, 79 satellite offices 206 self-determination theory trust and 171 self-esteem affective expression and 27
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 13:6
346 self-reports job autonomy 13 job characteristics 7, 8, 10–11, 13, 14 Skinner, B.F. 55, 64 social support coping behaviours and 23, 24 stereotyping counter-stereotypic individuating information 313–14 stereo-type inconsistent information/behavior 314–15 stereotype activation 319, 320 stereotype application 319, 320 stereotype threat 295, 310 stress and well-being 1–3 Affective Events Theory (AET) 17–18 affective expression 26–7, 28 affective states/reactions 4–5, 15, 17–18 affective well-being 4–5, 9 agency 3, 12, 22–8 affective expression 26–7, 28 as basis for intervention 28–30 coping 22–4, 28 problem-solving 24–5, 28, 29 recovery 25–6, 28 worker as active agent 22–8 as area of study 3–5 cognitive states 4 common health problems 4 coping see coping differences between individuals 10–12 experience sampling methods 1, 31–3 goal-based appraisal theories see goal-based appraisal theories health-related work limitation 11–12 interpretation as basis for intervention 28–30 worker as active interpreter 16–22 interventions implementation 30 interpretation and agency as basis for 28–30 intervention studies 9–10, 12 worker involvement 30–1 job characteristics see job characteristics job control see job control job design see job design job performance and 2, 4, 25, 26 job redesign 9, 21, 29–30, 33 national surveillance schemes 1–2 out-of-work activities and 25 physical health 5 presenteeism 12 problem-solving and 24–5, 28, 29
Printer Name: Yet to Come
INDEX
psychosocial hazards see psychosocial hazards recovery 25–6 research 2 risk assessments 6, 28–9, 31 subjective well-being 4 trust and 172 UK Management Standards for Work-Related Stress 2, 6–7, 11, 28–9, 30, 33 work hours 13, 25 sustainable buildings offices 221–2, 223–4 task restructuring 9–10 team efficacy perceptions 97–9 trait affect 7, 11 trust ability, benevolence, and integrity model 145–6 affective commitment and 172 antecedents 149 attribution theory 160–1 contextual factors and processes 152–61 cooperative behaviours 151 disposition to trust 149–51 high involvement work systems (HIWS) 156 human resource management policies and practices 155–7 individual differences 149–52 interactive justice 158 leadership 152–5 organizational antecedents 152–61 organizational control 152, 159–61 organizational justice 157–9 propensity to trust psychological contracts 156–7 attribution theory 160–1, 165 autonomous motivation and 170–1 behavioural perspective 146–7 beliefs 145–6, 148, 150 betrayal 147, 148, 150, 163, 176 calculative trust 147, 148 co-worker relationships 266–7 community exchange-based social/emotional trust 147 competence and 171 conceptual issues 177–8 consequences control and 152, 159–61 culture and 173–6 decision to trust 146 depths of relationships 147
P1: OTA/XYZ ind
P2: OTA/XYZ
JWST046-Hodgkinson
QC: IML/UKS January 21, 2011
T1: IML 13:6
Printer Name: Yet to Come
347
INDEX
discretionary behaviors and 169–70 displaced trust 176 dispositional trust 149–51 distrust 162–4 dynamic nature 162–4 economic exchange-based rational trust 147 future research 172–6 group functioning and 267 gullibility 151 identification-based trust 148, 153, 173, 174, 177 individual performance and 167–8 interactive justice 158 interpersonal trust 145–6, 147, 150, 160, 165–6, 167, 172, 173, 175 job satisfaction and 171–2 justice and 157–9 knowledge-based trust 148, 149 knowledge-sharing and innovation and 168–9 Mafiosi and 176 as manipulation device 176 measurement issues 177–8 misuse 176 motivation to act and 170–1 nature 144–5 organizational commitment and 172 in organizational contexts 143–78 organizational culture and 173–4 outcomes of employee trust 167 attitudinal impact 171–2 discretionary behaviors 169–70 knowledge-sharing and innovation 168–9 motivation to act 170–1
organizational commitment 172 performance 167–8 predictability 146, 148, 149, 156 processes 161–2 trust and distrust dynamics 162–4 trust repair 164–7 psychological contracts 156–7 relatedness and 171 relational-based trust 148 relationship-sensitive models 147 repair 164–7 self-determination theory and 171 self-reinforcing character 162–3 social equilibrium models 165–6 societal culture and 174–6 types 147–9 unit performance and 168 well-being and 172 United Kingdom Management Standards for Work-Related Stress 2, 6–7, 11, 28–9, 30, 33 utilitarian reinforcement see Behavioural Perspective Model well-being see stress and well-being work attitudes longitudinal assessment of change 93–116 work breaks 16, 25, 28, 30 work hours 13, 25 work performance see job performance working mother job applicants 299–300 workspace see office environment World Health Organization (WHO) 1, 7
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS
T1: IML
December 31, 2010
14:53
Printer Name: Yet to Come
International Review of Industrial and Organizational Psychology CONTENTS OF PREVIOUS VOLUMES VOLUME 25—2010 1. Implicit Leadership and Followership Theories: Dynamic Structures for Leadership Perceptions, Memory, and Leader–Follower Processes Sara J. Shondrick and Robert G. Lord 2. A Review of Leader–Member Exchange Research: Future Prospects and Directions Robin Martin, Olga Epitropaki, Geoff Thomas, and Anna Topakas 3. Corporate Communications Paul R. Jackson
1
35
89
4. The State of Play in Coaching Today: A Comprehensive Review of the Field Anthony M. Grant, Jonathan Passmore, Michael J. Cavanagh, and Helen M. Parker 5. Employee Selection in Times of Change Chockalingam Viswesvaran and Deniz S. Ones 6. Doing Diversity Right: An Empirically Based Approach to Effective Diversity Management Derek R. Avery and Patrick F. McKay 7. Positive Organizational Behavior at Work James Campbell Quick, Cary L. Cooper, Philip C. Gibbs, Laura M. Little and Debra L. Nelson 8. Team Cognition and Adaptability in Dynamic Settings: A Review of Pertinent Work Sjir Uitdewilligen, Mary J. Waller and Fred R.H. Zijlstra International Review of Industrial and Organizational Psychology, 2011, Volume 26. Edited by G. P. Hodgkinson and J. K. Ford. © 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd. ISBN: 978-0-470-97174-1
349
125
169
227 253
293
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
350
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
VOLUME 24—2009 1. Conceptualizing and Measuring Intuition: A Review of Recent Trends Erik Dane and Michael G. Pratt
1
2. Transfer of Training 1988–2008: An Updated Review and Agenda for Future Research Timothy T. Baldwin, J. Kevin Ford, and Brian D. Blume
41
3. Fifty Years of Psychological Contract Research: What Do We Know and What are the Main Challenges? Neil Conway and Rob B. Briner
71
4. Security in Organizations: Expanding the Frontier of Industrial–Organizational Psychology Edward G. Bitzer III, Peter Y. Chen, and Roger G. Johnston 5. Sensemaking in Virtual Teams: The Impact of Emotions and Support Tools on Team Mental Models and Team Performance Anat Rafaeli, Shy Ravid, and Arik Cheshin 6. Team Performance in Dynamic Task Environments Verlin B. Hinsz, Dana M. Wallace, and Jared L. Ladbury
131
151 183
7. Clarifying the Notion of Self-Regulation in Organizational Behavior Richard P. DeShon and Tara A. Rench
217
8. Individual Differences and Decision Making: What We Know and Where We Go From Here Susan Mohammed and Alexander Schwall
249
VOLUME 23—2008 1. The Psychology of Careers in Industrial and Organizational Settings: A Critical But Appreciative Analysis John Arnold and Laurie Cohen 2. Employee Recruitment: Current Knowledge and Directions for Future Research James A. Breaugh, Therese H. Macan, and Dana M.Grambow 3. Age and Learning in Organizations Margaret E. Beier
1
45 83
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
4. The Good, the Bad and the Ugly: Politics and Politicians at Work Jo Silvester 5. Building Better Workplaces through Individual Perspective Taking: A Fresh Look at a Fundamental Human Process Sharon K. Parker, Paul W.B. Atkins, and Carolyn M. Axtell 6. The Dawning of a New Era for Genuine Leadership Development Bruce J. Avolio and Adrian Chan 7. Health Protection and Promotion in the Workplace: A Review and Application of Value and Regulatory Focus Perspectives Lois E. Tetrick and Michael T. Ford 8. Personality as a Predictor of Work-Related Behavior and Performance: Recent Advances and Directions for Future Research Giles St. J. Burch and Neil Anderson
351
107
149
197
239
261
VOLUME 22—2007 1. Socialization in Organizational Contexts Blake E. Ashforth, David M. Sluss and Spencer H. Harrison 2. The Costs – and Benefits – of Human Resources Wayne F. Cascio 3. Strategies for Reducing Work-Family Conflict: Applying Research and Best Practices from Industrial and Organizational Psychology Debra A. Major and Jeanette N. Cleveland 4. Coping Research and Measurement in the Context of Work-Related Stress Philip Dewe and Cary L. Cooper 5. Organizational Learning Linda Argote and Gergana Todorova
1 71
111
141 193
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
352
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
6. Cultural Variations in Individual Job Performance: Implications for Industrial and Organizational Psychology in the 21st Century Rabi S. Bhagat, James R. Van Scotter, Pamela K. Steverson and Karen South Moustafa
235
7. Conducting Meaningful Research in a Fast-paced and Volatile World of Work: Challenges and Opportunities Anne Marie Ryan and Elaine D. Pulakos
265
VOLUME 21—2006 1. A Walk on the Safe Side: The Implications of Learning Theory for Developing Effective Safety and Health Training Michael J. Burke, David Holman, and Kamaljit Birdi 2. Task Analysis John Annett and Neville Stanton 3. Uncovering Workplace Interpersonal Skills: A Review, Framework, and Research Agenda Cameron Klein, Ren´ee E. DeRouin, and Eduardo Salas
1 45
79
4. Attribution Theory in Industrial and Organizational Psychology: A Review Mark J. Martinko, Scott C. Douglas, and Paul Harvey
127
5. International Management: Some Key Challenges for Industrial and Organizational Psychology Paul R. Sparrow
189
6. Women in Management: An Update on Their Progress and Persistent Challenges Karen S. Lyness and Jolie M.B. Terrazas
267
7. Advances in the Science of Performance Appraisal: Implications for Practice Gary P. Latham and Sara Mann
295
8. Qualitative Methods in Industrial and Organizational Psychology Catherine Cassell and Gillian Symon
339
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
353
VOLUME 20—2005 1. Mergers and Acquisitions: An Update and Appraisal Susan Cartwright
1
2. Social Identity in Industrial and Organizational Psychology: Concepts, Controversies, and Contributions S. Alexander Haslam and Naomi Ellemers
39
3. Personality in Industrial/Organizational Psychology: Not Much More than Cheese Jose M. Cortina and Michael J. Ingerick
119
4. Organizational Justice across Human Resource Management Decisions Stephen W. Gilliland and Layne Paddock
149
5. Contributions of Industrial/Organizational Psychology to Safety in Commercial Aircraft Don Harris and Lauren Thomas
177
6. Emotion in Organizations: A Neglected Topic in I/O Psychology, but with a Bright Future Neal M. Ashkanasy and Claire E. Ashton-James
221
7. Burnout and Health Review: Current Knowledge and Future Research Directions Arie Shirom, Samuel Melamed, Sharon Toker, Shlomo Berliner, and Itzhak Shapira
269
VOLUME 19—2004 1. Empowerment and Performance Toby D. Wall, Stephen J. Wood, and Desmond J. Leach 2. 25 Years of Team Effectiveness in Organizations: Research Themes and Emerging Needs Eduardo Salas, Kevin C. Stagl, and C. Shawn Burke 3. Creating Healthy Workplaces: The Supervisor’s Role Brad Gilbreath 4. Work Experience: A Review and Research Agenda ˜ Miguel A. Quinones
1
47 93 119
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
354
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
5. Workplace Experiences of Lesbian and Gay Employees: A Review of Current Research Brian Welle and Scott B. Button
139
6. My Job is My Castle: Identification in Organizational Contexts Rolf van Dick
171
7. Virtual Teams: Collaborating across Distance Carolyn M. Axtell, Steven J. Fleck, and Nick Turner
205
8. Learning at Work: Training and Development Sabine Sonnentag, Cornelia Niessen, and Sandra Ohly
249
VOLUME 18—2003 1. Flexible Working Arrangements: Implementation, Outcomes, and Management Suzan Lewis 2. Economic Psychology ¨ Erich Kirchler and Erik Holzl 3. Sleepiness in the Workplace: Causes, Consequences, and Countermeasures Autumn D. Krauss, Peter Y. Chen, Sarah DeArmond, and Bill Moorcroft 4. Research on Internet Recruiting and Testing: Current Status and Future Directions Filip Lievens and Michael M. Harris 5. Workaholism: A Review of Theory, Research, and Future Directions Lynley H.W. McMillan, Michael P. O’Driscoll, and Ronald J. Burke 6. Ethnic Group Differences and Measuring Cognitive Ability Helen Baron, Tamsin Martin, Ashley Proud, Kirsty Weston, and Chris Elshaw 7. Implicit Knowledge and Experience in Work and Organizations ¨ Andr´e Bussing and Britta Herbig
1 29
81
131
167
191
239
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
355
VOLUME 17—2002 Coping with Job Loss: A Life-facet Perspective, Frances M. McKee-Ryan and Angelo J. Kinicki; The Older Worker in Organizational Context: Beyond the Individual, James L. Farr and Erika L. Ringseis; Employment Relationships from the Employer’s Perspective: Current Research and Future Directions, Anne Tsui and Duanxu Wang; Great Minds Don’t Think Alike? Person-level Predictors of Innovation at Work, Fiona Patterson; Past, Present and Future of Crosscultural Studies in Industrial and Organizational Psychology, Sharon Glazer; Executive Health: Building Self-reliance for Challenging Times, Jonathan D. Quick, Cary L. Cooper, Joanne H. Gavin, and James Campbell Quick; The Influence of Values in Organizations: Linking Values and Outcomes at Multiple Levels of Analysis, Naomi I. Maierhofer, Boris Kabanoff, and Mark A. Griffin; New Research Perspectives and Implicit Managerial Competency Modeling in China, ZhongMing Wang
VOLUME 16—2001 Age and Work Behaviour: Physical Attributes, Cognitive Abilities, Knowledge, Personality Traits and Motives, Warr; Organizational Attraction and Job Choice, Highouse and Hoffman; The Psychology of Strategic Management: Diversity and Cognition Revisited, Hodgkinson; Vacations and Other Respites: Studying Stress on and off the Job, Eden; Cross-cultural Industrial/ Organisational Psychology, Smith, Fischer, and Sale; International Uses of Selection Methods, Newell and Tansley; Domestic and International Relocation for Work, Feldman; Understanding the Assessment Centre Process: Where Are We Now?, Lievens and Klimoski
VOLUME 15—2000 Psychological Contracts: Employee Relations for the Twenty-first Century?, Millward and Brewerton; Impacts of Telework on Individuals, Organizations and Families—A Critical Review, Kondradt, Schmook, and M¨alecke; Psychological Approaches to Entrepreneurial Success: A General Model and an Overview of Findings, Rauch and Frese; Conceptual and Empirical Gaps in Research on Individual Adaptation at Work, Chan; Understanding Acts of Betrayal: Implications for Industrial and Organizational Psychology, Pearce and Henderson; Working Time, Health and Performance, Spurgeon and Cooper; Expertise at Work: Experience and Excellent Performance, Sonnentag; A Rich and Rigorous Examination of Applied Behavior Analysis Research in the World of Work, Komaki, Coombs, Redding, Jr, and Schepman
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
356
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
VOLUME 14—1999 Personnel Selection Methods, Salgado; System Safety—An Emerging Field for I/O Psychology, Fahlbruch and Wilpert; Work Control and Employee Well-being: A Decade Review, Terry and Jimmieson; Multi-source Feedback Systems: A Research Perspective, Fletcher and Baldry; Workplace Bullying, Hoel, Rayner, and Cooper; Work Performance: A Multiple Regulation Perspective, Roe; A New Kind of Performance for Industrial and Organizational Psychology: Recent Contributions to the Study of Organizational Citizenship Behavior, Organ and Paine; Conflict and Performance in Groups and Organizations, de Dreu, Harinck, and van Vianen
VOLUME 13—1998 Team Effectiveness in Organizations, West, Borrill, and Unsworth; Turnover, Maertz and Campion; Learning Strategies and Occupational Training, Warr and Allan; Meta-analysis, Fried and Ager; General Cognitive Ability and Occupational Performance, Ree and Carretta; Consequences of Alternative Work Schedules, Daus, Sanders, and Campbell; Organizational Men: Masculinity and Its Discontents, Burke and Nelson; Women’s Careers and Occupational Stress, Langan-Fox; Computer-Aided Technology and Work: Moving the Field Forward, Majchrzak and Borys
VOLUME 12—1997 The Psychology of Careers in Organizations, Arnold; Managerial Career Advancement, Tharenou; Work Adjustment: Extension of the Theoretical Framework, Tziner and Meir; Contemporary Research on Absence from Work: Correlates, Causes and Consequences, Johns; Organizational Commitment, Meyer; The Explanation of Consumer Behaviour: From Social Cognition to Environmental Control, Foxall; Drug and Alcohol Programs in the Workplace: A Review of Recent Literature, Harris and Trusty; Progress in Organizational Justice: Tunneling through the Maze, Cropanzano and Greenberg; Genetic Influence on Mental Abilities, Personality, Vocational Interests and Work Attitudes, Bouchard
VOLUME 11—1996 Self-esteem and Work, Locke, McClear, and Knight; Job Design, Oldham; Fairness in the Assessment Centre, Baron and Janman; Subgroup Differences Associated with Different Measures of Some Common Job-relevant
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
357
Constructs, Schmitt, Clause and Pulakos; Common Practices in Structural Equation Modeling, Kelloway; Contextualism in Context, Payne; Employee Involvement, Cotton; Part-time Employment, Barling and Gallagher; The Interface between Job and Off-job Roles: Enhancement and Conflict, O’Driscoll
VOLUME 10—1995 The Application of Cognitive Constructs and Principles to the Instructional Systems Model of Training: Implications for Needs Assessment, Design, and Transfer, Ford and Kraiger; Determinants of Human Performance in Organizational Settings, Smith; Personality and Industrial/Organizational Psychology, Schneider and Hough; Managing Diversity: New Broom or Old Hat?, Kandola; Unemployment: Its Psychological Costs, Winefield; VDUs in the Workplace: Psychological Health Implications, Bramwell and Cooper; The Organizational Implications of Teleworking, Chapman, Sheehy, Heywood, Dooley, and Collins; The Nature and Effects of Method Variance in Organizational Research, Spector and Brannick; Developments in Eastern Europe and Work and Organizational Psychology, Roe
VOLUME 9—1994 Psychosocial Factors and the Physical Environment: Inter-relations in the Workplace, Evans, Johansson, and Carrere; Computer-based Assessment, Bartram; Applications of Meta-Analysis: 1987–1992, Tett, Meyer, and Roese; The Psychology of Strikes, Bluen; The Psychology of Strategic Management: Emerging Themes of Diversity and Cognition, Sparrow; Industrial and Organizational Psychology in Russia: The Concept of Human Functional States and Applied Stress Research, Leonova; The Prevention of Violence at Work: Application of a Cognitive Behavioural Theory, Cox and Leather; The Psychology of Mergers and Acquisitions, Hogan and Overmyer-Day; Recent Developments in Applied Creativity, Kabanoff and Rossiter
VOLUME 8—1993 Innovation in Organizations, Anderson and King; Management Development, Baldwin and Padgett; The Increasing Importance of Performance Appraisals to Employee Effectiveness in Organizational Settings in North America, Latham, Skarlicki, Irvine, and Siegel; Measurement Issues in Industrial and Organizational Psychology, Hesketh; Medical and Physiological Aspects of Job Interventions, Theorell; Goal Orientation and Action Control Theory, Farr, Hofmann, and Ringenbach; Corporate Culture, Furnham and Gunter; Organizational
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
358
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
Downsizing: Strategies, Interventions, and Research Implications, Kozlowski, Chao, Smith, and Hedlund; Group Processes in Organizations, Argote and McGrath
VOLUME 7—1992 Work Motivation, Kanfer; Selection Methods, Smith and George; Research Design in Industrial and Organizational Psychology, Schaubroeck and Kuehn; A Consideration of the Validity and Meaning of Self-report Measures of Job Conditions, Spector; Emotions in Work and Achievement, Pekrun and Frese; The Psychology of Industrial Relations, Hartley; Women in Management, Burke and McKeen; Use of Background Data in Organizational Decisions, Stokes and Reddy; Job Transfer, Brett, Stroh, and Reilly; Shopfloor Work Organization and Advanced Manufacturing Technology, Wall and Davids
VOLUME 6—1991 Recent Developments in Industrial and Organizational Psychology in People’s Republic of China, Wang; Mediated Communications and New Organizational Forms, Andriessen; Performance Measurement, Ilgen and Schneider; Ergonomics, Megaw; Ageing and Work, Davies, Matthews, and Wong; Methodological Issues in Personnel Selection Research, Schuler and Guldin; Mental Health Counseling in Industry, Swanson and Murphy; Person–Job Fit, Edwards; Job Satisfaction, Arvey, Carter, and Buerkley
VOLUME 5—1990 Laboratory vs. Field Research in Industrial and Organizational Psychology, Dipboye; Managerial Delegation, Hackman and Dunphy; Cross-cultural Issues in Organizational Psychology, Bhagat, Kedia, Crawford, and Kaplan; Decision Making in Organizations, Koopman and Pool; Ethics in the Workplace, Freeman; Feedback Systems in Organizations, Algera; Linking Environmental and Industrial/Organizational Psychology, Ornstein; Cognitive Illusions and Personnel Management Decisions, Brodt; Vocational Guidance, Taylor and Giannantonio
VOLUME 4—1989 Selection Interviewing, Keenan; Burnout in Work Organizations, Shirom; Cognitive Processes in Industrial and Organizational Psychology, Lord and Maher;
P1: IML/IKJ bm
P2: IML/UKS
JWST046-Hodgkinson
QC: IML/UKS December 31, 2010
T1: IML 14:53
Printer Name: Yet to Come
C ONTENTS OF P REVIOUS V OLUMES
359
Cognitive Style and Complexity, Streufert and Nogami; Coaching and Practice Effects in Personnel Selection, Sackett, Burris, and Ryan; Retirement, Talaga and Beehr; Quality Circles, Van Fleet and Griffin; Control in the Workplace, Ganster and Fusilier; Job Analysis, Spector, Brannick, and Coovert; Japanese Management, Smith and Misumi; Casual Modelling in Organizational Research, James and James
VOLUME 3—1988 The Significance of Race and Ethnicity for Understanding Organizational Behavior, Alderfer and Thomas; Training and Development inWork Organizations, Goldstein and Gessner; Leadership Theory and Research, Fiedler and House; Theory Building in Industrial and Organizational Psychology, Webster and Starbuck; The Construction of Climate in Organizational Research, Rousseau; Approaches to Managerial Selection, Robertson and Iles; Psychological Measurement, Murphy; Careers, Driver; Health Promotion at Work, Matteson and Ivancevich; Recent Developments in the Study of Personality and Organizational Behavior, Adler and Weiss
VOLUME 2—1987 Organization Theory, Bedeian; Behavioural Approaches to Organizations, Luthans and Martinko; Job and Work Design, Wall and Martin; Human Interfaces with Advanced Manufacturing Systems, Wilson and Rutherford; Human– Computer Interaction in the Office, Frese; Occupational Stress and Health, Mackay and Cooper; Industrial Accidents, Sheehy and Chapman; Interpersonal Conflicts in Organizations, Greenhalgh; Work and Family, Burke and Greenglass; Applications of Meta-analysis, Hunter and Rothstein Hirsh
VOLUME 1—1986 Work Motivation Theories, Locke and Henne; Personnel Selection Methods, Muchinsky; Personnel Selection and Equal Employment Opportunity, Schmit and Noe; Job Performance and Appraisal, Latham; Job Satisfaction and Organizational Commitment, Griffin and Bateman; Quality of Worklife and Employee Involvement, Mohrman, Ledford, Lawler, and Mohrman; Women at Work, Gutek, Larwood, and Stromberg; Being Unemployed, Fryer and Payne; Organization Analysis and Praxis, Golembiewski; Research Methods in Industrial and Organizational Psychology, Stone