Affective and Emotional Aspects of Human-Computer Interaction: Game- and Innovative Learning Approaches: Volume 1 Future of Learning

AFFECTIVE AND EMOTIONAL ASPECTS OF HUMAN-COMPUTER INTERACTION

The Future of Learning Learning is becoming more and more important as one of the indispensable tools to ensure future prosperity and well-being. This is the case not only for the individual, alone or as a member of a group, but also for organisational structures of all kinds. New learning paradigms and pedagogic principles, new learning environments and conditions, and new learning technologies are being tested in order to find the right combination of parameters that can optimise the outcome of the learning process in a given situation. This book series presents to all stakeholders the latest advances in this important area, based on a sound foundation. Schools, higher education, industrial companies, public administrations and other organisational structures, including providers of learning and training services, including life-long learning, plus all the individuals involved, researchers, students, pupils, citizens, teachers, professors, instructors, politicians, decision makers etc., contribute to and benefit from this series. Pedagogic, economics, structural and organisational aspects, the latest technologies, and the influence from changing attitudes and globalisation are treated in this series, providing sound and updated information, which can be used to further improve the learning process in both formal and informal contexts. Series Editors:

N. Balacheff, J. Breuker, P. Brna, K.-E. Chang, J.C. Cherniavsky, J.P. Christensen, M. Gattis, M. Gutiérrez-Díaz, P. Kommers, C.J. Oliveira, M. Schlager, M. Selinger, L. Steels and G. White

Volume 1 Related publications by IOS Press: M. Tokoro and L. Steels (Eds.), The Future of Learning: Issues and Prospects M. Tokoro and L. Steels (Eds.), A Learning Zone of One’s Own: Sharing Representations and Flow in Collaborative Learning Environments P. Kommers (Ed.), Cognitive Support for Learning: Imagining the Unknown

ISSN 1572-4794

Affective and Emotional Aspects of Human-Computer Interaction Game-Based and Innovative Learning Approaches

Edited by

Maja Pivec FH JOANNEUM Gesellschaft mbH, University of Applied Sciences, Department of Information Design, Graz, Austria

Amsterdam • Berlin • Oxford • Tokyo • Washington, DC

© 2006 The authors. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior written permission from the publisher. ISBN 1-58603-572-X Library of Congress Control Number: 2006920170 Publisher IOS Press Nieuwe Hemweg 6B 1013 BG Amsterdam Netherlands fax: +31 20 687 0019 e-mail: [email protected] Distributor in the UK and Ireland Gazelle Books Falcon House Queen Square Lancaster LA1 1RN United Kingdom fax: +44 1524 63232

Distributor in the USA and Canada IOS Press, Inc. 4502 Rachael Manor Drive Fairfax, VA 22032 USA fax: +1 703 323 3668 e-mail: [email protected]

LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

v

Foreword The aim of the interdisciplinary workshop “Affective and Emotional Aspects of Human-Computer Interaction: Emphasis on Game-Based and Innovative Learning Approaches” is to define new research directions related to affective and emotional approaches to computer-supported learning and human-computer interactions. The workshop is a unique opportunity to bring together, on the one hand, scientists and research contributions from psychology, educational sciences, cognitive sciences, various aspects of communication and human computer interaction, interface design and computer science and, on the other hand, educators and the game industry. The intensive exchange of information from various research fields should open the gates for evolutionary change and new research directions in technology-supported learning. Part of the research presented at the ESF workshop is published in this book. For the purpose of this book publication, an additional call for chapters was issued to the international research community, so as to obtain contributions worldwide that reflect the current research in this field. Gathered and reviewed, eighteen selected contributions are grouped into three topics: Game-based Learning, Motivation and Learning, and Emotions and Emotional Agents. To decide how to categorize the contributions was a very difficult task, hence all the contributions explore the learning process as an emotional and personal experience that is addictive and motivates the learner to proactive behavior. The major topics such as emotions, motivation, games and game-experience, though in different equivalents and various priorities, are present in all of the contributions, thus offering a variety of possible solutions for contribution classification. However, I leave the challenge to organize the contributions differently for the next publication. Game-Based Learning Jon Sykes, in his contribution Affective Gaming: Advancing the Argument for GameBased Learning, reflects upon the two-way interaction between game and student, thus enabling the game to react to the student’s emotional state. Having the possibility to detect and steer the emotional state of the student could have a positive impact on using digital games in education. In the chapter Didactic Analysis of Digital Games and Game-Based Learning, the author Matthias Bopp leads us through different games, starting from Pong by Atari to the latest developments in immersive 3D games. He regards these games as virtual environments that ‘teach’ the player to perform entertaining actions and proposes a schema to analyse the didactic methods of virtual games, featuring situational, temporal and social aspects. Furthermore, Bopp analyses the didactic methods in popular educational games. Concluding, the author asks if some of the didactic methods used in entertaining games may be used to improve educational games and points at some challenges of such an attempt. Recent research carried out at the UNITEC, University of Technology New Zealand, indicates that some commercial computer games increase cognitive skills and

vi

may enhance the participants’ ability to learn. Reported research results in the chapter Immersive Environments: What Can We Learn from Commercial Computer Games? by Paul R. Kearney, suggest that the immersive environment of Counter-Strike does in fact enhance multitasking abilities, but this may be in response to the immersive environment that is created rather than the game content itself. The ability of learners to multitask will require a rethink of traditional learning styles, argues the author. “Why do I identify with a yellow circular shape on my computer screen and feel that shape being a part of me when I play a game of Pac-Man? Why do I engage in the game play process to such a degree that when this yellow thing gets crammed into a corner with no way to run and no way to hide, I feel severe distress and call for help loud and clear and wake my kids in the middle of the night? ...” In his chapter What Is a Game Ego?, Ulf Wilhelmsson strives to find answers to these and similar questions by proposing a framework for understanding computer games from diverse fields of research: film theory (including theories on narration and narratives), theories on visual perception (which are also applicable to sound) and experientialist cognitive theory. Hakan Tüzün was inspired by the extreme interest and motivation of children to play the Quest Atlantis game (participating in the pure game activities as well as educational tasks with the same intensity). In the chapter Multiple Motivations Framework, he proposes an organizing framework from which to explain things of significance for motivating learners. His research study is based on the qualitative methods, and the research outcomes provide a very different perspective than what is available in understating motivation. The proposed framework is based on multiple elements that contribute to one’s motivation and that collectively constitute the activity of motivation, i.e. Duality of Subject, Duality of Activity, Duality of Outcome, Duality of Object, and Context of Support. In the chapter An Instructional Design/Development Model for the Creation of Game-Like Learning Environments: The FIDGE Model, the authors Göknur Kaplan Akilli and Kürsat Cagiltay tackle issues of the lack of available comprehensive design paradigms and well-designed research studies on the question of “how to” incorporate games into learning environments, that are experienced despite more than thirty years’ existence of computer games and simulations in the instructional design movement. Based on the formative research study results and with the inspiration from fuzzy logic, the authors propose an instructional design/development model for creating game-like environments, called the “FIDGE model”. “FIDGE” stands for “Fuzzified Instructional Design Development of Game-like Environments” for learning. Learning when Using Commercial Computer Games as Simulations: A Case Study Using a Simulation Game is a chapter by Preston P. Parker, where the author looks at using an off-the-shelf commercial computer game, Age of Empires II, as a simulation to facilitate learning Multimedia Production Management and explores the possibilities of implementing a structure—mapping elements of the game to elements in real-life. This study shows that it is possible that learning objectives can be reached when using a commercial computer game with some obvious mappings, but mostly analogous mappings, to reality. The described case also urges further pursuit of studies on how to use off-the-shelf commercial computer games and how to build an intervention around the game that can facilitate achieving a specific learning objective. Igor Mayer and Geertje Bekebrede review the use and usefulness of digital games and simulations for (e-) learning, training and decision & policy support of technological infrastructures, such as ports, container terminals, off-shore wind farms etc. In their chapter Serious Games and ‘Simulation Based E-Learning’ for Infrastructure Man-

vii

agement, several examples of such applications are presented. Three cases that bear relevance to infrastructure are discussed in more detail as follows: CONTAINERS ADRIFT is a computer-supported simulation-game revolving around the planning and design of an inland container terminal, VENTUM ON LINE is a multi-user on-line role playing game that is concerned with the planning and design of an off shore wind farm, SIM MV2 is an animated and network based simulation game commissioned by the Port of Rotterdam to explore and support the planning and design of its second harbour area (2nd Maasvlakte). Motivation and Learning In their contribution Learning and Motivation with Virtual Tutors, authors Manuela Paechter and Karin Schweizer report on their research concerned with social processes in the virtual classroom. “How important is information about a tutor or lecturer in an online seminar?”, “Is social information about the tutor’s appearance or his or her voice important for learning?” and similar questions concerning the role of a tutor and the form of communication between a tutor and students, were investigated in a university seminar. It was analyzed whether the absence or presence of social and personal cues in the communication between a tutor and his or her students influence students’ learning and their satisfaction with the tutor and the course. The research showed that not all types of personal information are equally important and possibly pictorial information is more important than audible information. At this point the authors conclude that further research needs to be carried out related to these questions. The aim of the pilot study presented in the chapter Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning was to investigate whether the sound and empirically valid knowledge space theory is able to cover learning and performance in two different motivational states, which were hope for success and fear of failure. These motivational states were combined with two different learning conditions e.g. pre-structured learning sessions within an adaptive tutorial system contrasting to rather free text-based learning. The data collected with 104 high school students in the domain of elementary probability theory indicate that knowledge space theory is able to represent the responses obtained in a post-test for both motivational states, as well as for both learning conditions. The authors Jürgen Heller, Dietrich Albert, Michael Kickmeier-Rust and Markus Kertz conclude that the results of the presented pilot study strongly encourage the research program to integrate cognitive and emotional/motivational aspects into a comprehensive psychological model for adaptive tutorial systems. An Interactive Dictionary of Concepts: An Exploratory Platform for Enhancing Communication Between the Concepts by Ania Lian, offers a description of a noncommercial, web-based communication platform. The platform was designed in order to facilitate and support a negotiation process between various interest groups and, as a result, between the concepts in which the interests of these groups are embedded. Unlike standard dictionaries, the Interactive Dictionary of Concepts does not come with ready-made entries. Instead, it enables individuals who are investigating various concepts and issues to create their own entries, which are then situated and organised within the constraints of the management structure of the Dictionary. These creative constraints themselves function as tools for stimulating and generating the critical reflection of authors upon the concepts which they are investigating. This is achieved by

viii

offering conditions which challenge, enrich and help to systematise the associations which inform the ideas of authors and their beliefs. The project outlined in the chapter Human-Computer Interaction: Sharing of Intergenerational and Cross-Cultural Knowledge is designed to investigate attitudinal changes amongst young children, undergraduate students, and senior citizens, in both synchronous and asynchronous learning environments when engaged in Web-mediated collaborative knowledge sharing activities. The global eMuseum System (GEMS) research project extends the principles of a Generative Virtual Classroom and Schank’s Sickle Cell Counselor, by combining social interaction through linguistics in a communicative collaborative model. The project team manages the complexity of dealing with diverse subjects by adopting three distinct experiential environments. The first involves senior citizens recalling traditional stories and games. The second defines undergraduate student interaction in a multi-cultural setting, while the third relates to young children’s propensity for playful sharing of experiential learning materials. The tracking of the collaborative nature of the involvement with the eMuseum will enhance the knowledge of interaction between generations, creating a capacity for emergent innovative global legends, concludes Elspeth McKay. Apart from providing the right functionality (being useful) and giving access to it (being easy to use), interactive products also provide hedonic qualities when being applied. The hedonic quality has a stimulation and an identity aspect. The central question of the study outlined by Michael Burmester and Annely Dufner in the chapter Designing the Stimulation Aspect of Hedonic Quality – An Exploratory Study is which type of product features support the stimulation aspect of hedonic quality and whether this quality can be systematically increased in the design process. This exploratory study had the goal to explore types of feature ideas, design principles, and methods in order to design products showing a high degree of the stimulation aspect of the perceived hedonic quality. Based on the combination of theoretic foundations and tacit knowledge of design experts with user centred evaluation of the outcome of design work, successful strategies to design attractive products can be derived. Emotions and Emotional Agents In the chapter On the Role of Self Esteem, Empathy and Narrative in the Development of Intelligent Learning Environments, Paul Brna discusses the production of learning environments which enhance the learner’s self esteem, ensure that the learner’s best interests are respected through paying attention to the narrative structure of the learner’s experience, and the ways in which communication can be enhanced through empathy with the learner. The author outlines experiences of using narrative in learning environments and observations made through the examples of the NIMIS project and the T’rrific Tales software, that was designed to both scaffold the development of narrative skills and to actively guide learners. The evidence of the benefits of empathic narrative-based design for explicitly supporting the development of the learner’s self esteem is not yet available even if the indications are favourable. The role of empathy in the construction of synthetic characters to interact with learners in intelligent learning environments is the main focus of the chapter Empathic Characters in Computer-Based Personal and Social Education written by João Dias, Ana Paiva, Marco Vala, Ruth Aylett, Sarah Woods, Carsten Zoll and Lynne Hall. The authors outline an example of an interactive learning virtual environment called Fear-

ix

Not!, that uses synthetic characters and role playing, developed as a set of bullying situations, which emerge from the actions and interactions between synthetic characters in a 3D virtual world. The system was designed to evoke affective responses by the users, in this case children, and has been evaluated with 345 children in June 2004. The evaluation results show that empathic interactions were achieved with synthetic characters. The effectiveness of intelligent tutoring systems, for instance on-line learning systems, can be improved when the learner’s emotions are taken into account. A necessary condition for this is that the system will be able to recognize the learner’s current emotional state. The authors Mohammed A. Razek, Soumaya Chaffar, Claude Frasson and Magalie Ochs present an extremely simple method that can be used for determining emotional state, the Emotion Recognition Agent (ERA), which is devoted to exploit the natural relation between emotions and colors. By giving a sequence of three colors, a person can express his/her emotion reliably. Based on the ordered choice of colors the ERA system determines someone’s emotion with 57.6 % accuracy. “In fact, the recognition of somebody’s emotions is known to be one of the central features of, and even a necessary condition for, Emotional Intelligence. Therefore, our system can be seen as a first step towards realizing online tutoring systems that are emotionally intelligent and can use this ability for the sake of improved learning efficiency.” is the conclusion of the authors in their chapter Using Machine-Learning Techniques to Recognize Emotions for On-Line Learning Systems. In their paper A Framework for Emotional Agents as Tutoring Entities, the authors Bogdan Florin Marin, Axel Hunger and Stefan Werner show how they integrate agent technology to support collaborative learning in distributed environments. The aim of their research is to provide the first steps in defining a method for creating a believable tutor agent which can partially replace human teachers and assist the students in the process of learning. The authors postulate that the application of emotional animated agents in online learning environments as a tutoring paradigm can be beneficial and increase the learners’ motivation. The prototype agent is able to display emotions by means of synthetic speech, facial display and gestures. Verbal and non-verbal behavior is synthesized in the agent’s mental model and interpreted in a learning-session. In the paper the authors also discuss the premises under which emotional agents can be pedagogically effective as tutors in a collaborative learning environment. Cyrus F. Nourani addresses affective computing with a new haptic computing logic. The author argues that “if there is a Gestalt model for the world decided on, the answer might be affirmative. From our published perspective we are where the objects are described with languages as Frege intended, modeled by structures, which can be examined by Kant’s transcendental Idealism, and their computability and reducibility areas Hilbert arithmetized. Hence there is a systematic basis to carry out conceptobject descriptions for machine discovery, a premise to an illusion logic is developed ”. In his paper A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees, the author discusses questions such as ‘are intelligent decisions based on emotions?’ He also suggests other related issues that influence creativity, planning, perception, and mood-congruent memory retrieval, with precise computing and cognitive models. These “foundations are applied to present a brief on Computational Illusion, affective computing, and virtual reality.”

x

Acknowledgements The ESF Standing Committee for the Social Sciences (SCSS) approved the workshop proposal (Ref.: EW03-210 SCSS) thus creating a unique and challenging opportunity to explore new research possibilities in the field of affective and emotional ways of human-computer interaction, with emphasis on application of games for learning. An international and interdisciplinary team offered interesting discussions and vivid exchanges of opinions and practices. The European Science Foundation (www.esf.org) financially supported the ESF SCSS Exploratory Workshop on Affective and Emotional Aspects of HumanComputer Interaction: Emphasis on Game-Based and Innovative Learning Approaches. The publication of this book is funded by the ESF. Thanks to Paul Brna, the idea of this book materialized into the publication that you hold in your hands, by giving me advice about the publisher and by establishing the first contact. Michaela Bernreiter, Ania and Andrew Lian, and Paul R. Kearney kindly offered their help to make my English sound proper. For all the support received I am much obliged – thank you, my friends! The ESF workshop was co-organized by Frank Thissen, Ph.D., University of Applied Sciences Stuttgart, Germany and by Konrad Baumann, Ph.D., FH JOANNEUM, University of Applied Sciences, Graz, Austria. Michaela Bernreiter, FH JOANNEUM, University of Applied Sciences, Graz, Austria, took the burden of all the administrative issues related to the workshop organization and financial issues. Many thanks to Department of Information Design at the FH JOANNEUM, University of Applied Sciences, Graz, Austria, especially Karl Stocker, for supporting my work.

xi

Supplementary Material Available via the Internet Information About the ESF Workshop The Scientific report, ESF workshop presentation slides and full list of workshop participants is available at: http://informations-design.fh-joanneum.at/images/esf_workshop_092004/. Resources About Game Based Learning SIG-GLUE (www.sig-glue.net) stands for Special Interest Group for Game-based Learning in Universities and Lifelong Learning. Visitors can register with the community and join the discussions, and exchange experiences with others. Furthermore, community members can share information and best practice via the newsletter and let others know about interesting new events, books, games and game platforms. SIG-GLUE: Special Interest Group for Game-based Learning in Universities and Lifelong Learning is an EC eLearning initiative Project (Agreement No: 2003-4704/ 001-001 EDU ELEARN).

This page intentionally left blank

xiii

Contents Foreword

v

Acknowledgements

x

Supplementary Material Available via the Internet

xi

Game-Based Learning Affective Gaming: Advancing the Argument for Game-Based Learning Jonathan Sykes

3

Didactic Analysis of Digital Games and Game-Based Learning Matthias Bopp

8

Immersive Environments: What Can We Learn from Commercial Computer Games? Paul R. Kearney

38

What Is a Game Ego? (or How the Embodied Mind Plays a Role in Computer Game Environments) Ulf Wilhelmsson

45

Multiple Motivations Framework Hakan Tüzün An Instructional Design/Development Model for the Creation of Game-Like Learning Environments: The FIDGE Model Göknur Kaplan Akilli and Kürsat Cagiltay

59

93

Learning when Using Commercial Computer Games as Simulations: A Case Study Using a Simulation Game Preston P. Parker

113

Serious Games and ‘Simulation Based E-Learning’ for Infrastructure Management Igor Mayer and Geertje Bekebrede

136

Motivation and Learning Learning and Motivation with Virtual Tutors. Does It Matter if the Tutor Is Visible on the Net? Manuela Paechter and Karin Schweizer Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning Jürgen Heller, Dietrich Albert, Michael Kickmeier-Rust and Markus Kertz

155

165

xiv

An Interactive Dictionary of Concepts: An Exploratory Platform for Enhancing Communication Between the Concepts Which Form and Inform Us Ania Lian

178

Human-Computer Interaction: Sharing of Intergenerational Wisdom and Cross-Cultural Knowledge Elspeth McKay

207

Designing the Stimulation Aspect of Hedonic Quality – An Exploratory Study Michael Burmester and Annely Dufner

217

Emotions and Emotional Agents On the Role of Self Esteem, Empathy and Narrative in the Development of Intelligent Learning Environments Paul Brna Empathic Characters in Computer-Based Personal and Social Education João Dias, Ana Paiva, Marco Vala, Ruth Aylett, Sarah Woods, Carsten Zoll and Lynne Hall Using Machine-Learning Techniques to Recognize Emotions for On-Line Learning Systems Mohammed A. Razek, Soumaya Chaffar, Claude Frasson and Magalie Ochs

237 246

255

A Framework for Emotional Agents as Tutoring Entities Bogdan Florin Marin, Axel Hunger and Stefan Werner

266

A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees Cyrus F. Nourani

286

ESF SCSS Exploratory Workshop on Affective and Emotional Aspects of Human-Computer Interaction: Emphasis on Game-Based and Innovative Learning Approaches

312

Author Index

317

Game-Based Learning

This page intentionally left blank

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

3

Affective Gaming: Advancing the Argument for Game-Based Learning Jonathan SYKES eMotion Laboratory, Glasgow Caledonian University, Glasgow, UK [email protected] Abstract. The argument for game-based learning (GBL) has stagnated. The justification for introducing games into the classroom has not developed in over 30 years. However, the introduction of digital game-based learning is not merely a change of gaming platform, but a significant advancement in learning technology. The two-way interaction between game and student provides unique paralinguistic communication, allowing the game to potentially react to a player’s emotional state. With learning being so dependant upon the student’s emotions, affective gaming could herald a revolutionary step in the delivery of education.

1. The Argument for GBL The argument for adapting games for the classroom is not new. Indeed, it would be fair to say that the proposition has advanced little in the past 30 to 40 years. It is remarkable, but we are rehearsing the same arguments for the use of digital games that our colleagues proposed for the use of board games during the 1960s and 1970s. For example, Coleman presented a clear case for why teachers should apply game-based learning back in 1971. His position, will appear all too familiar to proponents of digital gamebased learning: 1. “… people do not learn by being taught, they learn by experiencing the consequences of their actions.”

For over 30 years academics have been discussing the limitations of lecture-based education, and the need for a more experiential-based curriculum. Piaget (1973), Vygotsky (1978), Papert (1980) and Kolb (1984) all propose that actively involving the student in their learning is much more constructive than having them passively listen to a professor – re-affirming the Confucius dictum: “Tell me, and I will forget. Show me, and I may remember. Involve me, and I will understand.”

Digital games, through the necessity of market, are designed such that they encompass good principles of experiential learning (Gee, 2002). Every game offers a learning environment where mastery is rewarded – either by a high score, the introduction of a new item, or progression of the narrative. 2. “Games tend to focus attention more effectively than most other teaching devices …”

To distil those elements of the play experience that motivate and engage, and then have them “bottled and unleashed in the classroom to enhance student involvement,

4

J. Sykes / Affective Gaming: Advancing the Argument for Game-Based Learning

and commitment” has long been the driving force for research into game-based education. (Bowman, 1982, p. 14) It is also the thrust of Prensky’s argument for digital game-based learning (2001). He proposes that games and education have a yin and yang relationship. Games, claims Prensky, are high on engagement and low on content, while education, being high on content and low on engagement, is the exact opposite. 3. “The teacher’s role reverts to a more natural one of helper and coach.”

In the UK, students are entering further education in ever higher numbers. Lecturing staff are over stretched – typically being responsible for many hundred students, both in the delivery of material, and in the provision and subsequent marking of assessment. Although institutions might prefer to administer an experiential learning curriculum, the lecture format remains because the ratio of one teacher to many students is the natural economic solution. However, game-based learning could potentially change the teacher workload, and ultimately the student-teacher relationship. Game based learning is particularly suited to pedagogy where the student is active in their own learning. The game provides a learning environment, and introduces goals and tasks which the student must complete to progress. This frees the teacher from delivering content, allowing them to instead provide guidance, helping the student to find appropriate solutions to problems presented in the game. An element which enhances our engagement with games is immediate feedback. When playing games, we instantly see the consequence of our actions, and our subsequent score is a measure of our performance. Having the game provide student assessment reduces the confrontational nature of the student-teacher relationship. Currently the teacher provides the assessment, challenging the student with exams and coursework. Games-based learning allows the teacher to play a supportive role, rather than that of opponent. These above represents the main arguments used to support digital game-based learning. They are also the same arguments used to support board game-based learning over 30 years ago. Has anything changed to suggest that digital gaming offers anything above and beyond the traditional board game? The introduction of digital gaming could potentially change GBL considerably. Gaming in the digital age is ‘interactive’. The game and the player enter a dialogue, where the player identifies a state of play, and then initiates an action. The game accepts those actions, and in return presents a response. When entertaining a dialogue with fellow humans, we typically communicate verbally. However, to interpret the message we rely heavily on para-linguistic information. It is not just the words themselves that communicate meaning; it is the intonation of our voice, the pace at which we speak, eye contact, physical distance, the expression on our face, the way we gesticulate. Evidence shows that paralinguistic communication is actually crucial to the transfer of the message, and is perhaps more important than the verbal content (Mares, Henley, & Baxter, 1985). Saying ‘fire”, expresses a very different meaning to yelling “FIIIRRRE!” and running around madly waving your arms. Although not always recorded, human-computer interaction also contains a great deal of paralinguistic information – the way we strike the keyboard, the way we move the mouse, how frequently we look at the VDU, and so forth. By moving gameplay to the digital arena, it is possible for games to react to how players perform an action, as well as the action itself. This clearly makes digital GBL a significant improvement on traditional GBL. There is an immense difference between a confident and a pensive

J. Sykes / Affective Gaming: Advancing the Argument for Game-Based Learning

5

action. We associate a confident and correct answer with understanding. A student who is dithering, and provides an unconfident response, deserves greater attention, regardless of the actual answer they give.

2. Affective Gaming Affective computing is a sub-domain of paralinguistic computing. The goal of affective computing is to ascertain the emotional state of computer users. When we communicate with others, we consider our partner’s affective state in formulating our response. If our partner is upset, our words, our tone of voice, and our general behaviour will differ to an interaction where they are happy. Picard (1997) argues that this is a sign of emotional intelligence, and is a fundamental aspect of human intellect. She goes on to argue that that if computers are to demonstrate intelligence, they too must respond appropriately to their user’s emotions. Emotion is characterised by changes in physiology and behaviour. When scared, our heart rate will increase, our skin might change colour as blood is drawn from the surface of the skin, we might perspire more than usual, our voice would probably alter in pitch and fluency (Johnstone & Scherer, in press), our facial muscles would alter our appearance (Ekman, 2003; Hazlett, 2003), and our movement would differ (Camurri, Lagerlof, & Volpe, 2003). Humans are very good at noticing subtle changes that indicate a person’s emotional state, and machines can be programmed to pick up on the same information. Measuring of emotion by machine often involves extra peripheral equipment (such as heart rate monitors, electro dermal activity monitors, EEG, and so forth), but can also be implemented using more standard equipment. At the eMotion Laboratory we have been investigating the possibility of detecting the user’s emotional state by their use of peripheral equipment – such as keyboards, gamepads, and mice. We have already started our investigation into the emotional signature of button depression (Sykes & Brown, 2003), and are now investigating whether the motion of the gamepad during play can indicate player affect. Affective gaming is concerned with the application of affective computing techniques to the domain of digital games. Affective gaming involves both the ‘evocation’ of emotions, as well as the detection of player emotion. Research undertaken at the eMotion Laboratory has observed how strongly game environments can evoke emotion (Sykes & Wiseman, 2003). Noticing that some environments make us feel small and uncomfortable, where others make us feel warm and secure, a study was conducted taking reputably haunted environments in Edinburgh and modelling them in Unreal Tournament. When touring the virtual rendition of Edinburgh’s underground vaults, users would often report paranormal-like experiences – such as feeling another’s presence, feeling the breath of a ghost on the neck, suddenly feeling cold on entering a room, and seeing and hearing things which had not been added to the environment. An alternative approach is to evoke emotion through narrative. Freeman, an exHollywood script writer, believes that game design can apply many of the same techniques currently used in film to evoke emotion (2003). We watch movies and read books because the story moves us. By adapting a game to include strong narrative elements, it is possible to accomplish the same effect in games.

6

J. Sykes / Affective Gaming: Advancing the Argument for Game-Based Learning

3. Affective Gaming and GBL There is wide agreement that emotion is a significant component of the learning process. The Yerkes-Dodson Law suggests that the quality of student learning is related to their level of arousal (1908). Bloom’s taxonomy of learning (Bloom, 1964) places emotion at the heart of the learning experience. Others have found intellect and emotion to be inseparable when considering meaningful and sustained learning (McCombs, 2001), arguing that emotion drives attention, learning and memory. Clearly, the evocation and detection of emotion during play could potentially improve the quality of learning in GBL environments. The learning process can evoke a wide variety of emotions – both positive and negative. We become despondent when we make mistakes, we are hopeful when we try again, and we are thrilled when we finally succeed. Students experiencing negative emotions can find it difficult to learn because they find it hard to attend to the material (Goleman, 1995). The continual monitoring of student emotion, and the ability to counter negative emotions is therefore clearly beneficial. Conati (2002) and Kort et al. (2001) are looking at the use of intelligent agents to monitor student emotion when playing educational games, so that they can intervene when students become overly despondent. Should the student become ashamed or frustrated when they make a mistake the agent intervenes and attempts to make the student feel better about their performance. By changing the affective state of the player, the agent not only makes the learning experience enjoyable, but also productive.

4. Conclusion Although valid today as it was 30 years ago, the argument made for game-based learning has not advanced, making it hard to see how digital games are going to impact education, when board games have failed. However, it has been suggested here that the advancement of digital gaming technology has initiated new forms of interaction which support paralinguistic computing. In particular, it has been shown that games could potentially detect and steer the affective state of players offering significant improvement on current forms of game-based learning.

References [1] Bloom, B., Berram, M., & Krathwohl, D. (1964). Taxonomy of Educational Objectives. New York: David McKay. [2] Camurri, A., Lagerlof, I., & Volpe, G. (2003). Recognizing emotion from dance movement: comparison of spectator recognition and automated techniques. International Journal of Human-Computer Studies, 59, 213–225. [3] Conati, C. (2002). Probabilistic assessment of user’s emotions in educational games. Journal of Applied Artificial Intelligence, special issue on ‘Merging Cognition and Affect in HCI’, 16(7–8), 555–575. [4] Ekman, P. (2003). Emotions revealed: understanding faces and feelings, London: Weidenfeld & Nicolson. [5] Freeman, D. (2003). Creating emotion in games: the craft of emotioneering, Indianapolis: New Riders. [6] Goleman, D. (1995). Emotional Intelligence. New York: Bantam Books. [7] Hazlett, R. (2003). Measurement of user frustration: a biological approach. CHI extended abstracts 2003, 734–735. [8] Johnstone, T. & Scherer, K. (in press) The effects of emotions on voice quality. To appear in the Proceedings of the XIVth International Congress of Phonetic Sciences.

J. Sykes / Affective Gaming: Advancing the Argument for Game-Based Learning

7

[9] Kolb, D. (1984). Experiential learning: experience as the source of learning and development, New Jersey: Prentice Hall. [10] Kort, B., Reilly, R., & Picard, R. (2001). An affective model of interplay between emotions and learning: reengineering educational pedagogy – building a learning companion. Conference on Advanced Learning Technologies. [11] Mares, P., Henley, A. & Baxter, C. (1985). Health care in multicultural Britain, Cambridge: Health education Council & National Extension College. [12] McCombs, B. (2001). The learner-centred psychological principles: a framework for balancing academic and social and emotional learning. Center on Education in the Inner Cities Review. [13] Papert, S. (1980). Mindstorms, New York: Basic Books. [14] Piaget, J. (1973). To understand is to invent, Massachusetts: Harvard University Press. [15] Picard, R. (1997). Affective Computing, Cambridge: MIT Press. [16] Sykes, J. & Brown, S. (2003). Affective gaming: measuring emotion through the gamepad. CHI extended abstracts 2003, 732–733. [17] Sykes, J. & Wiseman, R. (2003). Deconstructing ghosts. In Mark Blythe, Kees Overbeeke, Andrew Monk and Peter Wright [Eds.] Funology: from usability to enjoyment, London: Kluwer Academic Publishers. [18] Vygotsky, L. (1978). Mind in society: the development of higher psychological processes, Massachusetts: Harvard University Press. [19] Yerkes, R. & Dodson, J. (1908). The relation of strength of stimulus to rapidity of habit-formation, Journal of Comparative Neurology and Psychology, 18, 459–482.

8

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

Didactic Analysis of Digital Games and Game-Based Learning Matthias BOPP

Abstract. The article looks at (non educational) digital games as learning and teaching environments that may provide helpful suggestions for the design of digital educational games. It suggests to use a three dimensional theoretical framework to identify didactic methods in digital games (situational, temporal and social dimension) and uses this framework to give a brief outline of didactic methods in commercial digital games today. After that it compares these methods with the use of didactic methods in common digital educational games today. Keywords. Digital games, video games, computer games, educational games, didactic analysis, game-based learning, teaching methods, game design

1. Didactic Analysis of Digital Games and Game-Based Learning Games and game-like activities have a long tradition within the theory and practice of pedagogy. Looking at the tremendous role and motivational power (Bowman 1982; Gage & Berliner 1988: 384; Zimbardo & Weber 1997: 325) of digital games (computer games, video games, handheld and mobile games) in youth culture today, it may be a promising idea to use learning and teaching principles of this based learning? In 1972 the manual of the classic comparatively new form of gaming to enhance the ways of learning in school, university etc. (Squire 2003). But are there actually any significant activities of learning and ‘teaching’ in digital games that may be used for serious game-computer game Pong (Atari/Atari 1972) (see screenshot 1) was made up of three sentences: “Insert coin”, “Press start” and “Avoid missing ball for high score”. In 1998, the manual for the flight simulation Fal-

Screenshot 1. Pong, © Atari.

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

9

con 4.0 (Hasbo 1998) contained about 600 pages. Of cause, flight simulations are not characteristic of digital games in general, but the contrast of this opposition nevertheless illustrates a point: the knowledge the player may use in a game is getting more and more complex. There still are digital games today with just a few rules that can be easily discovered by fumbling around with the game pad, and there still are games where the player learns the application of rules by just repeating specific actions again and again. This is especially true for eye-hand-coordination necessary for classic action games. Designers call this type of games “Easy to learn, hard to master”. But this type is getting rare. Even in the genre of action games the player today usually has to reflect throughout the whole gaming session. If you simply run into a new room in Doom 3 (Id 2004) and pull the trigger, you get humiliated (as a reviewer sardonically put it). A reason for this trend towards complexity is the accumulation of basic gameplay skills, the so-called game literacy. Playing a game usually leads to new knowledge about how to play, for example how to take an object and throw it away to distract an enemy. When playing another game (of the same genre), the player tends to expect to be able to do the same things he or she has learned to do in recent gaming. If the game does not give the opportunity to do so, he or she is disappointed. In addition to this, the player will expect to be able to do new things in the new game. In this way, the expectation towards games as interactive environments grows and game design has to live up to this by increasing the complexity of games. This has two consequences both for the player and the game designer. As for the players, they have to learn more and more gaming rules. Some of them are ‘official’ ones, documented in the game’s manual. Others they have to discover throughout the process of gaming. This may be called the demystification of a game (Friedman 1999). Secondly, they have to learn how to apply these rules. Knowing the rules of chess is not the same as being a good chess player. Pedagogically, this is known as the distinction between ‘knowing-that’ and ‘knowing-how’. For designers this trend towards complexity is a huge challenge, because the majority of players do not like to read manuals before the fun starts and even if they did do so, they would not get the knowing-how that the gameplay requires. How do they deal with this situation? They do not only create more and more complex games, but also develop more and more complex and efficient methods to support the learning process while playing. Currently this is a more or less unconscious behaviour, driven by the evolutionary processes of the capitalistic market, which the digital games industry is a part of. Games that trigger learning and thus allow players to do new interesting things are games that people like to play. The games people play are the games that are sold. (Learning) the designing of well-selling games is being copied and varied; it evolves and gains in market shares. Recapitulating, it is obvious that digital games, for economic reasons, are a media of fast evolving complex arranged learning environments and using these methods to enhance ‘serious’ learning may be promising. Nevertheless there are only a few theoretical attempts at describing games as arranged learning environments and how they may be used (Gee 2003; Prensky 2001). And these attempts are neither thorough nor up to date. Thus an important research question for game-based learning is: what design principles of (non-educational) games can be used for educational games? This article proposes a framework to answer this question by applying the method of didactic analysis to digital games.

10

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Graph 1.

Didactic analysis traditionally deals with three main questions that are important to educational game design (see graph 1): 1. 2. 3.

What actually is and what should be learned (learning goals)? What is and should be the material/themes used to reach these goals (learning content)? How should this learning content be learned (learning and teaching methods)?

These three subfields of didactic analysis are strongly interrelated. Nevertheless, this article mainly deals with the question of learning and teaching methods. This approach can be termed didactic-methodical analysis.

2. Methodological Basics Analysing the didactic methods by which digital games support learning is a methodological challenge. Up until now there has been no real awareness of this area within the profession of game designing (this is perhaps because being some kind of ‘teacher’ does not fit into a game designer’s self-concept, although they most certainly are at least a kind of ‘coach’). Design books contain some scattered hints, but nothing more (check, for example, Bates 2001). Beside this, studies on school teaching point to the fact that teachers (and creative professionals like game designers, as well) are not fully aware of what they actually do. This is due to the two sides of professional knowledge in general: it melts descriptive (knowing-that) and procedural knowledge (knowinghow) and it is usually very difficult or even impossible to verbalise procedural knowledge. That is why, for example, the attempts to program teaching expert modules as parts of intelligent tutoring systems (ITS) are still not very successful. Thus, didactic researchers cannot simply ask designers how they support learning. The next obvious thing to do is to analyse games with methods that are used in other arranged learning environments, for example, teaching, educational software, and edutainment. This, too, is demanding. Currently, there are two major attempts to analyse computer games in this way. The first one is Prensky (Prensky 2001: 157–163), who identifies fourteen different methods of learning and teaching in computer games: Practice and feedback; learning by doing (practice and feedback, exploring, discovery and problem solving, and task-based learning); learning from mistakes; goal-oriented learning (‘learning to do

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

11

Graph 2.

something’ instead of learning facts); discovery learning and ‘guided discovery’; taskbased learning; question-led learning (questions that induce a process of reflection); role playing; coaching; constructivist learning; multisensory learning; selecting from learning objects (learning of knowledge that can be hooked up on demand in any order by either an instructor or the learner); intelligent tutoring (intelligent tutoring systems). The second attempt is Gee (2003: 207–212) with a list of 36 learning principles in computer and video games and Gee (2004), which contains 12 principles. Gee’s principles are strongly centred on the traditional ways of guided learning by doing. Reading the (oddly assorted) lists of Prensky and Gee brings to light two methodological problems concerning the analysis of digital games as learning environments. Firstly, there are thousands of teaching and learning principles at hand. Thus, some kind of systematic ordering may be useful. Secondly, research on the effectiveness of teaching methods indicates that it is important to analyse teaching methods at a very detailed level. There are effective and ineffective ways of lecturing, of learning in small groups, of guided discovery, of using multi-modal ways of communication, and so on. Therefore, you do not know much about computer games when you can tell that they apply ‘guided discovery’. Tools of analysis need a much higher resolution than this. This is especially important when one wants to improve school learning and edutainment software by looking at digital games. Neither Prensky nor Gee reflect on these problems. I suggest focusing on three fine-grained dimensions in the description of games as learning environment (note that this is a contingent decision which has to justify itself through its fruitfulness in the course of research): • • •

the dimension of arranged learning situations (situational dimension), the temporal dimension in which situated actions are arranged, the social dimension in which actions take place.

All of these dimensions are interrelated and may be applied within a larger framework of didactic analysis which takes into consideration learning goals and learning content, as well (see graph 2). The following paragraph will deal with these methodical dimensions in detail.

12

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

3. Basic Dimensions of Games as Arranged Digital Learning Environments 3.1. The Situational Dimension The concept of a situational dimension of gaming is based on theories of microanalysis in sociology and pedagogy (see, for example, Goffman 1972). Microanalysis theories in turn are based on an action theory paradigm. They describe the phenomenon of gaming as a series of gaming situations. A (gaming) situation is characterised by • • •

a person’s concept of an environment (the environment may include other people and their actions), the goals the person wants to achieve within this environment, the tools and plans at hand to reach these goals, certain intentional (goalseeking) actions to actually achieve the goals (see Miller, Galanter, Pribram 1960 and in the field of Human-Computer Interaction (HCI) Suchman 1987).

An arranged learning situation is a certain type of gaming situation that is structured by a) the actions of the game designer (or a team of designers), who creates environmental learning stimuli for the player; and b) the actions of the player, who responds to these stimuli in a way that leads to the kind of learning which the designer wants to stimulate. Arranged learning situations can be distinguished from three other gaming situations that are important for the analysis of games as learning environments: a situation where the player does not learn the way that the designer wants him to, may be called an unsuccessful or pure teaching situation. An example of this may be a boring, unsuccessful situation within a tutorial. A situation in which the player learns something, which the designer did not intend him to learn, it is a pure learning situation. The latter often takes place, when a player is creative in a way the designer did not anticipated. The designer Harvey Smith gives an example of such a pure learning situation when talking about Deus Ex (Ion Storm 2000): “For instance, some clever players figured out that they could attach a proximity mine to the wall and hop up onto it (because it was physically solid and therefore became a small ledge, essentially). So then these players would attach a second mine a bit higher, hop up onto the prox mine, reach back and remove the first proximity mine, replace it higher on the wall, hop up one step higher, and then repeat, thus climbing any wall in the game, escaping our carefully predefined boundaries.” (Smith 2002)

Arranged learning situations are by definition characterised by an intention of the designer to prompt learning. This is the main difference to situations where the designer just wants to cause the player to do something but does not intend any long-term changes in player behaviour. These situations may be called guiding situations. With regard to the designer’s methods, both types of situations often use similar forms of instruction. Usually it is only the larger context of a situation that makes it possible to distinguish one from the other. The designer, of course, is not actually present within the gaming situation, but his or her prior actions lead to certain events within the game world and the player reacts to these events. Thus, typical arranged learning situations are made of specific pairs or patterns of designer and player actions. Some common types of patterns are described below.

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

13

Screenshot 2. Splinter Cell, © Ubisoft 2002.

3.1.1. Writing, Telling, Illustrating – Reading, Listening, Looking At The most obvious type of arranged learning situations in digital games is the presentation of learning content in a written or illustrative form. In the past, these were often simply presented by info screens that interrupted the play. Meanwhile it is common to motivate these presentations by simulating various media within the game world, for example PDAs as in Splinter Cell 2 (Ubisoft 2002; see screenshot 2) and Doom 3 (Id 2003), books, cell phones, records, computer files, sheets of paper, and wall or street signs as in Half-Life (Valve 1998) and Vampire. The Masquerade: Bloodlines (Activision/Troika 2004; see screenshot 3). A similar thing happens when, in a certain situation in Silent Hill 2 (Konami 2002), the avatar walks along a street and discovers the corpse of a slaughtered man and several sheets of paper on the ground. As soon as he reaches down, the text of the sheet is displayed on the screen (see screenshot 4). Learning content may also be presented audibly, for example by a NPC (Non Player Character), a character which is controlled by the computer. It is common to simultaneously display in written form what the NPC is saying (multisensory learning). These telling/writing-situations often occur extensively within tutorials and introduction levels, but may also be scattered throughout the whole game. FPS and stealthshooters like the Splinter Cell series (Ubisoft) use this method quite often. The hero communicates with his mission team via radio and receives situated advice. In these situations, the player usually just has to listen; whereas in the Metal Gear Solid series (Konami) he or she additionally has the opportunity to actively ask for advice by using the “codec” (feedback on demand). Of course, the players do not here formulate

14

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshot 3. Vampire. The Masquerade: Bloodlines, © Activision/Troika 2004.

Screenshot 4. Silent Hill 2, © Konami.

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

15

Screenshot 5. Detail of a screenshot from Metal Gear Solid 2: Substance, © Konami.

a question on their own. Instead, the game displays a scripted conversation, covering situated matters. Important information of this kind is often stored in a virtual notebook. 3.1.2. Demonstrating – Observing Another typical learning/teaching -situation is the display of a process or skill while the learner observes the display and acquires knowledge (see Bandura 1963). This can be realised in different ways. In an explicit way, the learner is informed about the learning character of the situation. Take for example a situation from Metal Gear Solid 2: Substance (Konami 2003). Here the Colonel (a member of the mission team in contact with the protagonist via radio) gives Raiden (the hero) audio advice to throw a chaff grenade to clear a connecting bridge. Simultaneously, this action is demonstrated in a little video clip (see screenshot 5). In a more implicit way, the model process may be displayed to the player without a direct comment about the situation. Take as an example a situation from Silent Hill 2. In the Woodside Apartment, the protagonist enters a rather dark room (see screenshot 6). A torch that is attached to a tailor’s dummy illuminates the part of the room in front of him. In the dark background, something is lying motionlessly on the ground. When the player takes the flashlight, the rear part of the room suddenly is illuminated and the object on the ground (a monster) comes to life and attacks the player. This may be interpreted as an implicit demonstration of the same learning content, which is presented in written form in screenshot 4. (The less overt a demonstration is, the closer it is to the didactic method of ‘guided discovery’.) Furthermore, these situations are an example of the possibility of using different types of didactic methods to teach the same kind of learning content. An important focus of didactic analysis and planning is the question of what learning content (and goal) may be used together with what kind of method.

16

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshot 6. Silent Hill 2, © Konami.

3.1.3. Stealth Teaching Situations Gamers do not like being taught or guided, because the enjoyment of gaming often comes from the enjoyment of solving problems on one’s own (or within a team of peers). Because of this, designers sometimes try to avoid the impression that games use didactic methods. As this camouflage is a didactic strategy as well, I propose to call this stealth teaching or stealth guidance. The terms are loosely affiliated to the concepts of implicit learning and teaching. Stealth teaching is a generic term, which covers several types of learning and teaching situations. The concept (not the term) can be traced back to the Gestaltpsychologist Maier and his classic two cords experiment (see Maier 1931). In Maier’s experiment, two cords were hung from the ceiling of a laboratory strewn with many objects such as poles, screwdrivers, and pliers. The subject was told his task was to tie the two ends of the cords together. The problem in doing so was that the cords were placed so far apart that the subject could not, while holding onto one cord, reach the other (see image 1). After the subject had been stumped for several minutes, Maier, who had been wandering around the room, casually put one of the cords in motion. Then, typically within 45 seconds of this cue, the subject picked up a weight, tied it to the end of one of the cords, set it swinging like a pendulum, ran to the other cord, grabbed it, and waited for the first cord to swing close enough for it to be seized. Immediately thereafter, Maier asked the subject how he had got the idea of a pendulum. The question elicited such answers as ‘it just dawned on me’, and ‘I just realised the cord would swing if I attached a weight to it’. All of the subjects denied that Maier’s cue had played any role in their solution.

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

17

Image 1. Maier’s two cords experiment, drawing.

A similar principle of ‘covered guided discovery’ can be found, for example, in the adventure game Broken Sword 3: The Sleeping Dragon (Revolution 2003). Here the hero searches a table in a scientist’s laboratory and picks up several items, for example a postcard from Glastonbury, England and a magnifying glass. When he takes the glass, he makes a casual remark on an episode from his youth in sunny California, when he set his father’s straw hat on fire. Several gaming situations later, outside the laboratory, the player has to solve the problem of making fire within a stone statue that serves as a fireplace. He must solve this problem by taking a dry bird’s nest, putting it onto the fireplace within the statue and using the magnifying glass to light up the bird’s nest (see screenshots 7–9). The main point here is that the casual remark several gaming situations earlier may not be remembered consciously but it, nevertheless, primes the correct solution for the problem (for literature on different ways of ‘priming’ see the experiments of Nisbett/Bellows 1977 and Nisbett/Wilson 1977). Games use stealth techniques not just in regard to problem solving, but also in regard to the guidance of the user, as well. Here a typical field of application is to lead the player from one place within the game world to another while giving the impression that the player has found out the way independently. An example of this can be observed in Deus Ex: Invisible War. On the first level of the game, the player enters a hallway and has to find apartment No. 454. As soon as he approaches the next crossing, an explosion occurs and a frightened janitor appears, runs down the hall and chooses the next corridor to the left. If the player follows him around the corner, he sees the janitor standing at the end of the hall, looking at him. After a short conversation about the explosion, when the player again starts to look for apartment 454, he realises that he is standing right in front of it (see screenshot 10–12). 3.1.4. Pointing at – Being Observant Digital games, especially 3D-games, display large areas of landscape or rooms that the player has to explore or deal with. To ease these requirements games sometimes ‘point at’ certain objects to guide the player’s attention. This may be done in various ways. For example, designers may point at objects by using alienation effects as in Grand Theft Auto: Vice City (Rockstar North 2003) (see screenshot 13).

18

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshots 7–9. Broken Sword 3: The sleeping dragon, © Revolution/THQ.

Screenshots 10–12. Deus Ex: Invisible War, © Ion Storm/Eidos.

They may label objects in the game world by a kind of underlining, in a way common to strategy games in unit labelling; and they may frame them as in XIII (Ubisoft 2003) (see screenshot 14). Or they may, for example, design a level in a special

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

19

Screenshot 13. Grand Theft Auto: Vice City, © Rockstar North/Rockstar.

Screenshot 14. XIII, © Ubisoft.

way, for example by using fluorescent sticks as in Jedi Knight III: Jedi Academy (Raven 2003) to indicate a path (see screenshot 15).

20

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshot 15. Jedi Knight III: Jedi Academy, © Raven/LucasArts.

Screenshots 16/17. Jedi Knight II: Jedi Outcast, details of two screenshots, © Raven/Lucas Arts.

And NPCs may be used for this task. For again example, in the first gaming situation of Jedi Knight II: Jedi Outcast (Raven 2002), the NPC Jan walks across a great square, kneels down behind some containers and looks in a certain direction. When the player follows him and looks in the same direction, he or she can see some storm troops – the first enemies within the game (see screenshots 16 and 17). Many of the mentioned situations are instantly followed by other situations in which the player must immediately apply the knowledge he or she has just acquired. This may be called the ‘just-in-time’ principle of teaching/learning in digital games (for

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

21

Screenshots 18–20. Deus Ex: Invisible War, © Ion Storm/Eidos.

analysis of such temporal relations between arranged learning situations see chapter 3.2). These situations are just a selection of types of arranged learning situations in digital games. Others are, for example, provoking the player, giving rewards (as tools, money, praise etc.), punishing (for example with the ‘time-out’-method of games: reloading) etc. As the technical basis of games develops, other traditional educational methods will be transformed and used in the virtual world of gaming. It is important to notice that a situational analysis of games should be integrated into a more complex didactic analysis which takes into account the learning content and the learning goals that are to be achieved. This is important as some combinations of methods, goals, and content suit each other better than others. Consider, for example, the following didactic problem from Deus Ex: Invisible War. This was one of the first games that contained a physics engine (the “Havok” engine) which allowed the player

22

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

to take objects and throw them, while the objects than bounced off more or less physically correct. The learning goal here is to teach the players that they may take objects in this way to solve problems. As a learning method, it suggests a learning-by-doing method to inform the player about this feature of the game instead of telling him about it. The interesting question now is how to stimulate the player to throw something around. A not very elegant way to do this would be to simply tell him to do so. Deus Ex: Invisible War applies another way. On the level of the learning content, it chooses a special object, which calls itself for being taken and thrown around: a baseball (see screenshots 18–20). 3.2. The Temporal Dimension and Phases of Learning Learning situations in games often last just a few seconds. Thus, analysing digital games suggests several questions. How are these learning situations arranged within the game temporally? In a linear sequence, a random order, a bottleneck pattern, or what? How fast is the learning curve within the game? What is the temporal relation between information and the opportunity to apply it? Are there phases, or steps of gameplay? Questions of this type concern the temporal dimension of games. A phase of learning is a part of the game that usually contains several individual situations which together serve a special function within a higher-ranking learning process plan. Digital games have systematically arranged learning situations in this way since their dawn 35 years ago. The most common way to do this is the hierarchical approach (see Eraut 1985 and Dick/Carey 1996). Here the designer has to ask: what does the player already have to know so that this challenge (for example fighting a ‘boss’) can be learned? By asking this question, the designer can identify one or more critical subordinate skills that the game has to teach the player prior to a main challenge. Learning situations may then be constructed and disposed accordingly. For example, simple learning situations concerning movement within the game world usually are the first to be presented within a game, followed by situations with more prerequisites. During the evolution of game didactics, the design of learning phases has got more and more complex. For example, in Space Invaders (Taito 1978), there are just two basic skills: shooting at aliens, which approach from the top of the screen, and avoiding getting in touch with their bombs and bodies. According to this, the speed of the attacking aliens accelerates from level to level (in this perspective the common term ‘level’ is a synonym for learning phase). Nowadays digital games usually imply not one but dozens of different basic skills, and the arrangement of learning situations is correspondingly more difficult. Splinter Cell (2002) level designer Clint Hocking gives an example (see screenshots 21–23): “The player must learn them [the skills; M.B.] once, than [he must; M.B.] be able to build on that knowledge… You can start with a simple challenge in an early level … like Level 1 … you learn how to climb a pipe. Come Level 3, you’ll want the player to have to climb that pipe again, but in a more challenging scenario. An enemy might be looking out a window. Then in Level 5 or 7 the player will climb a pipe with the enemy at the window and a moving dynamic light searching the area of the pipe. Once the player has learned a set of skills in one part of a map or over the course of a few maps you’ll then want to encourage the player to face all of these challenges together or in close proximity to each other.” (Hocking 2003)

Another example is Prince of Persia (Konami 1989).

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

23

Screenshots 21–23. Screenshots from a video documentation on Splinter Cell, © Ubisoft.

“[...] when the player first encounters a break-away floor in Prince of Persia falling through it is non-lethal. Similarly, spikes are introduced in such a way that the player is very likely to notice them and to be able to survive them. Subsequent encounters with spikes will not be so forgiving, but by then the player has learned of the threat they pose to his game-world character, and if he is clever he will be able to survive them.” (Rouse 133f.)

This type of arrangement is a typical large-scale variation of the hierarchical approach. Large-scale hierarchies organise long periods of gaming. Small-scale hierarchies can be found as well, usually within a single level. For example, on the fifth level of Forbidden Siren (Sony 2004) the player, for the first time, has to attack a zombie. A

24

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

1 3 2

Graph 3. Game time.

subskill for this is the ability to hit something that does not move. Hence, a situation is first arranged by the designer where the player must overcome an iron gate. To overcome it he or she has to smash the gate with a poker, using at least three heavy blows. This poker can then be used as a weapon to fight the zombies behind the gate. Smallscale hierarchical arrangements like this are typical for tutorials and introduction levels (see chapter 4). The hierarchical approach is closely related to the concept of learning curves in games. The term is usually used to describe the relation between the demands of gameplay situations and the progress of the gameplay in time. A curve gradient that reflects this relationship can be used to illustrate the didactic structure of a specific game design and related problems (see graph 3). For example, the starting point and steep gradient of curve 1 in graph 3 indicates a didactic design that relays strongly on the game literacy of the player. This is sensible, for example, in the case of an add-on. The gradient of curve 2 indicates that the game confronts the players with new challenges without having given them time to learn the skills they need to deal with these challenges. A decline in the learning curve indicates a part of gameplay that is easier than the previous ones. Curve 3 illustrates the skill curve of a player. For this individual, game 1 with learning curve 1 would probably be frustrating; game 2 would be too difficult in the beginning and too easy in the end. Methodically, it is difficult to quantitatively measure the demands and difficulties of a gaming situation or a player’s skills. Nevertheless, learning and skill curves such as in graph 4 are of heuristic value, as they point to the basic didactic problem of bringing game demands on level with player skills and vice versa. Describing games by means of the hierarchical approach or learning curves is just one of the two various possibilities of identifying phases of learning in digital games. Other categories to identify phases of learning are: •

Taking account of prior learning/adaptation: This usually happens at the beginning of a game. The player may be able to set the level of difficulty within the game, for example by identifying himself as a rookie, veteran etc. or by directly choosing between the levels of easy, normal or hard. This does not only apply to the action level of games, but, for example, to riddles as well (see the Silent Hill series). Alternatively, the game may ask the player whether he or she has finished a prequel of the game or whether he or she likes action in games (both in Metal Gear Solid 2: Substance); or it may pose personality related questions about general wishes and fears (both in Kingdom Hearts,

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

•

• •

• • • • •

•

25

Squaresoft 2002). The game may also take in account the performance of the player throughout the whole game and automatically adjust the challenges in one way or another (for the importance of prior knowledge and a player’s vocabulary from a designer’s perspective, see Johnson 2001). Motivation/reinforcement: Motivation in games appears in many ways. Significant phases for this are situations which reward the player, usually with helpful new tools, or socially, for example by praising him or provoking him using NPCs. Others are cutscenes and the non-interactive story parts of a game in general that can serve this purpose by establishing an interesting setting for the game challenges. Informing the learner of the objective of an upcoming part of the game: This, for example, may be done by using briefings at the beginning of a mission or during the course of a level. Repeating and testing: Both of these types, of course, are very common in games. Repetition may be arranged by designing series of more or less similar challenges or by a reload of a game passage after the player has failed a test. In games like the Tomb Raider and Indiana Jones series, falling to the ground and the necessity to climb up again is a typical variation of this ‘failed-testand repetition’ pattern without an actual reload. Stimulating recall of prerequisite learning: Here usually an NPC points at the skills the player has learned before (for example at the beginning of the first level of Deus Ex.) Directing attention phase: This may, for example, happen visually (see screenshots 13–15) or audibly (when a hostile NPC guard behind a corner talks to a colleague or asks: “Who is there?”). Providing direct learning guidance: This is a phase usually employing tellingwriting-situations. Relaxation phases Providing explicit informative feedback: This is usually done at the end of a mission; for example, by using score lists or statistic tables of the player’s performance as in Thief: Deadly Shadows (see screenshot 24), a general verbal summary at the end, for example of Tropico (Pop Top Software 2001) or a “replay” in a racing game. Providing emotional feedback: This happens, for example, by praising or punishing the player with a ‘time out’ while reloading a previous state of the game. Analysing feedback in games includes the question whether feedback is given remotely or on the spot. This difference may have an influence on the learning process (see McKeachie 1974). In general, digital games provide a lot of instant feedback.

As the examples above indicate, most phases of learning can be realised by different types of learning situations. In general, it is typical of digital games that information is presented in a situated and ‘just-in-time’ form, which means that the player has to use that information quite soon. By repeated usage, this declarative knowledge then is transformed into implicit knowledge and long time memory. Learning situations are relatively easy to recognise in games because they are constituted by behaviour that is observable. By contrast, the identification of learning phases and other temporal aspects of gaming refer to psychological theories about

26

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshot 24. Thief: Deadly Shadows, © Ion Storn/Eidos.

learning on the part of the player and to designer’s theories of learning. Therefore, their identification is often based on disputed theory. 3.3. The Social and Parasocial Dimension Man is a social being and he learns differently depending on whether he is alone or not. Especially learning together with others, with a teacher or in a group can support learning significantly. Thus, it is important to take in account the impact of the social dimension on learning in games. Relying on theories of social aspects in other learning environments, four basic forms of social arrangements and learning in digital games can be distinguished: • • • •

playing alone and learning by self instruction (in a way this, of course, is just the absence of a social situation), playing with a partner, which may involve various forms of partner learning, playing in various forms of groups, which may lead to various forms of group learning, playing within a virtual social situation which may support parasocial learning.

The first three terms of learning are quite common in educational science. In contrast, the term “parasocial learning” is a rather unconventional concept and the focus of this paragraph. As learning often is much more motivating when it takes place in a social situation, it would be a grave disadvantage for digital games not to be able to use this didactic

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

27

Screenshots 25. Maniac Manson (1987), © Lucasfilm; Tomb Raider II (1997), © Core Design/Eidos; Broken Sword III – The sleeping dragon (2003), © Revolution/THQ.

dimension for arranging their learning environment. However, how should they do this in single-player games? Here, at first sight, there is no social situation at play. At second sight, however, a functional equivalent can be found. It is a well-documented phenomenon, that “[…] individuals apply social rules and social expectations to computers [and other media and objects as well; M.B.] That is, individuals use the same social rules to assess and respond to the performance of computers that they use when assessing or responding to other individuals, even when they are fully aware that they are interacting with machines.” (Nass/Steuer 1993)

A well-known historical example of this behaviour is the interaction of players with the conversation program Eliza, (http://i5.nyu.edu/~mm64/x52.9265/january1966. html, for instance) developed by Joseph Weizenbaum at the Massachusetts Institute of Technology (MIT). Some people got really emotional during the “conversation” and some psychologists suggested developing it and using it as a virtual therapist (see Weizenbaum 1976). Feelings and behaviour that are typical for social interaction and that occur during interaction with computers may be called “parasocial”. Originally, the term was used by Horton and Wohl (1956) to describe a special form of designing and watching TV shows: when the host of a show says “Good evening, how are you?” etc., he simulates a social interaction where there actually is none. Nevertheless, some viewers regard him as a good friend, a member of the family. TV shows intend to support parasocial feelings by simulating social interaction between real people. In contrast, digital games support parasocial behaviour (and feelings) by simulating social actors: the NPCs. They do this by simulating cues that are fundamental to human interaction. “Among the […] primary cues that appear to be important are the use of language […], interactivity […], filling of roles traditionally held by humans […], and voice […].” (Nass/Steuer, 1993)

Obviously, all of this is more and more the case with NPCs in digital games. Nowadays, NPCs talk to the player, they take orders from him, accompany and support him; they praise, insult, challenge, and flatter him. Additionally, there are improvements in the graphic surface of game characters (see screenshot 25), which allow for more convincing facial expressions, a feature which is very important for lifelike faceto-face-interaction. All this seems to be more and more effective. Thus, for example, a

28

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshot 26. Knights of the old Republic, © BioWare/LucasArts.

warning in a walkthrough (Shunichiro 2004) for Forbidden Siren (Sony 2004) that goes: “Don’t get attached to the characters. That way you will feel much better when the game gets cruel...” (Shunichiro 2004). The tend towards parasocial cues in games will go on in the future, due to such technical developments like microphones that allow the player to actually talk to NPCs, and (in the future) cameras that scan the body language and facial expression of the player and use it as an input. Besides these technological developments, the strengthening of the narrative elements and the deepening of NPC’s characters support parasocial relationships as well. An example is the role-playing game Knights of the old Republic (BioWare 2003). It explicitly asks the player to “talk to” NPCs in his party, because “something seems to worry” them (see screenshot 26). An interesting allusion to the phenomenon of parasocial relations in games can be found in Deus Ex: Invisible War. Here a lonely NPC citizen talks to a hologram AI (artificial intelligence), modelled after a famous singer (“NG Resonance”) within the game world (see screenshot 27). He asks the hologram if NG Resonance ever listens to the conversations between her holograms and her fans and the hologram answers: “Of course she does. I love to listen to my fans.” (translation from the German version of the game). The importance of parasocial phenomena for didactic methods lies in its potential to create motivating learning situations, among others. It is, for example, usually easier to take advice from someone you like than from someone you are not affected by. It is more motivating to command and take care of a team if you care about its members. It is more challenging to fight a bad guy that is convincingly nasty. It is an effective reinforcement to praise a player by means of an NPC he is affected by, and so on. I suggest using the term parasocial didactics for methods and content, which systematically support and build on parasocial feelings and behaviour to support learning. Parasocial didactics can be seen as a development of several traditional classroom methods; for example, the teaching story, a method sometimes used to make history alive by telling it from the perspective of real or made up persons from the past (for basic theory on thinking and narration, see Bruner 1990). Other examples are the broad tradition of educational role-play and the anchored-instruction approach. The best-

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

29

Screenshot 27. Deus Ex: Invisible War, © Ion Storm/Eidos.

known example of the latter is the “Jasper-Project” that produced a series of teaching films narrating “The adventures of Jasper Woodbury”. Jasper is engaged in several practical problems the learner then has to solve by using mathematics. All the necessary information for this is to be found in the story itself (see CTGV 1997). These traditional methods (except for role-playing games) do not allow the learner to interact with the social setting they create – digital games do. Compared to the arrangement of real social situations, parasocial didactics have some obvious advantages and disadvantages. On the one hand, in a real social setting behaviour and characters of the involved persons are not extensively controllable. Role players may act “badly”, teachers may behave unprofessionally (get tired, tease learners etc.). This, in return, often affects the effectiveness of learning. Compared to this, parasocial environments promise a much higher degree of controllability, as the NPC “cast list” can be modelled in whichever way to support learning and then to behave exactly in the way defined by their role. On the other hand, NPCs do remain virtual and this, of course, may reduce the player’s engagement significantly. Besides, the AI of NPCs today is still poor, especially in regard to all kinds of verbal exchange. Thus, the challenge for designers today is to create believable virtual social situations where the expectations of players towards their virtual social companions fit their limited AI skills. 4. Didactic Conventions: Tutorials and Introduction Levels Didactic-methodical elements are usually not distributed randomly within digital games. They build patterns. During more than 30 years of game evolution, several of these

30

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

patterns have developed and crystallised into didactic conventions which are to be found in lots of games. Didactic conventions are an important target of didactic analysis of digital games. They may have a short or long range. Tutorials and introduction (the topic of this paragraph) levels are two long-ranged forms. Tutorials may be seen as remediations of another didactic convention in games: the manuals. Manuals have accompanied computer games since their beginning, they have slowly been remediated into the game world itself and have became what may be called in-game tutorials. The reason for this remediation is that learning goals and content of games mainly cover practical skills, and practical skills are learned more easily by instructed doing than by just reading. Nevertheless, manuals are sometimes still important in the genres of role-playing games and simulation. Here the learning goals often cover not only the practical skills but the knowledge of complex mathematical sets of rules, which cannot easily be learned only by instructed doing. An early historical example of a tutorial in the action genre is Karate Champ (Nihon Bussan/AV Japan 1984), one of the first beat ‘em ups. It starts with a short tutorial at the dojo where the computer makes a move and the player has to copy it to score extra points (demonstration). When this training is complete, the player is able to mix with the big boys and the tournament begins. Another early example in the adventure genre is Robot Odyssey (The Learning Company 1984). The objective of the game is to escape from a dangerous underground robotic city. To escape the player has to program his or her robots to solve various puzzles, and, before that, there is a facultative ingame tutorial to learn how to program his or her bots. Within the situational dimension, tutorials are game levels with a high degree of situated writing and telling, usually followed by an opportunity to apply the newly acquired knowledge. Additionally, there is just a small threat of failure and ‘time out’ feedback. Temporally tutorials are usually a dense and linear chain of learning locations, where the player literally has to walk from room to room, get instructions, and master one task after the other. Usually a pattern is used that is typical of conventional telltest-learning software (or programmed learning): explanation, instructions to do something, player’s action, feedback, next explanation. However, in regard to the use of weapons, tutorials may provide voluntary phases of repetition. Tutorials are often optional and serve as introductions at the beginning of a game. But there may be tutorials later on in the game, as well, if important parts of the gameplay change during the game. An example of this is Jedi Knight II: Jedi Outcast. Here the player does not get his first light sword and magical powers at the beginning of the game, but later on, and he has to complete a tutorial to learn how to handle his new tool and skills. In regard to the narrative setting, tutorials nowadays share some common features, most of which are to be found, for example, in the early sequels of the Tomb Raider series. Here Lara Croft’s stately home is equipped with a huge training hall (see screenshot 28 from Tomb Raider 3, Core Design 1998), in which the owner may learn and train basic moves. Often, games may vary this ‘training theme’ by simulating institutional training situations like military camps and entrance examinations. Other games like Max Payne (Remedy 2001) do not try to motivate the tutorial in such a narrative way. The game simply places the protagonist on two lone streets, where the player learns how to move, interact with objects, use weapons etc. The setting here is unconventionally surreal because of a public telephone, which gives and provides instruc-

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

31

Screenshot 28. Tomb Raider 3, screenshot from the tutorial, © Core Design/Eidos.

Screenshot 29. Max Payne, © Remedy/3D Realms.

tions, and an “enemy dispenser” button on a wall, which the player may use to get additional cannon fodder (see screenshot 29). Some tutorials try to create a setting which avoids ‘school connotations’. Note that these attempts sometimes are a bit clumsy, for example, when the protagonist’s superiors in Deus Ex and Splinter Cell literally apologise for forcing the player to do the training, or the entrance examinations, or when, as in Half-Life, the training is (ironically?) labelled “Hazard Course”. What tutorials have in common is that the learning does not take place within the actual story or narrative of the game. They share the protagonist and the setting with the actual story but they are not part of the chain of events in the story. As learning within a narrative setting has several advantages, training mission and introduction level take the place of the classic tutorial. Introduction levels are early parts in a game which include the same writing/telling-situations and content as tutorials do, but which are part of the game story

32

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Graph 4.

and avoid educational connotations. Modern examples of this are the first level of Warcraft III (Blizzard 2002) and Halo (Bungie 2002). Looking at the transformation of manuals into tutorials and from tutorials into training missions and introduction levels a trend towards the fusion of tutorial elements and regular gameplay can be identified. If this trend continues, the typical training mission or introduction level will in the future be replaced by game structures that distribute typical tutorial elements throughout the first parts of the game. An example of this is Indiana Jones and the Emperor’s Tomb (The Collective 2003). A quantitative situational analysis of the number of written explanations (typical of tutorials) in each one of the first ten (sub-)levels of the game results in the graph below (see graph 4) and reveals the smooth systematic drop of the amount of written instructions in the first parts of the game.

5. Didactic Methods in Current Game-Based Learning In general, all the didactic methods of entertaining digital games described in this article may be applied to the design of educational games as well and there are a lot more of these methods to discover due to the fast evolution of entertaining games today. Some examples are games and augmented reality, mobile gaming and online games that are not addressed in this article. But even a short look at current game-based learning software demonstrates that today these games usually does not take advantage of the sophisticated didactic methods used in non educational games. Take for example the games by publisher BrainGame, the German market leader in the area of learning games for adolescents and adults.

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

33

Screenshot 30. The ‘game world’ in Physicus, © BrainGame.

Screenshot 31. The corresponding ‘learning part’ in Physicus, © BrainGame.

All these games basically consist of two layers: the ‘game world’ and a corresponding so called ‘learning part’ (see screenshots 30 and 31). The game world is a

34

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

Screenshot 32. Mission: Schatztaucher, © BrainGame.

more or less immersive setting where the player has to solve several riddles or carry out tasks to be able to move on. The dominant form of gameplay is the adventure. Lots of these worlds use a 2-D Myst (1993, Cyan) style retro science fiction setting (see Biolab, 001, Geograficus 2003, Physicus 2002) but there is a growing number of 3-D games as well (see Babylons Fluch, 2003, 10 hours left, 2005, Mission: Schatztaucher, [see screenshot 32] and Flipps galaktische Abenteuer, 2005). As a form of entertainment, these worlds usually feel ‘low budget’: graphics are not state of the art and the world is not very ‘interactive’. Regarding didactic methods, there are manuals instead of in-game tutorials or introduction levels (the only exemption in the sample is Flipps galaktische Abenteuer), the games rarely use NPCs for parasocial didactic methods and they reduce the social dimension to the recommendation to play the game with friends or members of the family, the gameplay itself does not support a multiplayer mode – which may be a drawback as well designed collaborative learning is one of the most promising ways of game-based learning (see Pivec/Koubek/Dondi 2004). All in all, the didactic arrangement of the game world is not very sophisticated. What about the other layer, the so called learning part? This is the main difference between these games and similar non educational adventures (besides less performance of violence): the player has to use academic knowledge to solve the riddles or to carry out his tasks and in order to acquire this knowledge he or she has to use the learning part of the game. Basically, this learning part is a traditionally multi-media schoolbook on certain topics (ancient history, physics, biology, etc.), which usually can be used independent from the game world as well. Sometimes it is necessary to search for the important information within this multimedia

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

35

Screenshot 33. A typical multiple-choice test device of ‘knowing that’ – knowledge from Geograficus, © BrainGame.

learning part; sometimes there is a direct hint in the game world towards a special chapter. Then the player just has to remember this knowledge, seldom he has to do a little transfer. The level of difficulty concerning these ‘academic riddles’ does not change throughout the game. Concerning the situational dimension of learning, this demonstrates that there is some discovery or hands-on (‘handlungsorientiertes’) learning in these kinds of educational games but this learning mainly refers to gameplay knowledge (i.e. ‘What am I supposed to do to make this machine work?’ … ‘O.K., I have to select three species that are typical for the Cretaceous [Kreide] period.’) (see screenshot 33). In contrast, academic knowledge (‘typical species of the Cretaceous period are …’) is mainly acquired by traditional reading, listening and testing (see for such a ‘testing device’ screenshot 33). This is not only caused by the genre of the above mentioned games which are all adventure game. The same goes with strategic games or simulations: games as Mission: Schatztaucher (BrainGame 2003) or the famous ecological simulation ecopolicy (2000, Westermann Multimedia) heavily call for a large amount of knowledge that has to be acquired by reading. Basically, all these games provide more or less (mostly less) attractive fictional settings to motivate reading a schoolbook. This dominance of traditional learning and teaching in educational games may be caused by a lack of knowledge about the didactic design of non educational games or a lack of financial resources necessary to use them. Or maybe it is difficult to actually use these didactic methods to support academic learning content and not just entertaining learning content. Regarding this last crucial point, more research is necessary.

36

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

References [1] Bates, Bob (2001) Game Design. The Art and Business of Creating Games. Premier Press. [2] Bowman, R.F. (1982) “A ‘Pac-Man’ theory of motivation: Tactical implications for classroom instruction”. In: Educational Technology, 22 (9), 14–16. [3] Bruner, Jerome (1990) Acts of Meaning. Cambridge, MA: Harvard University Press. [4] CTGV (Cognition and Technology Group at Vanderbilt) (1997). The Cognition and Technology Group at Vanderbilt. The Jasper Project: lessons in curriculum, instruction, assessment, and professional development. Mahwah, NJ: Lawrence Erlbaum Associates. [5] Dick, Walter & Carey, Lon (1996) The systematic design of instruction (4th ed.). New York: Harper Collins College Publishers. [6] Egenfeldt-Nielsen, Simon (2004) A framework for the role of narratives in educational use of computer games. Download: . [7] Eraut, Michael (1985) “Programmed Learning”. In: The International encyclopedia of education, Vol. 10, Oxford: Pergamon Press Ltd., 4096–4105. [8] Eskelinen, Markku (2001) “The Gaming Situation”. In: Game Studies. The International Journal of Computer Game Research, 1(1), July 2001. Download: . [9] Friedman, Ted (1999) Semiotics of SimCity. Download: . [10] Gage, Nathaniel L. & Berliner, David C. (1988) Educational Psychology. (4th ed.). Boston: Houghton Mifflin, 12–31. [11] Gagné, Robert M. & Briggs, Leslie J. & Wagner, Walter W. (1992) Principles of Instructional Design (4th ed.). Holt, Reinhardt, and Winston Inc. [12] Gee, James Paul (2003) What Video Games Have to Teach Us About Learning and Literacy. Palgrave Macmillan. [13] Gee, James Paul (2004) “Learning by Design: Games as Learning Machines”. In: Gamasutra, March 24. Download: <www.gamasutra.com/gdc2004/features/20040324/gee_01.shtml>. [14] Goffman, Erving (1972) “Fun in games”. In: Goffman, Erving (1972) Encounters. London: The Penguin Press, 15–31. [15] Hocking, Clint (2002) Level Design. Bonus material of Tom Clancy’s Splinter Cell, DVD for Xbox, UBI Soft. [16] Horton, Donald & Wohl, Richard R. (1956) “Mass Communication and para-social interaction. Observations on intimacy at a distance”. In: Psychiatry. Journal of the Study of Interpersonal Processes, 3(19), 215–229. [17] Johnson, Brett (2001) “Great Expectations: Building a Player Vocabulary”. In: Gamasutra, July 16, 2001. Download: . [18] Maier, Norman R.F. (1931) “Reasoning in humans II. The solution of a problem and its appearance in consciousness”. In: Journal of Comparative Psychology, 12, 181–194. [19] McKeachie, Wilbert James (1974) “The Decline and Fall of the Laws of Learning”. In: Educational Researcher, March 1974, 7–11. [20] Miller, George A. & Galanter, Eugene & Pribram, Carl (1960) Plans and the structure of behavior. New York: Holt, Rinehart and Winston. [21] Nass, Clifford & Steuer, Jonathan (1993) “Voices, boxes, and sources of messages: Computers and social actors”. In: Human Communication Research, 19, 504–527. [22] Nisbett, Richard E. & Bellows, N. (1977) “Verbal reports about causal influences on social judgments: Private access versus public theories”. In: Journal of Personality and Social Psychology, 35, 613–624. [23] Nisbett, Richard E. & Wilson, Timothy D. (1977) “Telling More Than We Can Know: Verbal Reports on Mental Processes”. In: Psychological Review, 3 (84), May 1977, 231–259. [24] Pivec, Maja & Koubek, Anni & Dondi, Claudio (Ed.) (2004) Guidelines for Game-Based Learning. Lengerich et al.: Pabst. [25] Prensky, Marc (2001) Digital Game Based Learning. New York: McGraw-Hill. [26] Poole, Steven (2000) Trigger Happy. London: Fourth Estate. [27] Rouse, Richard III (2001) Game Design. Theory and Practice. Wordware Publishing. [28] Shunichiro (2004) Forbidden Siren. Walkthrough Version 1.30. Download: . [29] Skinner, Burrhus Frederick (1954) “The science of learning and the art of teaching”. In: Harvard Educational Review, 24, 86–97. [30] Smith, Harvey (2002) The Future of Game Design. Moving Beyond Deus Ex and Other Dated Paradigms. Download: .

M. Bopp / Didactic Analysis of Digital Games and Game-Based Learning

37

[31] Squire, Kurt (2002) “Cultural Framing of Computer/Video Games”. In: Game Studies. The International Journal of Computer Game Research, 1(1). Download: . [32] Suchman, Lucy A. (1987) Plans and Situated Actions: the problem of human-machine communication. Cambridge, NY: Cambridge University Press. [33] Weizenbaum, Joseph (1976) Computer Power and Human Reason. San Francisco, CA: W.H. Freeman. [34] Zimbardo, Philip G. & Weber, Ann L. (1997) Psychology (2nd ed.). New York: Longman.

38

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

Immersive Environments: What Can We Learn From Commercial Computer Games? Paul R. KEARNEY School of Computing and Information Technology, Unitec New Zealand

Abstract. It is known that puzzle games, such as Tetris, enhance the player’s cognitive abilities, but today’s students opt for more action filled games such as first person shooters or role playing games. Recent research suggests that some commercial computer games increase cognitive skills and may enhance the participants ability to learn. However, it appears that it is not attributed to the gameplay of such games, but to the immersive environment created by the realism and the intense situation that the player experiences. This article suggests that these facets found in commercial games should be embraced by educators and supplement the learning environment of today’s digitally literate students. Keywords. Immersive environments, cognitive abilities, 3d worlds, games

1. Introduction According to Mulligan and Patrovsky [1], research analysts suggest that over 100 million people in the United States play commercial computer games. The China Center of Information Industry Development (CCID) estimates that there will be 48 million online gamers in China by the end of 2005 [2], and the popular gaming web site “GameSpy” maintains that the action computer game Counter-Strike is played by eight times more players than any other game [3]. It is widely accepted that educational computer games are a valuable resource for learning and commercial computer games are often viewed as mindless entertainment, but a study completed recently by researchers Green and Bavelier [4] showed other benefits are gained from computer games, such as the enhancement of peripheral vision. It has long been known that puzzle games, such as Tetris, enhance the player’s cognitive abilities. Okagaki and Frensch [5] used Tetris in their research; they found that spatial visualisation abilities were improved in college students after six hours of playing. Research done by De Lisi and Cammarano [6] showed that students improved their mental rotational skills playing a game called Block Out; a three dimensional version of Tetris. Earlier research completed by Dorval and Pepin [7] suggests that students with greater spatial visualisation abilities are generally high achievers and excel in subjects like mathematics and science. Unfortunately, today’s generation do not play Tetris or Block out. They are more captivated by action games, known as first-person shooter games, like Counter-Strike and Quake. In a survey of 25 computer game players, it was suggested that these games “not only enhanced hand-eye co-ordination, but also increased their ability to multitask.

P.R. Kearney / Immersive Environments

39

A typical first-person shooter involves controlling the player movement, aiming and firing the chosen weapon, evading being a target for other players, monitoring health status and ammunition supplies, and devising a seek and destroy strategy in order to complete the level. All this is done in unison, in a pressure situation” [8]. If action computer games do enhance basic cognitive abilities, like multitasking and hand-eye co-ordination, and increased attention span, further research could identify specific games that enhance specific abilities. The results would be of value to people who possess learning or motor skill disabilities and those involved in working with them. But is it the gameplay of such games or is it the environment created through using this technology?

2. The Question of Learning Educators need to understand the pedagogy delivered by commercial computer games and utilise them in class, and they must re-define the term “literacy”. Kress [9] states that while the spoken language will remain; the written language will evolve into a visual language in the form of images. Smith, Curtin and Newman [10] agree and suggest that the new literacy is multimedia, computer gaming, the Internet, and anything that encompasses technology. They state that instead of “the three R’s”, students need skills of analysis, evaluation and prediction at an early age. All this points to studentcentered learning using technology with educators facilitating a technology-rich environment in the future; but the future needs to be now or these students will surpass their teachers. In more recent times, Brown [11] states that learning comes as the result of a framework or environment that fosters learning rather than as a result of teaching. He maintains that today’s students look upon technology as an integral part of life and a tool that they take for granted; for many of them computing has been part of their learning since early childhood. Brown suggests that there is a shift in the way that these students learn, and this has only been embraced by a handful of tertiary institutions, creating a problem in student retention. The shifts include literacy, from text to multimedia; lectures, from teacher-centered to student-centered; reasoning, from deduction to transforming; and reading, from solitary to social exploration. Prensky [12] agrees and suggests that this decline in education is attributed to the change in the students themselves. Using the term “Digital Natives”, he states that today’s graduates have spent an average of less than 5,000 hours reading books, yet over 10,000 hours playing computer games. Because of this, Prensky concludes that these students think differently, process information differently, and get bored with traditional schooling techniques. “Digital natives” multitask, prefer multimedia over text, and thrive on computer games. Unfortunately, even today, educators still assume that students learn in traditional ways, and Prensky maintains that this must change. One perspective of educational psychology is that of information processing. Buchanan [13] notes that game designers challenge players by using “bots” or non-player characters (NPCs) in their games that mimic human behaviour. Behaviour is learnt and decisions are made by evaluating the situation and considering all the options. Bots do this by using what programmers call a decision tree. Human players do it intrinsically by monitoring the situation and manipulating it based on their own thoughts and perceived skill set. This is meta-cognition. Buchanan claims that experienced players consciously increase their mental space for visualisation and manipulation of problems. He

40

P.R. Kearney / Immersive Environments

suggests that game players possess an increased ability to multitask and mentally sort information. Buchanan concludes that computer games include all the underpinning for interactive learning but suggests that the controversy around violence related to firstperson shooter computer games is a barrier to education uptake. If the today’s digitally literate screenagers are opting out of tertiary education because the traditional delivery methods are not reaching them, environments such as MIT media laboratory’s games-to-teach project need to be encouraged. Similar research and development is also continuing at the University of Abertay’s ICCAVE, the University of Minnesota’s GRAVEL, and Unitec New Zealand’s UniCave – all games research laboratories using commercial computer games for learning. Oblinger [14] suggests that educational environments involving commercial computer games lead to deeper learning, and Buchanan [13] states that the cognitive conflict from computer games enhances learning; however, with the slow uptake of computers and computer games in education, do educators believe that students learn from games, or do games have a detrimental effect? 3. Commercial Games Versus Traditional Education Beazzant [15] suggests that commercial computer games create an environment where players are compelled to play to the extent of forming addictions. Is it this immersive environment that leads to the deep learning that Oblinger suggests, or should educators opt for the safe choice of purpose built educational games? Gee [16] suggests that many of the educational games found in today’s schools have very few learning attributes and do not deliver what is expected. But some recreational computer games, including commercial first-person shooters that digitally literate students spend much of their time at home immersed in, do possess all the learning principles. Gee [16] lists games like Rise of Nations (Microsoft), Prince of Persia (Ubi Soft), Metal Gear Solid (Konami), and even Halo (Bungi), the latest first-person shooter for Microsoft’s Xbox, as having many of the learning qualities built in. He categorises learning into three major headings: • • •

Empowered learners Problem solving Understanding

Gee suggests that there are learning facets that can be found in each of the various games; facets like the ability to customise the game, the facility to manipulate features within the virtual worlds, and well-ordered problems that challenge the player are all achievable. Gee suggests that game designers must be encouraged to continue to include such learning principles into commercial games. However, Clark [17] maintains that game designers focus on engaging the player and making the game fun to play. He states that it is the design of the interactivity that engages the player. This can be achieved at an emotional level or an intellectual level, but for the player to learn from the game, Clark states that the game design must include action and consequence; learning will then be achieved through cognitive conflict. Antonietti and Mellone [18] conclude that literature suggesting that the act of using technology on its own increases the cognitive learning abilities of the user is not justified. According to Kasvi [19], it is the game itself that promotes the learning. Kasvi lists many of the attributes in computer games that promote learning, including:

P.R. Kearney / Immersive Environments

• • • •

41

Motivation and Challenge – the addictive nature of the game Feedback and Interactivity – the experience of the environment Goals and Objectives – clear and unambiguous instructions Communication and Collaboration – co-operative tasks with individual accountability

And while not all games have all of these qualities, Kasvi suggests that multiplayer games have the best results. Kasvi notes that while attempts have been made to market computer games that have co-operative themes and no conflict, violent or otherwise, they have not been commercially successful. Kasvi concludes by suggesting that more research is required into multiplayer games to ascertain if learning takes place while collaboration is facilitated. Nova [20] used Counter-Strike (Valve Software) and Quake III (id Software) in a study of awareness tools in multi-user workspaces. Nova suggests that these multiplayer games are more successful at creating collaborative workspaces and collaborative learning environments than many GroupWare products. Nova observed and interviewed ten students involved in first-person shooter games. The results showed that much could be learnt from this environment that may be useful in collaborative GroupWare; non-verbal communication, work progress indicators, and artifact status. Cramer, Ramachandran, and Viera [21] concur that first-person shooter games create such an environment as stated above; the player is engaged to the extent of being motivated to solve strategic problems. They suggest that the world created inside a three dimensional game simulates the real world to such an extent that it is an extremely effective learning environment. They suggest that the computer “game is the cornerstone of computer based learning” (p. 5). However, they also say that while commercial games engage the player, to be effective as learning tools they must offer feedback and evaluation. America’s Army does just that, by allowing players to progress through and attain the status of “Green Beret”. But Cramer et al. suggest that America’s Army is little more than entertainment for the game player. The United States Department of Defense would disagree, with their online community web site, http://www.dodgamecommunity.com/, listing over 50 computer games currently used for training; America’s Army is top of the list.

4. Cognitive Learning from First Person Shooters A recently completed study to test the cognitive abilities of players of action computer games before and after playing such games [22], focused on commercial computer games rather than educational titles. The game Counter-Strike was chosen for this study because of its apparent immersive qualities and Quake III because of the perceived similarities to Counter-Strike. Participants were selected from various sources. Members of the International Game Developers Association (IDGA) – New Zealand Student Chapter, were requested to participate because of their extensive game experience in first-person shooter games. Students enrolled in a graduate diploma of game development also volunteered to assist in this study. Other interested parties volunteered to participate and those with no desire to be involved in the game playing, became part of the control group. In total, 38 people participated in this study, ages ranging from 14 to 60, both male and female, with varying experience in computer games. The product chosen for the assessment was SynWin from Activity Research Services.

42

P.R. Kearney / Immersive Environments

This computerized neuro-psychological assessment software creates a synthetic work environment and records increases in memory, arithmetic, visual, and auditory monitoring, and provides an overall composite score of the participant’s multitasking ability. The experiments completed by the groups playing Counter-Strike clearly showed statistically significant increases in multitasking ability. The increase appears in each of the monitoring tasks and the overall composite score. A parametric test using the data from these groups results in a p-value less than 0.05. These groups showed significant increases in scores compared to that of the control group. However, the experiment completed by participants returning after three weeks only suggests that the increase in cognitive ability is retained and does not provide conclusive results. This study also showed that playing the first-person shooter Counter-Strike appears to increase the player’s multitasking ability more than playing the other firstperson shooter computer game Quake III. The group who played Quake III between assessments showed an overall increase that was significantly less than the groups who played Counter-Strike. These two games differ in the area of realism and in the player intensity created within the multiplayer environment. Quake III is set in a fantasy environment whereas Counter-Strike simulates a real world situation. The game play of Quake III also contributes to the reduced intensity created by allowing the player to rejoin the game immediately after their character is eliminated, rather than having to wait until the end of the current round. The reduced level of intensity and player discipline also made it difficult to control the environment that the experiment was being conducted within. The immersive environment created by Counter-Strike captivated the attention of the players in this study. The participants were completely focused on the game and this concentration appeared to influence the results of the subsequent multitasking test. If, in fact, games such as Counter-Strike and America’s Army do simulate the real world, the United States military are justified in using these games for recruitment and training. In a real life combat situation, a soldier’s ability to multitask and focus on multiple applications, and to increase these abilities throughout the exercise would be imperative. The military have concluded that this does happen when playing these computer games but do not understand what characteristics within the game promote this ability. Although Quake III does not appear to create an environment conducive with increasing cognitive skills, other games such as Starcraft from Blizzard may do. Starcraft is not a first-person shooter, but is a real time strategy game that can be played in either single or multiplayer modes. Some of the participants in the above study, who scored high initial composite scores, listed the game that they prefer to play (and spend many hours playing) as StarCraft, and not a first-person shooter. When played in multiplayer mode, either via a local network or over the Internet, Starcraft also creates an immersive environment and when observing players of this game, it can be seen that it has a level of intensity similar to Counter-Strike. However, as the name suggests, Starcraft is set in a fantasy world like Quake III. Therefore it may not be the realism surrounds Counter-Strike, but the level of player intensity that the realism creates. This research has suggested that Counter-Strike creates an environment that promotes an increase in cognitive abilities and suggests that other computer games with a high level of player intensity may achieve the same results. Okagaki and Frensch [5] suggested that Tetris improved spatial visualisation abilities. Anyone who has ever played Tetris would agree that this game not only has an addictive quality but also immerses the player to such an extent that they are completely focused on the task at

P.R. Kearney / Immersive Environments

43

hand. Tetris has now been modified into a multiplayer game called Tetrinet. This game pits player against player and requires the participant to monitor multiple screens simultaneously. Tetrinet may also lead to an increase in multitasking abilities after prolonged playing. But although Tetrinet requires hand-eye co-ordination, visual monitoring skills, and multiple-screen monitoring, the skills needed to strategize and monitor sounds are minimal. A real time strategy game such as Starcraft includes all of these characteristics, especially when played in a multiplayer environment.

5. Conclusions and the Future of Learning Squire [23] contends that educators have compared computer games to the act of teaching and not embraced the cognitive learning that this platform offers. Even though Squire suggests that computer games form part of today’s student’s culture, it is unlikely that many educational institutions would accept Counter-Strike as part of their curriculum. But with Prensky suggesting that the current decline in student retention is due to student dissatisfaction with traditional teaching methods, educators may wish to rethink their strategies and revise their current delivery methods. Research indicates that younger people do in fact have increased cognitive skills (Kearney, 2005), probably accelerated through the playing of computer games. Their ability to multitask will require a rethink of traditional learning styles. Results suggested that the immersive environment of Counter-Strike does in fact enhance multitasking abilities, but this may be in response to the immersive environment that is created rather than the game content itself. Further research is required to ascertain what other cognitive abilities are improved by games similar to Counter-Strike, which other games hold the same promise, how the relevant characteristics can be easily identified and quantified, and how they can be developed into commercial computer games, marketed solely for entertainment. This article does not advocate the inclusion of the game Counter-Strike into the school curriculum, but does suggest that the immersive environments created by such games have beneficial implications for learning that warrant further study. Cramer et al. [21] agree with this and suggest that the future of learning revolves around three dimensional worlds that inherently promote learning; Counter-Strike appears to be one of those worlds.

References [1] Mulligan, J., & Patrovsky, B. (2003). Developing Online Games: An Insider’s Guide (1st ed.). Indianapolis: New Riders. [2] China Daily. (2004). Why virtual entertainment means real money. Retrieved 2oth January, 2005, from http://www.chinadaily.com.cn/en/doc/2004–01/02/content_295214.htm. [3] GameSpy. (2004). GameSpy Stats. Retrieved 15th August, 2004, from http://archive.gamespy.com/ stats. [4] Green, C., & Bavelier, D. (2003). Action video game modifies visual attention. Nature, 423, 534–537. [5] Okagaki, L. & Frensch, P.A. (1994). Effects of video game playing on measures of spatial performance: Gender Effects in Late adolescence. Journal of Applied Developmental Psychology. 15. 33–58. [6] De Lisi, R., & Cammarano, D.M. (1996). Computer experience and gender differences in undergraduate mental rotation performance. Computers in Human Behavior, 12, 351–361. [7] Dorval, M., & Pepin, M. (1986). Effect of playing a video game on a measure of spatial visualization. Perceptual Motor Skills, 62, 159–162.

44

P.R. Kearney / Immersive Environments

[8] Kearney, P. (2003). The impact of Computer Games on Children’s aggressive behaviour and learning abilities. Bulletin of Information Technology Research. 1(1), 1–8. [9] Kress, G. (2003). Literacy in the New Media Age (Literacies). Routledge: New York. [10] Smith, R., Curtin, P., & Newman, L. (1997). Kids in the kitchen: The educational implications of computer and computer games use by young children. Paper presented at the Australian Association for Research in Education Annual Conference, Brisbane, Australia. [11] Brown, J. (2002). Learning in the digital age. Paper presented at the The Internet & the University: Forum 2001. [12] Prensky, M. (2001a). Digital natives, digital immigrants. On the horizon, 9(5), 1–6. [13] Buchanan, K. (2004). How an educator thinks about computer games. Retrieved 21st December, 2004, from http://www.msu.edu/~buchan56/games/educator_thinks_games.htm. [14] Oblinger, D. (2004). The next generation of educational engagement. Journal of Interactive Media in Education – Special Issue on the Educational Semantic Web, 2004(8), 1–18. [15] Beazzant, S. (1999). Dissertation: Children and Video Games: What’s the fuss? http://www. scottbezzant.btinternet.co.uk/Downloads/Dissertation.htm. (Retrieved 12 April 2003). [16] Gee, J. (2004). Learning by design: Games as learning machines. Paper presented at the Game Developers Conference, San Jose, CA. [17] Clark, C. (2004). The principles of game based learning. Paper presented at the NETC/LSC Conference, Crystal City, VA. [18] Antonietti, A., & Mellone, R. (2003). The difference between playing games with and without the computer: A preliminary view. Journal of Psychology, 137(2), 133–144. [19] Kasvi, J. (2000). Not just fun and games – Internet games as a training medium. In P. Kymäläinen & L. Seppänen (Eds.), Cosiga – Learning With Computerised Simulation Games. (pp. 23–34): HUT: Espoo. [20] Nova, N. (2001). Awareness tools: Lessons from quake-like. Retrieved December, 2004, from tecfa.unige.ch/perso/staf/nova/awareness_games.pdf. [21] Cramer, M., Ramachandran, S., & Viera, J. (2004). Using computer games to train information warfare teams. Paper presented at the Industry/Interservice, Training, Simulation & Education Conference (I/ITSEC 2004), Orlando, Florida. [22] Kearney, P. (2005). Cognitive callisthenics: do FPS computer games enhance the player’s cognitive abilities? Paper presented at the DiGRA 2005 Changing Views: Worlds in Play International Conference, Vancouver, Canada. [23] Squire, K. (2003). Video games in education. International Journal of Intelligent Simulations and Gaming, 2(1), 49–62 [17] Clark, C. (2004). The principles of game based learning. Paper presented at the NETC/LSC Conference, Crystal City, VA.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

45

What Is a Game Ego? (or How the Embodied Mind Plays a Role in Computer Game Environments) Ulf WILHELMSSON, Ph.D., Senior Lecturer School of Humanities and Informatics, University of Skövde, Sweden e-mail: [email protected] Abstract. What is a Game Ego (or How the embodied mind plays a role in computer game environments) addresses questions concerning why and how a computer game player identifies her or himself with a Game Ego in computer game environments. The theoretical framework used to address these questions is drawn mainly from three fields of research: film theory (including theories on narration and narratives) theories on visual perception (which are also applicable to sound) and finally experientialist cognitive theory. The central claims of this paper are: the process of identification with a manifestation of a Game Ego has a bodily basis. The Game Ego is primarily a bodily based function that enacts a point of being within the game environment through a tactile motor/kinesthetic link. The human conceptual system shows a relationship to the motor system of the human body and is tightly connected to the emotional system so that no clear-cut boundary can be drawn between them. The more direct and immediate the control of this agent is, the stronger the identification is as well.

1. What Is a Game Ego? (or How the Embodied Mind Plays a Role in Computer Game Environments) I will begin this article by asking some questions that at some point might have haunted you while you where playing computer or video games at late night sessions crumbed over your keyboard desperately trying to get past level 12 in Pac-Man [1]. “Why do I identify with a yellow circular shape on my computer screen and feel that shape being a part of me when I play a game of Pac-Man? Why do I engage in the game play process to such a degree that when this yellow thing gets crammed into a corner with no way to run and no way to hide, I feel severe distress and call for help loud and clear and wake my kids in the middle of the night? Why is it that when I use that yellow little thing and its chewing mouth in order to eat other shapes on that very same computer screen I feel so invincible and self assure that I feel I could take on the world? And why do I think of the shape as having a mouth and the act of eating in the first place? And how come I am not bothered by the sound of the game while my girlfriend stuff her ears full with pieces of a newspaper and hide her head underneath the pillow?” In the following I will try to provide some possible explanations to the above questions elaborating on my concept of the Game Ego [2]. There are of course several ways and methods that will provide a number of possible explanations. We could for example use a method with its roots in semiotics and thereby describe the game as a sign

46

U. Wilhelmsson / What Is a Game Ego?

system with codes that needs decoding by the game player cf. [3]. We could use formal logic to some extent and thereby understand a game as a special and more formal mode of play with a set of rules cf. [4]. We could also use methods from the field of psychoanalysis and describe the process of game play in terms of the unconscious, the other self and so on cf. [5]. I will take a different path in this attempt to find the answers though. I advocate the idea that narratives may be thought of as enacted rather than told. In summary there are three main research fields here that make up the theoretical stamina for the study: the field of film theory (including theories on narration and narratives), the field of cognitive theory and the field concerned with the psychology of visual perception (however the particular theory of interest for this study is applicable to a much broader field of research than visual perception only). Before we commence the actual quest of finding answers or explanations let us briefly examine the basic content and relevance of these fields of research and identify the scholars that will be of particular interest in this context. We will start out with film theory and the work of Hugo Münsterberg [6], David Bordwell [7], and Edward Branigan [8,9]. This way leads to a section on point of view and point of being. In that discussion there is a short comment on James Gibson’s ecological approach to visual perception [10,11]. Finally, before starting the main discussion on the Game Ego, we will briefly encounter Lakoff and Johnson’s experientialist theory of cognition [12–17] in that very same context i.e. the point of view and the point of being discussion. Since these paths cross each other the discussion to follow will actually forego this succession slightly at critical points. This is done in purpose of clarifying the connections/relations between the theories. Then follows the main discussion on the Game Ego and we reach the end of this path, the final point at the conclusions. Deposit quarter. Here we go. 1.1. Some Words on Film Theory (or Backdrop for My Hypothesis) Film theory is a broad and interdisciplinary bundle of methods and competing theories that is not easily described as a whole congruent system. In some cases the sole bound between theories are what they try to provide answers to i.e. questions concerned with film. For the present work the following film scholars are the most important: Hugo Münsterberg, David Bordwell and Edward Branigan. They are scholars focused on the cognitive capabilities of the human mind. I suggest that a reason for using a cognitive approach in film theory (and of course also in the study of computer and video games!) would be that it can provide alternative answers to questions addressed by for instance psychoanalytic theories i.e. questions on subjectivity, identity, interaction, the meother-relation, the foundation for human concepts etc. Hugo Münsterberg was a forerunner to film theory with this particular focus even if he is often placed in the category of psychology scholars interested in film. In Münsterberg the world found an early (if not even the first) attempt to a scientific study of film and especially the human mind’s role in construing the narrative from the filmic events displayed on the screen. In his 1916 piece [6] he did put forth the idea that the spectator was the most interesting part for a scientific study since the narration of a movie is an internal mental process. In other words the role of the director of a movie was not as interesting to Münsterberg as were the spectator and the human mind. Münsterberg’s aim was to explore what goes on in the human mind while taking part in a photoplay (which was the term he used for narrative film). There are rather strong connections to be made between the theories on film put forth by Joseph D Anderson [18], David Bordwell [7], Edward Branigan [8,9],

U. Wilhelmsson / What Is a Game Ego?

47

Torben Grodal [19] and as well as other scholars interested in the human mind’s associative capabilities such as Vannevar Bush [20].1 As Joseph D. Anderson notes, Münsterberg recognized that “the motion picture is structured in a way that is analogous to structuring processes of the mind” [18] p. 4. See also [2] p. 21. The material of movies is the human mind and its narrative capabilities. This is at the core of the theories put forth by Münsterberg, Bordwell, Branigan, Anderson and Grodal. The human mind in Münsterberg’s understanding has several levels of organization where the higher ones depend on the lower ones. Film has its basis not in technology but in the human mind. Cf. [24] pp. 18–20. Much like the Gestalt Psychologists who he preceded, Münsterberg argued that all experiences are PART–WHOLE structures and that attention is an important factor in organizing the perceptual field. Münsterberg used what has later become classical examples of figure and ground reversal in order to show how we willingly are able to shift our attention towards a stimulus in our perceptual field and make the figure become the ground. Just to mention one of his examples, the movement of a cloud in front of the moon, can be reversed as to become the movement of the moon behind the clouds [24] pp. 16–17. This provides a link to the work of Lakoff and Johnson. This is one of the critical points on this path where there is an intersection between theories why we have to make a short stop here and study the terrain before moving on. Lakoff and Johnson’s experientialist approach to cognition acknowledge that: Rational thought is the application of very general cognitive processes – focusing, scanning, superimpositioning, figure-ground reversal, schema etc. [to certain well structured aspects of bodily and interactional experience to abstract conceptual structures]. [15] p. 121. (My italics). So it is noticeable that Münsterberg’s work relates to the experientialist theory of cognition. Both theories acknowledge that such primary levels of the human mind incorporate very important and fundamental functions in structuring our experiences. Cf. [2] chapters 2–5. So far we have recognized that the human mind will be one part of this study and that there are a number of film scholars that are important in how we will position the discourse. The main discussion of the present work is to elaborate on the concept of a Game Ego. A Game Ego is a bodily-based function that provides and allows a game player a function of agency within a game environment and hence is one source for subjectivity and identification within the game environment. There are theories within film studies that are concerned with questions of identification and subjectivity. The next section will therefore make some brief and summary comments of such traits and comments on their specific relevance in the context of exploring the Game Ego function in computer and video games. To begin with I will keep the discussion within the discourse of David Bordwell and Edward Branigan respectively.

1

Vannevar Bush’s famous article “As we may think” (1945) [20] plays a great role in how computers are used in our daily life today in the beginning of the 21st century. Bush was interested in how we create associative structures in our search for information about specific topics. He envisioned a technical device that he called “Memex” that in turn has inspired the work of Nicholas Negroponte, Douglas Engelbart and Ted Nelson in their attempts to build user-friendly computer systems. The associative function of the human mind is a common interest for Münsterberg and Bush and also for Bush’s followers. What Bush describes in his article is also interesting for the study of computer games, especially games that involve strategic thinking based on associative structures in a narrative presentation such as Day of the Tentacle [21], Sam & Max Hit the Road [22], Full Throttle [23] and the likes (all from LucasArts). (Cf. [2] pp. 48 and 61).

48

U. Wilhelmsson / What Is a Game Ego?

1.2. Point of View Versus Point of Being (or How Terminology Confuses Things) One central concept within film theory when discussing identity, identification and subjectivity is the concept of point of view. However this concept is problematic to use when studying computer games. In short: point of view is a sense-focused term and as such it often leads to theories that do not take into account other sensations of the human mind and perceptual system that might be the result of synesthesia.2 In some studies point of view designate the visual and auditory vantage point of a character without a clear distinction between these two. This is the case in the work of David Bordwell and Edward Branigan respectively. The most stunning example of this is actually when Branigan attempts to be distinctive and comes up with the term auralpoint of view [9] p. 186. Cf. [2] pp. 39–40. Furthermore there is a confusion concerning whether point of view is the so called optical point of view of a character (Bordwell) or if the term shall be defined as a way of reading the text (Branigan). David Bordwell’s approach is a camera analogy that is body based [7] p. 60. We incorporate the camera and the camera is a character within the diegetic space of the narrative. We occupy a point in space that is equal or close to the point of location of a character in the film (or in the game to keep it within the context of the present work). Branigan’s metaphorical point of view is significantly wider and tends to confuse a more literal conception (i.e. Bordwell’s approach) with the logic of reading i.e. the overall organizing structures of the narrative as well as the human mind’s capacities for understanding and making sense of the narrative [8]. The human mind is for Branigan and Bordwell what actually construes the narrative. The human mind is the material for narratives just as Hugo Münsterberg claimed. The point of view designates a point in space from which the environment might be observed. Observed shall in this context mostly be understood as “something is seen”. And seeing means that the one who sees is situated in the environment or simply that he or she is. Seeing equals observation and observation equals being. To express this in a more logical format: Seeing = Observation = Being gives Seeing = Being James J. Gibson however elaborates on what he calls “points of observation” in his theory on ecological visual perception. An observer, in Gibson’s understanding, is not only a spectator but also a subject using all the sense modalities of the human mind. Cf. [11] the discussion on Medium, pp. 16–19. The use of Gibson’s point of observation will make the logical formula different: Perception = Observation = Being gives Perception = Being 2 It seems reasonable to adhere to the idea that the language we use also affect the way we act and think and that language itself is one way of studying the human thought and conceptual system as Lakoff and Johnson suggest in their studies of cognitive semantics [12,13,2]. See also [25] the introduction. Using “view” as a broad concept for understanding something might result in that people, especially film scholars, get locked inside a visual language and system which also excludes other sense modalities.

U. Wilhelmsson / What Is a Game Ego?

49

When I use the concept Point of Being I take into account not only vision but all other sense modalities available to human beings. Our motor capacities play a great role in the manifestation of being. Lakoff and Johnson’s theory on experientialist cognition (cognitive semantics) shows that our conceptual system has its foundation in the use of the human body (i.e. motor action) and that we live within societies. This matches Gibson’s theory in that Gibson acknowledges motor performance in the form of visual kinesthesis to be a fundamental factor for visual perception. Visual kinesthesis combines body action and perception, while stressing the locomotion of the body as a factor in perceptual processes. If we combine this with Lakoff and Johnson’s theory we have a framework that merges perception with cognition at the motor level of the human mind. Accordingly, what I suggest is a way of thinking about the issue of subjectivity, identity, and identification at a bodily basis. The concept of a Game Ego is one step on such a line of thought where motor performance and cognitive capabilities add up to a Point of Being. In addition, and as a consequence of using James Gibson’s theory of affordances [10] computer games are better understood as interactable than interactive. This terminology highlights the “affordance” of the media. Interactable means that the media affords interaction – and therefore also involves the user in a higher degree than the term interactive would suggest. Gibson reduced an environment to consist of three entities: substance, medium and a surface that separates the two former from each other [11]. These three entities prove to be useful in the study of computer and video games. What we have here is the foundation for a model of understanding computer games as consisting of a primary surface for actions taken. Gibson notices that it is at the surface of the substances that actions are performed. The surface is the interface that not separates but connects the user and the computer so as to make the user and computer a system in a part/whole configuration. 1.3. The Embodied Mind and the Conceptual System (or the Bodily Basis for Understanding and Meaning) The basic statement of experientialist cognition is that the human mind is embodied and that this also has its implications on our human conceptual system. Lakoff and Johnson’s metaphor theory states that the languages we use show traces of motor activities in the way that language describes relations between objects, which can be metaphorically understood as human relations, relations in space and time etc. Lakoff and Johnson propose that the concepts we use in understanding the world are in fact embodied. An embodied concept is a neural structure that is actually part of, or makes use of, the sensorimotor system of our brains. Much of conceptual inference is, therefore, sensorimotor inference. If concepts are, as we believe, embodied in this strong sense, the philosophical consequences are enormous. The locus of reason (conceptual inference) would be the same as the locus of perception and motor control, which are bodily functions. [13] p. 20. (Their italics). Human beings make the world meaningful by metaphor and metonymy. Our ordinary conceptual system, in terms of which we think and act, is fundamentally metaphorical in nature [12] p. 3.

50

U. Wilhelmsson / What Is a Game Ego?

Equally important for experientialism is the recognition of basic level concepts and the image schema structures that preconceptually organizes and structure the human mind.3 1.

2.

Basic level concepts. This is a result of findings in theories on categorization. Categories are not firm and rigid things in which all category members share the same features. For instance there are prototypical examples of category members. This indicates that there is a basic level of categorization. Image schemas i.e. schematic structures of containment and containers, paths, links, part-whole schemas, force dynamics etc. These schemas have a nonfinitary internal structure. Such schematic structures play a great role in structuring our embodied mind. For instance the container schema has its basis in our bodily containment. To put it in a very simplified way: the surface of our skin is separating us from the rest of the world.4

We are physical beings, bounded by and set off from the rest of the world by the surface of our skins, and we experience the rest of the world as outside us. Each of us is a container with a bounding surface and an in-out orientation. [12] p. 29. This describes the physical basis for a computer game situation and is of importance in understanding computer games. This schema is a structure from which we understand the world as relations between our body as a container and objects outside this container. These objects can also be thought of as containers [15] p. 141. That is: the body is a container that could be inside or outside other containers. Consider the quote from Lakoff and Johnson once again: We are physical beings, bounded by and set off from the rest of the world by the surface of our skins, and we experience the rest of the world as outside us. Each of us is a container with a bounding surface and an in-out orientation. [12] p. 29. The structural elements of this schema are Interior, BOUNDARY, and Exterior. When studying computer games we could have a container schema of the following properties: Bodily experience: the game player has a character to control. The game player is outside (exterior) the computer controlling actions inside (interior) the computer. The game player’s movements extend into (interior) the computer via the control device (the boundary). These outside actions then loops back to the outside via the interface through various channels (boundaries). The game player could also be thought of as being inside the game environment since he or she is able to control and manipulate characters or objects within this world. The game player is inside the game environment and is manifested by a Game Ego i.e. an agent within the game. To execute these actions the player needs a connection to the game environment. A link schema can describe this. The structural elements of a link schema are two elements, A and B, and a link that connects them. Social relations and interpersonal relationships are often understood as links. Making connections and breaking social ties are examples of this [15] p. 143. 3 The use of the word image generate yet another language problem much in the same line up as point of view. Mark Johnson explicitly comments on this however and notice that this term is not something that only has to do with the visual sense but that “It would seem that image schemata transcend any specific sense modality, though they involve operations that are analogous to spatial manipulation, orientation, and movement.” [17] p. 25. 4 Gibson’s discussion on substance, medium and the surface that separate these two from each other have the same basic outline as the container schema. See [11].

U. Wilhelmsson / What Is a Game Ego?

51

To summarize and elaborate: in a computer game we can have some object or objects that we are in control of as game players. We are agents acting upon them. There is a link between the game player and the controllable object. This means that there is a social and psychological link to this object that rest on motor activity. This link is often so strong that the object in control ceases to be understood as an external object to the game player but rather is understood as an integral part of him or her while playing the game. This is what I call the tactile motor/kinesthetic link. This link is the relation between perception, cognition and action. It is the foundation for my model of a Game Ego. 1.4. The Game Ego (or the Embodied Mind and Its Manifestations Through Enactment) The Game Ego is not by necessity something that has a visual presence or manifestation as such but something that might exert force.5 It need therefore not in itself be visible, or audible for that matter, but the result of its actions need to be manifest in one way or another through the tactile motor/kinesthetic link or through visual or auditory channels. There is a strong relation between the tactile motor/kinesthetic link and the Game Ego. The link is what connects the game player’s action to the Game Ego. The Game Ego is the agent within the game environment. This link is an extension of the motor capacities of the game player and the Game Ego performs the actions in the game environment. This also goes the other way around. The manifestation of the Game Ego’s actions or the game environment’s exertion of force upon the manifestation of the Game Ego might be communicated to the game player through the tactile motor/kinesthetic link. Examples hereof are “force feedback” or “dual shock” depending on which game system we are discussing.6 This motor/kinesthetic feedback of force is one possible enhancement of the experience of presence in the game environment. The use of sound might be another strategy to achieve this, as we shall soon get back to. To temporarily leave the discussion on the container and the link schemas, we can think of them in this way: the link makes the boundary between the inside and the outside of the container transparent and unnoticed by the game player when the system performs well, i.e. without glitches of any kind, and the response to actions taken is conceived as being of a mechanical and immediate nature to the game player. One important factor to achieve this mechanical and immediate sensation of response is the duration between an action performed and it to show a result via the interface. To keep it brief, would you not be surprised if, when you try to drop something to the ground, a stone or something that normally would drop with a specific velocity, stayed in the air a long time after you have let it go and then maybe only slowly start to descend? Since many computer games are built on the idea of immediate response to actions performed, the boundary will get conceptually protruding if the bound between 5

This is a basic discourse in Mark Johnson’s conception of interaction. In [17] he goes from an argument on how complex structures of meaning is generated by image schematic structures of experience and arrives at “a second ever present dimension of our experience, that of forceful interaction” [17] p. 41. Forceful interaction supplements the analyses of bounded ness and the container schemas with considerations of motion, directedness of action, degree of intensity and structure of causal interaction. Johnson uses the term interaction in explaining the pre-conceptual gestalt structures characteristics for force. [17] pp. 41–43. I.e. in his words “their nature as coherent, meaningful, unified wholes within our experience and cognition” [17] p. 41. 6 On the Windows XP platform “force feedback” is maybe the most well known term for this. “Dual Shock” is a trademark of Sony and is associated with the Sony PlayStation system.

52

U. Wilhelmsson / What Is a Game Ego?

the game player and the game is distorted.7 And even if the game is not construed on a basis of direct motor control but on a point and click concept, the delay between a chosen action and its result must not have a too long duration. When playing a computer game the game player is goal-oriented and in order to reach the goal he or she uses the Game Ego manifestation within the game environment. If then this agent does not respond to the goal-oriented strategy of the game player and his or her motor input, a feeling of loss of self-control is probably beginning to thrive in the player’s mind. When this happens the address towards the Game Ego is often no longer “I” but “he, she or it” (or something worse better not mentioned). This also connects to “the split and bifurcated person” that Lakoff and Johnson describe with the Subject/Self metaphor system. [13] chapter 13 and [2] chapter 8, The Game Ego. In a simplified model of what Lakoff and Johnson are arguing the Self is a person’s own body and the Subject is that quality of a person that is able to reason about the world: “The Self acts in the world.” [16] p. 110. A projection of the Subject quality of the game player onto a Game Ego seems to be necessary in the process of playing computer games. The Game Ego is another Self and/or an extension of the game player’s own Self. Playing computer games is a process of subjective performative interaction. It is also a subjective experience. The mental space for this projection, the possible space of the game environment, is where the Game Ego resides and acts. 1.5. Step Forward Please (or the Front Is in the Front) Our visual system is able to focus our attention in the direction we call forward. Forward is a meaningful concept in regard to our bodily constitution. Our eyes, our nose, our mouth and our ears aimed and all attuned in the same direction. Also our limbs are positioned in a way that encourages movement in that very same direction. Most of our bodily attention and radius of action literally lies in front of our feet, legs, torso, arms, hands, neck, head and mind. No wonder then that many computer and video games feature exactly this. This is a result of our bodily constitution and in turn this also gives a conceptual structuring that provides an experiential bias for associating forward with something good and positive. Cf. [12] p. 132 and [2] pp. 82–83 and 200–204. Proprioception is yet another factor that plays a role in the construction of a Game Ego. In the processes of manipulation and controlling a Game Ego’s visual manifestation on the screen e.g. in the form of Lara Croft [26] or one of the Mario Brothers, [27] the game player enacts a motion in the game environment by pressing and moving his or her fingers. The important issue here is that the game player might associate one specific finger/keyboard configuration not at all with the visual Game Ego’s hand or finger but other parts of the visual manifestation and its functionality for any specific purpose. This is a process of achieved reflexes and part of the game play to learn and train to perfection. If we combine the frontward orientation with this tendency to associate specific actions with specific keyboard/finger configurations what do we get? I suggest that we get the exertion of distributed locomotion. The tactile motor/kinesthetic link between game player and Game Ego is a distribution link. It distributes the possibility of locomotion. The most probable way for a human to move inside an environment is in the same direction as vision directs the human in motion. This is due to that humans use 7

See also [2] p. 126.

U. Wilhelmsson / What Is a Game Ego?

53

vision to avoid bumping in to other substances in the form of other humans, rocks, cars, cats, dogs, pillars or vendor machines etc. This is of course not the exclusive use humans or other living beings make of vision. Vision can also be used to localize food, water, mating partners and other necessities of life. Vision is one of the life supporting systems. If the game interface is based on the idea to let the game player see (one mode of observation) the environment through the eyes of a Game Ego manifestation he or she might understandably identify the situation as something that lies close to the normal way of perceiving the environment. Most probably the game player will move the Game Ego in his or her frontal direction if possible since this in most cases is functionally easier to perform than moving backwards or sideways. It is also conceptually more difficult to move backwards or sideways. It is all too easy to rush ahead into the line of fire in a session of Facing Worlds in Unreal Tournament [28] instead of adopting the much wiser tactic of moving sideways, backwards and jumping. Probably this is due to the idea that it is conceptually easier to move in the frontward direction since one can see the blocking obstacles of different kinds (enemies with BFGs, laser weapons, squares and dots, cars, policemen, old ladies, mummies and zombies, trees, etc.). Even if the game interface does not rely on an inhabited first person visual manifestation the very same tactics are often adopted. In some games, like in Spyro the Dragon [29], the manifestation of the Game Ego is seen from behind and this is very close to see through the same visual perspective as the character. However, if a game player observes the manifestation of the Game Ego from above, below (which is probably very rare, I can not come up with any example hereof) or from one of its sides, the game player will probably construe a front-back orientation for the Game Ego and name one of the directions as frontward and its opposite backward or reverse. This is due to humans wanting objects within the visual field to have this orientation. As Lakoff and Johnson point out: Moving objects generally receive a FRONT–BACK orientation so that the front is in the direction of motion or in the canonical direction of motion. [12] p. 42. The front-back orientation is very strong. A car for example retains its front in the front even when it is going in reverse. The physiognomy of a concrete visual manifestation of a Game Ego has its importance for the game play. The physiognomy has consequences for what actions could be performed and how they will be performed. The front-back orientation is one factor. Our perceptual capabilities as well as our cognitive basis in motor performance, transcend into how the Game Ego can be manifest. A Game Ego with limbs of a human kind may for instance afford manipulation that is close to how a human actually perform certain kinds of manipulation in an environment. That is, a fully visible Game Ego that is anthropomorphic (i.e. in the shape of a human being) will probably be able to mimic how human beings do things. If we take this one step further, this could affect the setup of the game interface. And it does. There are games that use the game player’s body to greater extent than pressing keys or manipulating joysticks, steering wheels etc. This is of course the case in simulators of different kinds where there might be some kind of foot control and such devices. The motion of the legs and feet is used to control the Game Ego function. A more advanced technology is the Sony EyeToy [30] where you actually see yourself as a Game Ego on the screen and you use your whole body to control the game. Having discussed motion and its concrete visual manifestation it is time to turn the attention to motion of a somewhat different kind, namely sound.

54

U. Wilhelmsson / What Is a Game Ego?

1.6. A Texturized Wall of Sound (or Is There Some Kind of Way Out of Here?) Since I have made critical remarks about scholars providing only half theories that are very focused on vision I better say something about the auditory system and its relevance to the idea of a Game Ego function in computer and video games. Sound is also a means to construe a Game Ego manifestation. Sound has functionality both as the result of game player agency within the game environment and the information that makes a game player take action in the first place. Furthermore, an objective of a specific game could very well be based on sound manipulation rather than image manipulation. I argue that not only visual objects have substance and surface but that sound also has these properties. Let me exemplify this with the expression “a dense” sound. Sound is clearly considered as an object within the environment since it can obviously have a conceptual density. Such a sound can be understood and expressed as a wall of sound.8 That kind of inference means that we have interpreted the sound as having a rigid structure and some kind of texture as well.9 This connects to Gibson’s theory of affordances and constraints. The affordances of an environment are what it offers an animal/living organism [10] p. 68. A constraint is what hinders a specific action to be performed. A door is an opening in a wall and affords locomotion through that wall. The door is walk-through-able to use a Gibsonian language. The wall blocks locomotion and constitutes a constraint for locomotion in a specific direction. It might however have other affordances such as being paint-able, break-able etc. In this context however let us settle with the binary situation of walk-trough-able and blocking. The use of sound can have the same function in a game. A sound can afford locomotion in a specific direction and block another one. The use of surround sound might be one way of enhancing this phenomenon. A very “dense” sound from one of the speakers might block a direction while the sound from another speaker has a “lighter” quality that affords locomotion in that specific direction. Sound and not only concrete imagery can have affordances and constraints considering locomotion. And not only locomotion in the meaning of getting from one specific point to another but also the manipulation of the game environment can have its origin in the sound and the auditory sense system rather than in the concrete images displayed on a screen. Hearing a sound may result in an internal visual process that in turn may have a result in concrete motor action. I also argue that sound is possible to categorize in a system of basic level concepts in accordance with the experientialist theory of cognition. We have so far only briefly touched upon this part of Lakoff and Johnson’s theory. Drawing on the work of Eleanor Rosch, they argue that the human mind categorizes the world around a central basic level of categories. This level is where we have most of or knowledge about the world and the objects in the world. Cf. [14] p. 13, [13] pp. 27–28 and [2] chapters 5–10. An example of this is the concept of chair. It can be ordered in the following sequence of levels from general to specific:

8 “A wall of sound” is precisely how the producer Phil Spector thought of sound when producing music for the Tamla/Motown record label in the 1960’s. 9 This is something that would be interesting to study further; the relation between real sound and mental imagery. The processes of internal visualization would be most interesting to study more and also to do empirical neurological work in this field.

U. Wilhelmsson / What Is a Game Ego?

55

General level: Furniture Basic level: Chair Specific level: Red chair with white dots The basic level captures the concept chair very well in one single mental image of that object we call chair. This also connects to the affordance of the object. The basic level is the highest level where humans use similar motor actions to interact with category members. Lets us change the general level of furniture to something else: direction General level: Directionality Basic level: Up Specific level: Up to the sky How does this connect to sound? To envision a direction is quite easy. We see objects around us and they have a specific spatial position. They can for instance be to the right or left, behind or in front of us. I argue that we do make meaning from sound in directional terms as well. Sound in itself is a perceptual/cognitive process stemming from information that has its origin in motion of an object in an environment. Sound is motion and is also often conceptualized as such. We can say “a rising pitch” and by that we mean a sound that goes from a low point to a higher point. We use a visual terminology to make meaning from this sound. The understanding of “a rising pitch” is also a bodily tension. If you think about it you will probably experience a slight tension in your muscles when you project such a rising pitch. And the opposite as well: a falling pitch is a relaxation of the muscles. To raise something takes effort and your body “knows” this. It is part of your embodied mind and your conceptual system. 1.7. This Is a Game Ego (Answers to the Initial Questions and My Conclusions) I started out by posing a number of questions. Through out this paper I have tried to answer them but to avoid any misunderstandings I will summon the answers in this last paragraph which are my conclusions. Questions 1–4: “Why do I identify with a yellow circular shape on my computer screen and feel that shape being a part of me when I play a game of Pac-Man? Why do I engage in the game play process to such a degree that when this yellow thing gets crammed into a corner with no way to run and no way to hide, I feel severe distress and call for help loud and clear and wake my kids in the middle of the night? Why is it that when I use that yellow little thing and its chewing mouth in order to eat other shapes on that very same computer screen I feel so invincible and self assure that I feel I could take on the world?” Conclusion: The process of identification with a manifestation of a Game Ego has a bodily basis. I understand the Game Ego as a bodily based function that enacts a point of being within the game environment through a tactile motor/kinesthetic link. The more direct and immediate the control of this agent is, the stronger the identification is as well. As a player I incorporate an agent, a Game Ego function, within the game environment. In Pac-Man the visual manifestation of the Game Ego is a yellow circular shape that is controllable. It is up to me as a game player

56

U. Wilhelmsson / What Is a Game Ego?

to direct its motion. This exertion of control is an extension of my own sensory motor system via a tactile motor/kinesthetic link, why it is not only the controlled and perceived motion on a screen but also the experience of locomotion that is the result of this control. Hence the third person address of the Game Ego turns into the first person of me as a game player. I am there and there is not only there but also here so to speak. The distress I feel as a game player when things are not going to well for my Game Ego is due to this strong identification. It is a part of me that is acting within the game environment: a motor part and an extension of my sensory motor system. Sigmund Freud might have called this castration anxiety but I prefer to call it a very natural response to an equally concrete emotion. And when things go my way in the game and my Game Ego is successful there is no wonder that I feel good and invincible for the very same reason. The human conceptual system shows a relationship to the motor system of the human body and is tightly connected to the emotional system so that no clear-cut boundary can be drawn between them. The locus of reason is also the locus of experience [13] p. 20 [2] p. 72. Questions 5–6: “And why do I think of the shape as having a mouth and the act of eating in the first place? And how come I am not bothered by the sound of the game while my girlfriend stuffs her ears with newspaper and hide her head under the pillow?” Conclusion: Even Pac-Man is in a way anthropomorphic. Pac-Man has a trait in common to me in that it also has a thing in the direction of motion that seems to be able to consume other things. I call this thing a mouth as I call my mouth a mouth. I have my mouth in my front i.e. in my preferred direction of locomotion. And not only so but I have my mouth in my face which in turn I have at the front of my head. My head is more round than square. Pac-Man is more round than square. In this way we are alike. Of course these maneuvers are actions of basic level of categorization. At the basic level there are many similarities to be drawn between my looks and the looks of PacMan. Pac-Man moves and eats and tries not to be eaten in the attraction mode of the game. When I play the game it is my task to perform these actions or to control the motion that Pac-Man already posses, since this Game Ego moves automatically when the game begins. To answer the last question: “… how come I am not bothered by the sound of the game while my girlfriend stuff her ears full with pieces of a newspaper and hide her head underneath the pillow?” I am not bothered by the sound since it is there for my sake and it helps me to maneuver the game. Of course people can be very maddened by the sound of Pac-Man. It might be conceived as too loud and too much of blip blip blopping, lacking structure, having harsh timbre and being only painfully electronic to its nature. On the other hand, the sound of Pac-Man is very meaningful. It has clear connections, links, to what actually is going on in the game. Moreover, I am not only a perceiver of the soundscape when I play the game, I am also in a way the producer of it. Without my actions there would be no sound generated. My actions do not only have a result in a change in the visual flow of the game but in the auditive flow as well. I am an integral part of the game as a system. My Game Ego is my agent that allows me to take action upon the visual or auditive information from the game. This information is meaningful to me as a game player in a higher degree than any bystander, especially someone that does not know the game. The sound functions on a basic level of categorization as well as on a specific level.

U. Wilhelmsson / What Is a Game Ego?

57

1.8. Game Over? (or Can the System Be Rebooted?) In the above I have suggested a framework for understanding computer games from diverse fields of research: film theory, theories of visual perception and experientialist cognitive theory. This does of course not mean that all questions are asked or even that the answers suggested by me are the only ones of interest. The questions are not even the only ones of interest. The expanding field of computer game studies is here to stay, reform and reboot several times. The game is by no means over; in fact it has just begun. Insert another quarter. Game loading.

References [1] Pac-Man (Namco 1980). [2] Wilhelmsson, Ulf, Enacting the Point of Being. Computer Games, Intaraction and Film Theory. Ph.d dissertation. University of Copenhagen, 2001. [3] Mayer, Paul, A Social Semiotic approach to the analysis of computer media. Ph.d dissertation. University of Copenhagen, 1999. [4] Alexander, John, Screen Play – Audiovisual Narrative and Viewer Interaction. Ph.d dissertation. Stockholm University, 1999. [5] Turkle, Sherry, Life on the screen: identity in the age of the Internet. Simon & Schuster, New York, 1995. [6] Münsterberg, Hugo, The Photoplay: A Psychological Study. D. Appleton, New York, 1916. [7] Bordwell, David, Narration in the Fiction Film. Routledge, London, 1985. [8] Branigan, Edward, Point of View in the Cinema: A Theory of Narration and Subjectivity in Classical Film. Mouton Publishers, New York, 1984. [9] Branigan, Edward, Narrative Comprehension and Film. Routledge, London/New York, 1992. [10] Gibson. James J. ‘The Theory of Affordances’. In Shaw and Bransford. Perceiving, Acting and Knowing. Hillsdale, LEA, New Jersey 1977. [11] Gibson. James J. The Ecological Approach to Visual Perception. LEA, New Jersey, London, 1986. [12] Lakoff, George, Johnson, Mark, Metaphors We Live By. University of Chicago Press, Chicago, 1980. [13] Lakoff, George, Johnson, Mark, Philosophy in the Flesh Basic Books, New York, 1999. [14] Lakoff, George, Women, Fire and Dangerous Things, University of Chicago Press, Chicago, 1987. [15] Lakoff, George, ‘Cognitive semantics. Two views on cognition’. In Eco, U., Santambrogio, M. and Violi, eds. Meaning and Mental Representations 1987. [16] Lakoff, George, ‘Sorry I’m Not Myself Today’. In Fauconnier and Sweetser eds. Spaces, Worlds and Grammar, University of Chicago Press, Chicago, 1996. [17] Johnson, Mark, The Body in the Mind, University of Chicago Press, Chicago, 1990. [18] Anderson, Joseph D. The Reality of Illusion: An Ecological Approach to Cognitive Film Theory. Southern Illinois University Press, Carbondale, 1996. [19] Grodal, Torben, Moving Pictures: A New Theory of Film Genres, Feelings, and Cognition. Clarendon Press, Oxford, 1999. [20] Bush, Vannevar, ‘As We May Think’. In The Atlantic Monthly July 1945. Downloaded version of the text prepared by Denys Duchier, at the University of Ottawa, Canada in April 1994 and later updated 1995. http://www.isg.sfu.ca/~duchier/misc/vbush/. [21] Day of the Tentacle (LucasArts 1993). [22] Sam & Max Hit the Road (LucasArts 1993). [23] Full Throttle (LucasArts 1994). [24] Andrew, James Dudley, The Major Film Theories: An Introduction. Oxford University Press, Oxford, New York, London, 1976. [25] Jay, Martin, Downcast eyes: the denigration of vision in twentieth-century French thought. University of California, 1994. [26] Super Mario Bros. (Nintendo 1986). [27] Tomb Raider (EIDOS 1997).

58

U. Wilhelmsson / What Is a Game Ego?

[28] Unreal Tournament (Epic Mega Games/GT Interactive 1999). [29] Spyro the Dragon (Insomniac Games 1998). [30] EyeToy: Play (Sony 2003).

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

59

Multiple Motivations Framework Hakan TÜZÜN, Ph.D. Computer Education and Instructional Technology, Hacettepe University, Turkey [email protected]

Any theory of motivation must consider a large set of interactive processes if it is to provide an adequate explanation of human behavior. (Bandura, 1986, p. 243)

1. Introduction It is 4 a.m. in the morning. A pair of bloodshot eyes stares at the TV monitor, which is connected to a Commodore 64 personal computer. The owner of the eyes is very exhausted; he has been trying to finish a computer game called “Henry’s House” for the past 48 hours. While doing so his hands are almost integrated with the joystick. Although he is hungry and sleepy he refuses to leave the scene until he finishes the game. He has the task of organizing Henry’s messy house room by room. He is currently in the cellar, the eighth and final room. Until finishing the game he will have no peace on earth; everything beyond Henry’s House lies beyond his horizon of interest or concern. The person in the previous paragraph could be one of millions of children caught up in today’s video game dominant world, but this specific instance is based upon the author’s personal experience. Computers and especially computer games have been a major part of my life, since the age of ten. Much water has passed under the bridge and computer game technologies have grown exponentially since then. I eventually finished Henry’s House, the Commodore became obsolete, new technologies have been invented for playing computer games, and the video game revenues surpassed movie box office revenues in the U.S. (Greenspan, 2002). Video games have gone so mainstream that Wal-Mart, the top retailer in the world, has a “Video Games” section in its online store among its other big sections. Recently Personal Computers are being used for gaming in addition to solving many problems of life, but there have also been dedicated systems, called game consoles, for playing video games. Among those, Sony with its PlayStation 2, Nintendo with its GameCube, Microsoft with its Xbox (and recently with its Xbox 360), and Sega with its DreamCast are the reigning technologies in the gaming world of today. The transformation is still in progress; the diffusion of the Internet in the 1990’s added the multiplayer element to video games. Meanwhile, my gaming adventures have continued with these new technologies, and as time passed I eventually became a grown up. In the beginning of 2002, I joined the Quest Atlantis (QA) project, an educational computer game described in the following pages, both as a developer and researcher. Since then I have interacted with many kids who loved playing QA and who loved learning in the QA context. Observing the interest, devotion, and consequent motivation of these kids, I became curious about

60

H. Tüzün / Multiple Motivations Framework

their reasons for playing this game. What was so motivational in this computer game for almost five thousand kids even though this game was educational? What led them to continue participating in the game activities, including educational ones? In answering these questions, a review of the literature revealed that the major theories of motivation are confined within either the individual or the environment. In addition, the building blocks of these major theories differ significantly. For example, some theories like Keller’s (1983) ARCS motivational design model include the issue of “relevancy” while others mention no word of it. Therefore even if these theories, and specifically the ones based on empirical data like the hierarchy of needs (Maslow, 1987), are good at explaining what makes learners motivated, probably all of them are missing the big picture since each one of them is providing a partial explanation. For that reason, an explanation of motivation that included a broader spectrum of variables, which were both within the individual and within the environment, seemed to be needed. Past research on motivation in educational computer games was dominated by Malone and Lepper’s (1987) “taxonomy of intrinsic motivations for learning.” Their taxonomy asserted that challenge, curiosity, control, and fantasy were the motivational elements for the players of the educational computer games. However, this assertion was limited to isolated individuals. In contrast, recent learning theories emphasize the importance of the social and contextual factors in the learning process. In alignment with this emphasis, after reviewing sixty years of research on motivational research in education, Weiner (1990) concluded that: • • • • •

Older grand formal theories, such as drive, psychoanalytic, and associationistic conceptions, have faded away because they lacked cognitive approaches. Motivational research on individual difference variables was diminishing. Achievement was at the center of the study of motivation. Cognitive variables were starting to be incorporated into motivation theories more and more. There was a growing interest in the incorporation of emotions into motivation theories.

Besides, Weiner (1990) indicated that limiting the motivation studies in learning just with the individual was a narrow focus. He put emphasis on considering frameworks larger than the individual and thinking about extra motivational constructs. He also added that there were “many uncharted areas to incorporate” (p. 622) into motivational theories. Author of this chapter aspired to explore these uncharted areas in motivation and to put these constructs together into a framework of motivation. As a result, “Multiple Motivations Framework” was introduced to explain complex human activities while learning in computer games and learning in general. In this manuscript, I focus on details of introducing this framework and basic presumptions behind it. 2. Quest Atlantis Quest Atlantis (QA, http://www.QuestAtlantis.org) is an educational computer game that immerses children in a 3-D virtual environment for completing educational activities. The purpose of the game is to save mythical Atlantis from an impending disaster (Barab, Thomas, Dodge, Carteaux, & Tuzun, 2005). According to the back story of the game, as the learners complete the educational activities called “Quests,” they help with saving Atlantis from this disaster (Fig. 1).

H. Tüzün / Multiple Motivations Framework

61

Figure 1. Structure of Quest Atlantis.

Quest Atlantis combines play, role playing, adventure, and learning, allowing learners to immerse themselves into virtual 3-D worlds where they select or are assigned developmentally-appropriate Quests, interact with other learners and mentors live from around the world, and build virtual personae (Turkle, 1995; Bers, 2001). The flexibly adaptive (Schwartz, Lin, Brophy, & Bransford, 1999) nature of QA seeks a balance between complete control by designers and easy configuration by teachers and other users. With this approach the game can be adapted to the needs of local contexts. As a result, Quest Atlantis has been implemented in many different contexts, including elementary schools as part of the curriculum, and after school programs as a volunteer activity (i.e., Boys and Girls Clubs of America). In order to participate in QA, children must be related to a particular center (participating elementary schools, Boys and Girls Clubs, or local libraries) and must register on the web site. After registering, they can participate at a centre or from anywhere through a computer with Internet access. On the surface, QA might be perceived as a Multi-User Virtual Environment (MUVE) with rich multimedia elements. In their QA participation, students plunge into this multi-user virtual environment that is divided into different virtual worlds. Each world is further divided into three villages, each of which hosts about 25 Quests. These Quests range from simulation to application problems of differing levels of complexity. All worlds and villages have a theme. For example, when children visit the villages in Culture World, they can explore many different expressions of culture, especially art, music, and writing. Villages in culture world include artists’ village, sound of music village, and words of meaning village. Each village hosts a range of Quests in alignment with the theme of that village. Some of the Quests are combined into thematic unit plans to provide further structuring for teachers.

62

H. Tüzün / Multiple Motivations Framework

Menu and Tool Bars Used for functions such as • Choosing among first and third person view • Controlling avatar gestures • Personalizing settings

Avatar Represents the user in the virtual space and can be controlled by keyboard or mouse

QA Side Windows Used for displaying content such as Quester homepages, Quest content, homepage content, Quest responses, and Q-mail messages

Chat Window User can chat with others

Quest Symbol Shows Quest content in the side window when clicked on

Figure 2. OTAK Interface. A screenshot from Quest Atlantis, showing a scene from a virtual world on the left and the homepage for a student on the right.

At its core, Quest Atlantis would better be described as a virtual environment designed to support an online community as well as multiple face-to-face communities, instead of conceptualizing it as simply a computer software, or a computer “game” utilizing a MUVE and multimedia elements. The Quest Atlantis storyline, its virtual worlds and villages, policies, participant structures, activity sets, and social commitments make up the brand of QA (Barab, Arici, & Jackson, 2005). This brand contains the following key components: • • • • • •

A mythological legend that provides a back story for Quest Atlantis activities. A number of 3-D worlds and villages through which learners, mentors, and the fictional Quest Atlantis council members can interact with each other (Fig. 2). A Personal Digital Assistant (PDA) for each learner, serving as a portfolio of their participation and learning. An advancement system centered on pedagogically valid activities that encourage academic learning, entertainment, and social commitments. Extrinsic rewards structure. A globally-distributed community of participants.

Quest Atlantis lies at the intersection of education, entertainment, and social commitments (Fig. 3). Our QA team have worked to understand how to develop a “computer game that transcends the computer,” that includes inquiry-based and experiential activities that are in alignment with academic standards and that can be assessed for learning gains, that provides entertainment without violence, a girl-friendly environ-

H. Tüzün / Multiple Motivations Framework

63

Figure 3. Quest Atlantis Foundations.

ment that still is attractive to boys, and that is committed to making the world a better place (Barab, Thomas, Dodge, Carteaux, & Tuzun, 2005). Quest Atlantis has many components that can be categorized under different major groups: for example, communication, collaboration, and ownership. Within the game the channels of communication are chatting in the 3D space, the internal Q-mail system, telegramming, and other discourse within the physical space through various means (i.e., talking within the computer lab, or learners talking over the phone). The methods of collaboration are co-questing (i.e., doing Quests together as a team), being part of a guild, requesting help from others, and helping others related to different QA tasks. The modes of ownership are having a personal PDA with various elements on it (emoticons, awards, etc.), Q-points that learners collect after successfully completing Quests and other activities, having a unique representation, called an avatar, through customization, renting virtual land and building on it, artifacts created as the result of the Quests, and QA merchandise (QA trading cards, QA rulers, QA pencils, etc.) that can be purchased from the Quest Atlantis trading post in exchange for the Q-points.

3. Research Methodology for Generating Multiple Motivations Framework The study from which data were obtained to introduce multiple motivations framework can be characterized with multiple labels. First of all, since the study aimed to characterize a group (Fetterman, 1998) it can be described as an ethnographic research. For this characterization I spent two months at the data collection site. In addition, a year of irregular site visits preceded this time frame. The study integrated common elements of

64

H. Tüzün / Multiple Motivations Framework

ethnographic studies, like participant observation, interviews and field work. At the same time the study can also be characterized as a naturalistic research study (Lincoln & Guba, 1985) because the data were collected in a natural setting, in which there were no variables to confirm or disconfirm a priori hypothesis. Quest Atlantis implementation requires developing a vision collaboratively with the stakeholders at local centers. During this process the designers of the Quest Atlantis game are more than participant observers. We call this process as “critical design ethnography,” referring to “an ethnographic process involving participatory design work aimed at transforming a local context while producing an instructional design that can be used in multiple contexts” (Barab, Thomas, Dodge, Newell, and Squire, 2005, p. 254). This process is based upon a collection of methods including ethnographic research, naturalistic research, and action research. The goal of critical design ethnography is to change or empower the culture under study. The purpose of the researcher is to support a transformational process. Towards this end, the researcher acts as a change agent (Rogers, 1995) and she/he participates both within the culture as an “active member” and outside the culture as a “peripheral member” (Adler & Adler, 1987, p. 50). This position, which requires us to be more than a researcher, complicates our role as researchers and presents challenges in addition to those traditionally associated with ethnographic or naturalistic research (Clifford & Marcus, 1986; Fielding & Fielding, 1986; Silverman, 1993). Design ethnographers need to consider three ongoing focal points, which are trust, intervention, and sustainability. Trust evolves based on many factors, including adopting a participatory position, developing multi-tiered relationships, and having an evolving agenda. During the second focal point, a socially responsive design with the purpose of supporting change is carried out. The third focal point requires making the commitment to support sustainable change. The study at this site was conducted during this third focal point of our critical design ethnography. 3.1. Research Questions Since the basic elements of a theory are concepts (Mark, 1996), it was essential to reveal these concepts first. For this purpose, the following analytical research question was asked: “What are the motivational elements of Quest Atlantis, whether intrinsic or extrinsic, in terms of student-defined motivation?” Since motivation is a hypothetical construct (Martin & Briggs, 1986; Good & Brophy, 1997) and its definition differs among academicians, I needed to define motivation as individuals’ showing their willingness to initiate and sustain participation in Quest Atlantis activities. In this sense, I was more interested in continued motivation (Brugman & Beem, 1986; Malouf, 1988), as opposed to momentary motivation. Initiating engagement and sustaining it over time are different phenomenon, and we know little about the latter (Garris, Ahlers, & Driskell, 2002). To discover the relationships between the concepts discovered, a second analytical question was asked: “How do high, medium, and low participating groups differ in their responses with respect to the motivational elements found after answering the first research question?” 3.2. Context Selection I critique many motivational research studies, including the ones by Malone (1980) and Lepper and Malone (1987) that paved the way for their “taxonomy of intrinsic motiva-

H. Tüzün / Multiple Motivations Framework

65

tions for learning” (Malone & Lepper, 1987), in that these research studies offered the innovation, the computer games, to the learners and then measured the momentary motivation of the learners quickly after these games were used. This problem in media studies is known as the novelty effect (Clark, 1983). In studies including the novelty effect, the validity of the findings of the factors that explain the motivation of users playing these games becomes questionable, especially from the perspectives of sustainability and persistency. Bandura (1986) advises that motivation toward activities can be measured at different points in time, which decreases the risk of misinterpreting shortterm changes in motivation. I wanted to eliminate the novelty effect in this study by a purposeful selection of the research context. Since the novelty effect tends to disappear over time (Krendl & Broihier, 1991), the major criterion for context selection was selecting a Quest Atlantis center that was enrolled in the Quest Atlantis program for at least six months. There was just one such Quest Atlantis center, a U.S. Midwestern after-school program, in geographical proximity to the researcher. This center was selected for this study given their willingness to participate. Choosing an after-school context also made it an interesting research context for examining motivation in that learners were not forced by teachers to participate. The after-school context was a member of the national Boys and Girls Clubs of America. The club had 645 members (36% girls, 65% boys). The age group of the children attending the club ranged between six and eighteen years old. There were 346 members between the age of 9 and 12 (54%), which is the target age group for QA. Once members are at the club, it is up to them what to do at the club during the rest of the day. They can participate in activities in game room, canteen area, teen room, library, art room, gym, outside, or computer lab. Quest Atlantis educational game was one of the options in the computer lab among other software titles or activities such as Magic School Bus educational software series, Reader Rabbit’s Math, Amazon Trail, Civilizations II, SimCity 3000, and using the Internet. 3.3. Participant Selection When conducting this study, there were 133 QA members at the club. Altogether, these members had logged on to the game 6344 times and completed 319 Quests, 197 of which were accepted. Among these QA members, those, who had played the game at least five different sessions, and who had spent at least three hours within the game, were chosen for conducting the interviews. With these selection criteria, I wanted to make sure that the learners had accumulated the prerequisite skills necessary to play the game at a basic level. To answer the second research question appropriately, which required having game participants with high, medium, and low participation frequencies, the sampling of the interview participants was done conveniently so that equal number of participants would be obtained in each of the high, medium, and low participation categories. A total of twenty members were interviewed, five of whom were female and fifteen of whom were male. 3.4. Data Collection Ethnographic methods such as interviews, observations, and document analysis were used for collecting data. Semi-structured interviews were the main data collection method of the study. To provide validity for a research study the ethnographer needs to ask the right questions (Fetterman, 1998). To provide this kind of validity, I visited the

66

H. Tüzün / Multiple Motivations Framework

after-school context for a year at different times before this study took place so that I could form the interview questions based on what people did in this context in their daily lives. A total of 20 interviews were completed during a 36-day period, which lasted between 15 and 45 minutes. After their transcription, these interviews resulted in 161 single-spaced pages of data. The attention spans of the members at this age group (nine to twelve) were very low. For that reason, a demographics questionnaire form was developed for collecting factual data. This questionnaire included open and closeended questions related to their club life, Quest Atlantis use, and information technology use. This demographics questionnaire was completed by the interviewees at a time after the interview. Using the questionnaire prevented asking redundant questions that would make the interview length longer. Observations were conducted while the members of the after-school context were interacting with Quest Atlantis in the computer lab. The after-school context was open daily from 3pm to 8pm. My observations took place daily, Monday through Friday, for two months. During the observations, I usually spent the time between 3pm and 7pm in the computer lab, taking notes, summarizing events and interactions that took place physically in the computer lab and virtually in the online QA space. I spent the remaining hour between 7pm and 8pm for entering these notes into an electronic QA database available through the Internet. These observations resulted in 76 single-spaced pages of data. Document analysis included examining materials from the club and digital data from the Quest Atlantis servers. Documents from the club included member information, such as members’ social and economic status, and annual meeting reports. Digital data from the QA servers included all electronic data related to Questers’ participation within the game. These electronic data can be categorized within two groups: the frequency of participation (such as, time spent in the 3D space, the number of logon times to the game, the number of e-mails received and sent, and the number of Quests done) and the content of participation (such as, responses to the Quests, the contents of the emails, and the contents of their chatting in the online game space). 3.5. Data Analysis I used the constant comparison method of grounded theory (Glaser & Strauss, 1967) for data analysis. This method of data analysis seemed to be the most appropriate for several reasons. First of all, this approach is exclusively tailored for producing theory about a substantive area in social sciences. Specifically, grounded theory is “… a general methodology of analysis linked with data collection that uses a systematically applied set of methods to generate an inductive theory about a substantive area” (Glaser, 1992, p. 16). Since my intention was to provide an organizing framework from which to explain things of significance for motivating learners, this kind of data analysis fitted well with this purpose. In this kind of data analysis, categories inductively emerge out of the data rather than being decided prior to the data analysis (Patton, 1987). Possible data sources for producing theory can include interviews, field observation records, and documents (Strauss & Corbin, 1994). All of these data collection techniques were employed, as it was indicated in “Data Collection” section. To analyze the data through the constant comparison method, three stages are followed (Glaser, 1992; Strauss & Corbin, 1998): open coding, axial coding, and selective coding. These three stages were applied to the interview and observation transcription records as explained below.

H. Tüzün / Multiple Motivations Framework

•

•

•

67

During the open coding stage, data are broken down into their parts. To do so, incidents in the data are inspected closely, by comparing for their differences and similarities. During this stage data are conceptualized, so that mountains of data are reduced into manageable pieces. The data may be broken down into parts by three different methods. They can be analyzed line by line, as a whole sentence, or as a whole document. The first five interview documents were openly coded by three researchers (author of this chapter, another doctoral candidate, and a faculty member), while the remaining fifteen interview documents were openly coded by two researchers (author of this dissertation, and another doctoral candidate). The same two researchers coded one of the observation records. All of the remaining observation documents were coded by myself. Before the coding, the researchers read all interview documents to explore and to grasp the content in them. When coding for an interview document the researchers first read the question and the answer for it. Then the answer part of this chunk was coded as a whole paragraph. Data stated by the researcher, like questions or clarifications, were not coded. The researchers negotiated the codes within the chunk until they arrived at a 100% agreement. This process of open coding was independent of the research questions. Open coding of the observation documents was done in a fashion similar to that described above. Naturally, the majority of the codes emerged during the open coding of the interview documents. For that reason, existing codes from the interview documents were used for open coding of the observation documents. Since there were no questions in these documents the coding was done at the paragraph level. After the open coding of the interview documents there were 202 codes. Open coding of the observation records added 32 new codes. At the end a total of 234 codes were obtained. During the axial coding stage, categories are systematically developed and related to each other along their properties and dimensions. It needs to be emphasized that open coding and axial coding are not sequential stages. One needs to move between open coding and axial coding, during which the researcher continues coding for properties and dimensions while developing relationships between categories. Saturation is reached when a category does not seem to produce any more properties or dimensions. Just after completing the coding of the interview documents and just before starting the coding of the observation documents two researchers preliminarily organized 202 codes in 16 categories. These 16 categories were: store items, edutainment, building, 3D, identity, social, different from others, QA extras, people, homepage, Boys and Girls Club context, design, implementation, control, motivation, and feelings. Further, these categories were collapsed then again under 5 of these original categories so as to create more parsimony and usefulness to the codes: The different from others category included identity, social, building, and edutainment; design category included homepage, 3D, and QA extras; motivation category included store items and control; implementation category included Boys and Girls Club context and people; and feelings category stood alone. During the selective coding stage, the theory is integrated and refined. For this purpose, the emerging story is explained around a core category, while all the other categories are linked to this core category. The three researchers got together to discuss 16 categories found during the axial coding stage and the

68

H. Tüzün / Multiple Motivations Framework

codes within them. Since the open coding and the collapsing of categories were done independent of the research questions, these 16 categories and the larger 5 categories were characterizing the data well in general but not well with respect to the research questions of the study. For that reason, the researchers re-debated the codes and the categories in light of the data, by using their own characterization of motivation based on salient themes and the research questions, and by re-visiting the current literature on motivation theories. This re-debate was a dialectic intersection of the categories grounded in the data, our intuitive responses to the research questions, and the current theories of motivation. After the re-debate a number of changes were made. Nine of the categories were kept but renamed: building as creativity, social as social relations, identity as identity presentation, store items as rewards, motivation as achievement, 3D as immersive context, Boys and Girls Club context as context of support, different from others as uniqueness, and control as control and ownership. Six categories were dropped and they were collapsed under other categories: QA extras, people, homepage, design, implementation, and feelings. The remaining edutainment category was huge in size; therefore it was split into playing and learning categories. Two new categories emerged which were previously nested under one of the 16 categories: curiosity and fantasy. In the end, we obtained thirteen categories all of which related to and were placed under one of the research questions: identity presentation, social relations, playing, learning, achievement, rewards, immersive context, fantasy, uniqueness, creativity, curiosity, control and ownership, and context of support. These thirteen categories, their dimensions in QA, and the total number of responses coming from the interviewees related to each of the categories and their dimensions in QA are presented in Table 1. 3.6. Measuring Participation and Assigning Participants to Groups The second analytical research question required assigning the interviewees into high, medium, and low participating groups. The difficulties with measuring participation in research studies have been documented in education as well as other domains (see, for example, Mussino, 1999). The biggest difficulty is the selection of indexes of participation, i.e. what activity or activities show participation in a specific field. To overcome this complexity in education Fin (1989) developed a four-part taxonomy to identify the forms of student participation in schools. Participation in the first level is basic and it involves students’ tendency to attend the class. At the second level students take initiative in the class. For example, they might ask questions to teachers or they might do extra school work. The third level of participation occurs outside the class. For example, students might participate in social or extracurricular school activities. The fourth level involves the empowerment of students by involving them in the school’s disciplinary system or school government. Finn, Folger, and Cox (1991) developed an instrument for elementary school students to measure their participation based on the fourpart taxonomy. They also examined the empirical relationships among the parts of the taxonomy. They found the correlations among the parts of the taxonomy sufficiently high, concluding that any one of them could be used as a single participation index. In alignment with Finn’s taxonomy of participation, I chose two indexes to calculate Questers’ participation in QA. These indexes were the total amount of “time spent in QA” and the number of “Quests” undertaken. Since all QA activities take time

H. Tüzün / Multiple Motivations Framework

69

Table 1. Thirteen Categories and Sub-Categories as Motivational Elements.

Category

1. Identity Presentation

2. Social Relations

3. Playing

4. Learning

5. Achievement

6. Rewards

7. Immersive Context 8. Fantasy

9. Uniqueness 10. Creativity 11. Curiosity 12. Control and Ownership 13. Context of Support

Sub-category Avatars Usernames Homepages Interaction with others Sharing Competition Showing off Groups Privacy Security Multimedia Points Pushball Unique learning Meaningful learning Active learning Feedback Multimedia Attitudes Challenge Recognition Awards Points Trading cards Open market Support structures 3D QA myth Council members Unique opportunity Different from others QA vs. others Building End of the game Quest status Secret places Control Jobs School vs. club differences Username Trading post items

Total number of responses (phrases) 22 83 10 51 166 40 3 260 28 13 5 5 5 147 138 4 4 17 231 288 23 5 8 9 36 27 9 185 269 7 68 6 170 164 19 31 12 19 133 16 98 129 129 3 28 5 20 38 52 14 9 14 2 3

70

H. Tüzün / Multiple Motivations Framework

to complete, Questers need to spend time within the game as a basic requirement. For that reason, the total amount of time spent in QA is equivalent to the first level of Fin’s (1989) participation taxonomy. Doing Quests is the salient activity in this educational game and initiative is required to do them. For that reason, the number of Quests undertaken is equivalent to the second level of Fin’s (1989) participation taxonomy. Since kids attend the club with varying frequencies, kids attending the club more frequently would have more chance to spend time in QA. For example, a youth visiting the club everyday would have more chance to spend time in QA than a youth visiting the club just once a week. This would create a measurement error while calculating the participation in QA. To eliminate this error for “time spent in QA,” I divided “time spent in QA” by “club attendance per week” for each interviewee (Table 2; since there were three kids playing QA once a week at their schools, I added a day to their “club attendance per week,” as bolded in Table 2) — this gave me equalized “time spent in QA.” Then I calculated standard scores for equalized “time spent in QA” and “Quests.” For this purpose, I calculated Z scores for equalized “time spent in QA” and “Quests” by using SPSS v11.5. Since the cognitive and intellectual development of kids are dependent on doing educational activities more than other activities in the game, Questers’ doing the Quests in QA are given the most importance by the designers of QA. In alignment with this philosophy, I gave more weight to the index of “Quests” while calculating the overall participation score. To obtain the participation score I doubled the Z score of “Quests” and added it to the Z score of equalized “time spent in QA.” Table 2 shows interviewees sorted by their participation scores from high to low. The participation scores were used as a mean to sort the interviewees based on their participation. For that reason, the numbers themselves are not meaningful beyond that purpose. After sorting the interviewees by their participation scores it was apparent that Questers towards the top of list would belong to the high participating group, Questers towards the bottom of the list would belong to the low participating group, and Questers between these groups would belong to the medium participating group. The difficulty was in choosing the cutoff points. To determine the cutoff points I compared the Questers in the list next to each other starting at the top of the list. While comparing them I depended on my intuition, which was based on my longitudinal observations of the kids. I asked the following analytical question during the comparisons: “Did I observe a difference in the participation of Questers in row X and row X+1?” After following this methodology, it became evident that the first seven Questers belonged to the high participating group, while the next seven Questers belonged to the medium participating group, and the last six Questers belonged to the low participating group. Each Quester’s belonging to a specific group highly correlated with my intuition. In order to check the validity of these results, I asked the most experienced member of computer lab staff to put these twenty Questers into high, medium, and low participation groups based on their participation in QA. I asked him not to look into QA usage statistics and just to depend on his observations. He placed 17 of the 20 Questers into three categories in alignment with the assortment in Table 2. This high agreement validated the methodology for dividing the Questers into three participation categories. 3.7. Trustworthiness Lincoln and Guba (1985) summarized four areas for considering the importance of any scientific study: truth value, applicability, consistency, and neutrality. Since qualitative

H. Tüzün / Multiple Motivations Framework

71

Table 2. Questers Sorted by Their Participation in QA. Time spent in QA

Club attendance per week

Equalized time spent in QA

Quests

Z of equalized time spent in QA

Z of Quests

Partici-pation Score

Frequency of participation

Quester Andrew

71

5

14.2

22

1.03

2.93

6.88

High

Emily

142

5

28.4

5

3.27

–0.14

3

High

David

27

5

5.4

15

–0.36

1.67

2.97

High

Jason

27

4

6.75

13

–0.15

1.31

2.46

High

John

47

5

9.4

10

0.27

0.77

1.8

High

Ryan

80

5

16

6

1.31

0.05

1.4

High

Kevin

18

2

9

6

0.21

0.05

0.3

High

Anthony

21

2

10.5

5

0.45

–0.14

0.18

Medium

Jennifer

12

3

4

6

–0.58

0.05

–0.49

Medium

Thomas

41

5

8.2

4

0.08

–0.32

–0.55

Medium

Rebecca

8

5

1.6

6

–0.96

0.05

–0.87

Medium

Eric

6

1

6

4

–0.27

–0.32

–0.9

Medium

Brian

16

5

3.2

5

–0.71

–0.14

–0.98

Medium

Amy

10

2

5

3

–0.42

–0.5

–1.41

Medium

Mark

9

1

9

1

0.21

–0.86

–1.5

Low

Scott

18

3

6

2

–0.27

–0.68

–1.62

Low

Tyler

6

1

6

2

–0.27

–0.68

–1.62

Low

James

17

5

3.4

0

–0.68

–1.04

–2.75

Low

Luke

4

5

0.8

0

–1.09

–1.04

–3.16

Low

Sarah

3

4

0.75

0

–1.09

–1.04

–3.17

Low

research and quantitative research differ in their world views, they require different kinds of paradigms to evaluate their worth. Creswell (1998) documented that multiple perspectives existed for the verification of results in the qualitative research paradigm. Lincoln and Guba (1985) recommended using the word “trustworthiness” to refer to the verification in qualitative studies. They defined the trustworthiness as persuading the audience of a research study that the findings of the study are worth paying attention to. Aligned with their taxonomy for considering the importance of a scientific study, Lincoln and Guba (1985) offered the term credibility to deal with the truth value, the term transferability to deal with the applicability, the term dependability to deal

72

H. Tüzün / Multiple Motivations Framework

with the consistency, and the term confirmability to deal with the neutrality, for considering the importance of qualitative research studies. Since multiple realities are involved in a qualitative research study Lincoln and Guba (1985) offered “credibility” as an operational term. The implementation of credibility requires two tasks: doing the research in such a way that the possibility of finding credible outcomes is enhanced, and showing this credibility by having the results agreed to by the constructors of multiple realities (i.e., participants and other researchers). With respect to the former task, I followed prolonged engagement with the research site (frequently for a year), persistent observation in the research site (daily for two months), and triangulation of methods, sources, and researchers. Fetterman (1998, p. 36) agreed that “working with people, day in and day out, for long periods of time is what gives ethnographic research its validity and vitality.” With respect to the latter task, I followed up with peer debriefing. During the peer debriefings I was challenged by the QA design and research team. This helped me with increasing my awareness to “substantive, methodological, legal, ethical, or any other relevant matters” (Lincoln & Guba, 1985, p. 308) about the research study, with defending my assertions, and with developing the methodology. The implementation of transferability requires providing a thick description of the culture (Lincoln & Guba 1985; Gilbert 1993; Creswell, 1998; Merriam 1998). If the researcher knows the rules and norms of the culture under study and if he can convey this information to the readers such that they can integrate themselves into the culture by following the description of the researcher, transferability is established (Gilbert 1993). Further, since the researcher can not apply the findings to many other substantive fields, providing a thick description of the research site allows others “to compare the ‘fit’ with their situations” (Merriam, 1998, p. 211). Dependability criterion reflects the consistency of a qualitative study. The existence of multiple realities and the changing nature of the research site create “instabilities” (Lincoln & Guba, 1985, p. 299), which prevent applying the traditional reliability techniques into qualitative research. Lincoln and Guba (1985) use the term “confirmability” to refer to the neutrality of a qualitative study. They claim the emphasis of objectivity should be on the data instead of the researcher. Therefore, confirmability becomes an issue of checking the characteristics of the data. Lincoln and Guba (1985) recommend using external audit trails to establish both dependability and confirmability at the same time. In this sense, an outsider can examine data, findings, and interpretations just like a fiscal auditor checks the process and the product of a business account (Creswell, 1998). Approving the process provides dependability while approving the product provides confirmability. During the selective coding stage, the third researcher provided such an audit trail. He examined the products of the two researchers, which were the data and the sixteen categories, and also the process to produce those products.

4. Participation Patterns of High, Medium, and Low Group Members over Time To answer the second research question, each of the high, medium, and low participating groups was characterized based on the thirteen motivational categories. These three groups were then compared for similarities and differences to see to what degree their participation differed. Figure 4 provides the mean number of responses for the motiva-

H. Tüzün / Multiple Motivations Framework

High

Medium

Low

73

All

20 18 16

Phrases

14 12 10 8 6 4 2 0

Identity Presentation

Social Relations

Playing

Learning

Achievement

Rewards

Immersive Context

Fantasy

Uniquness

Creativity

Curiosity

High

4.6

15.6

14

12.3

1.9

14.4

7.1

3

9.1

11.1

1.3

3

Medium

5.7

12.6

13.6

18

2.6

14.6

9.3

0.4

6.1

4.4

1.4

3.3

Low All

Control and Ownership

4

10.5

8.3

11.3

0.8

9.8

9.2

1.2

4.3

3.3

0.2

1.3

4.8

12.9

12.0

13.9

1.8

12.9

8.5

1.5

6.5

6.3

1.0

2.5

Figure 4. A Visual Comparison of the Mean Number of Responses from High, Medium, and Low Participating Group Members with Respect to the Motivational Categories.

tional categories for each of the high, medium, and low participating group members in a visual form. Members of the high participating group wanted to lead other kids, in this way they can be described as innovators or early adopters in Roger’s (1995) terms. Members in the medium and low groups seemed not to fit into this category. The high group members spent significantly more time (59 hours) in the game than the medium group (16 hours) or the low group (10 hours) members. Although there was a huge difference between the high and medium groups in terms of time spent in the game, the understanding of the kids from both groups about the participant structures in the game was pretty close. The low group members seemed to need more time to know more about the game. The mean age for the high and low groups was both 10.5, but this was 12 for the medium group. The family income for the high and medium groups was very close, $36,000 and $34,000 respectively, however it declined noticeably to $21,500 in the low group. However, it should be noted that the family income of the most participating Quester was $20,000 and the family income for the least participating Quester was $40,000. Therefore, the family income did not necessarily correlate positively with the participation. QA as a whole seemed a fun game for all three group members. When asked about the three most favorite activities in QA, the themes that included these activities matched for all three groups: learning, social relations, and immersive context. There was an additional category of creativity for the high group. Although their most favorite activities matched, the order of these differed among the groups. The order for the three most favorite activities in QA was:

74

H. Tüzün / Multiple Motivations Framework

• • •

Social relations, creativity, learning, and immersive context for the high group. Social relations, learning, and immersive context for the medium group. Immersive context, learning, and social relations for the low group.

This preference was reflected in the interviews. For example, low group members talked most about the 3D worlds and villages, navigation in 3D, and avatars which constituted the immersive context category. On the contrary, members of the high and medium groups complained about the paths in 3D that became routine and suggested placing shortcuts to eliminate this problem. The least favorite activities for the medium group were related to the learning category and the least favorite activities for the low group were related to reading. The high group did not complain about the learning. Considering the identity and social relations, the mean number of responses for the social relations category was much higher than the identity category for all three groups. The interview and observation data supported this outcome; all kids loved interacting with others through various communication modes, shared information, competed with others, showed off their own game artifacts (points, awards, selfinformation, etc.) to others, and experienced conflict at times as the result of all these interactions. Their usage statistics with respect to using e-mail, chat, and number of entries on the friends page correlated positively with the time spent in the game, and these numbers decreased while going from the high to low participating groups. There was a slight difference between the high and –medium and low– groups in that the members of the former included others in the same context in their gameplay more than the members of the latter. Avatars had a higher impact on the identity of the Questers in the medium and low groups. High group members reflected their identity more in their homepages; each of them detailed their identity on their homepages. This using of homepages for identity decreased towards other groups; for example, half of the medium group and one third of the low group entered their self information into their homepages. Related to using the functions on the homepage, all three group members seemed to have limited information processing capability. One of them specifically indicated that she forgot about some functions during her usual play, or that some functions were just out of focus at a specific time. For the members of the medium participating group and low participating group the mean number of responses for the learning category was higher than the playing category. On the other hand this was the opposite for the high participating group; the number of responses related to the playing category was higher than the learning category. The mean number of responses for the playing category for the high group (14) and medium group (13.6) was very close. The mean number of responses for the rewards category was much higher than the achievement category for all three groups. This is a clear indicator that members of all groups valued extrinsic rewards. These rewards included QA points, awards, and trading post items. Members of the high group actively exchanged their points with items in the trading post; however, the majority of the members in the medium and low groups did not know about the trading post and participated in limited transactions. Although they did not know about the trading post and items in it, these kids heard about the availability of some items from others. The mean number of responses for the rewards category for the high group (14.4) and medium group (14.6) was very close.

H. Tüzün / Multiple Motivations Framework

75

The mean number of responses for the immersive context category was close for the medium and low groups, being 9.3 and 9.2 respectively. This was 7.1 for the high group. All group members knew about the worlds and villages that made up the 3D space, but members of the high group were able to discriminate between the worlds and villages better than the other group members. The fantasy category was another category that seemed to be different for the high group. In this category dimension, the medium and low groups were close to each other in that they did not absorb the legend well, and they did not remember about fictitious game characters. Many high group members on the other hand linked their efforts to helping the Atlantis people. Related to the uniqueness category, the number of responses decreased gradually through the high group to low group, from 9.1 to 6.1, and to 4.3. Of particular note, the members in the high and medium groups pointed to the unique learning and social interaction opportunities afforded by the game. For the low group, QA did not seem any different from other games, excluding the immersive context mostly provided by the 3D feeling of the game. The mean number of responses for the creativity category was 11.1, 4.4, and 3.3 for the high, medium, and low groups respectively. Almost all members of the high group actively built on virtual land, while almost all other Questers planned to rent land and build on it. The enjoyment of the building was not limited just to the builders; many others indicated that visiting these structures was an activity in which they regularly engaged. The mean number of responses for the curiosity category was very close for the high group (1.3) and the medium group (1.4). This closeness was reflected into the group characterization equally; both groups indicated secret places was the most exciting activity to do, and three to four kids from each group talked about the excitement of finding secret places. The low participating group members did not frequently engage in finding secret places, the mean number of responses being just 0.2. Related to the control and ownership category, the numbers for the high and medium groups were close. This was 3 for the high group, 3.3 for the medium group, and 1.3 for the low group. All group members from the high participating group felt the tension of the controlling elements in the game and complained about them. Meanwhile, tension was a noted factor for only half of the medium group members. Lastly, none of the low group members felt any control tension. Related to jobs, two members from the high group signed up for a job and an equal number of people from both the high and medium groups indicated their intentions to sign-up for a job soon. Low group members were not aware of the availability of the jobs.

5. Multiple Motivations Framework When I looked into the results of my data analysis, it was surprising that such a broad range of categories that motivated children in QA emerged. Traditionally, research regarding motivation in computer games has characterized motivation in a smaller number of categories, usually challenge, curiosity, control, context, and fantasy. Since I have been a close follower of the computer games since childhood, I was expecting new emergent categories such as interaction based on my observation of the popularity of online multi-player games. Initially I had no ideas or expectations much beyond that.

76

H. Tüzün / Multiple Motivations Framework

Because of this diversity of motivational reasons to play an educational game, I want to refer to my findings as the “multiple motivations framework” for playing educational computer games. Since the data helping me understand the construct of motivation were collected from a single context, I would prefer to call it a framework rather than a theory. I continue to conduct studies involving learners with different age groups, different subject matters, and different contexts to evolve my framework into a theory. Any scholar in education or social sciences can guess that I was influenced by the theory of multiple intelligences developed by Gardner (1993) while naming my framework. This guess would be correct. I believe such a name selection reflects learners’ multiple reasons for learning. A theory tries to explain something that is not easily visible (Gilbert, 1993; Mark 1996). When theories are used to explain something, they provide reasons for a specific phenomenon. For example, information-processing theory describes, among other things, that information goes through the short-term memory before entering the long-term memory. It does not inform how to facilitate learning (Reigeluth, 1999). Theories that are descriptive can be used for prediction, i.e. given an initial event what event will likely to occur (Reigeluth, 1999). Prescriptive theories on the other hand offer guidelines for using the best methods to achieve a specific goal. For example, a prescriptive theory might recommend relating new information to learners’ previous knowledge for helping the retention of that information in the long-term memory (Reigeluth, 1999). There could be other methods for the retention of information in the long-term memory. Therefore, while the major concern for the prescriptive theories is their preferability (i.e., does this method achieve the goal better than any other known method?), validity is the most important concern for descriptive theories (Reigeluth, 1999). The theory of multiple motivations is prescriptive in nature since it includes categories and methods that motivate learners in educational computer games. To a certain degree it is also descriptive at the same time, because it describes the relationships between these categories. When a theory is explicated, it is necessary to clarify its content and form (Steiner, 1998). The basic elements of a theory are concepts (Mark, 1996). Concepts are related to each other to produce universal generalizations. Concepts and these universal generalizations are the content of a theory. Relationships between the concepts and relationships between universal generalizations give a theory its form (Steiner, 1998). The purpose of my multiple motivations framework is to provide an organizing framework from which to explain things of significance for motivating learners (Fig. 5). In this sense, its content includes categories which are formed by relating concepts through the constant comparison analysis of the data. Although my framework is inducted from multiple interviews and observations, it is only in relation to one study so I offer it as a descriptive framework for making sense of and interpreting this dataset and not yet as a theoretical framework with generalizable power. The categorical content of multiple motivations framework contains the categories of identity presentation, social relations, playing, learning, achievement, rewards, immersive context, fantasy, uniqueness, creativity, curiosity, control and ownership, and context of support. After a comprehensive examination of these thirteen categories in the light of my data, further relationships between some categories, universal generalizations so to speak, emerged. These generalizations are characterized as an organizing framework and presented as a series of dualities. Along these relationships are those between identity presentation and social relations, playing and learning, and achievement and rewards. In addition immersive context, fantasy, and uniqueness came closer

H. Tüzün / Multiple Motivations Framework

77

Figure 5. Multiple Motivations Framework.

while creativity, curiosity, and control and ownership made another group. The fourth relationship is between these groups, with three categories in each. I will refer to the relationships between the categories as dualities. Wenger (1998) defines a duality as “a single conceptual unit that is formed by two inseparable and mutually constitutive elements whose inherent tension and complementarity give the concept richness and dynamism” (p. 66). In this sense, dualities are not polarizations in a Hegelian or Marxist sense. They exist together. Dualities do not make up a spectrum. For example, going from one side to the other does not imply leaving the other. The effective functioning of one side of a duality necessitates, and is dependent on, the existence of the other (Barab, Makinster, & Scheckler, 2003). Sides of a duality describe an interplay. Understanding this interplay between the sides of a duality is of primary interest. For example, Wenger (1998) and Barab et al. (2003) utilized this understanding as a framework to understand the community life. Here I utilize the interplay between the sides of dualities in Multiple Motivations Framework to understand motivation in learning.

78

H. Tüzün / Multiple Motivations Framework

I will define the universal generalization between the categories of identity presentation and social relations as the “subject” duality, the one between the categories of playing and learning as the “activity” duality, and the one between the categories of achievement and rewards as the “outcome” duality. Additionally, I discuss the universal generalization between identification and negotiability as the “object” duality. Here I need to note Leontiev’s concept of activity. Vygotsky and later his student Leontiev struggled to explain the differences between human beings and animals. For this explanation, Leontiev came up with the concept of activity. In his explanation he conceptualized activity as a collective process between the individual and community (Hedegaard, Chaiklin, & Jensen, 1999). In this sense, I need to clarify that my concept of activity is different from that of Leontiev’s in that activity is not necessarily a collective process. However, I find Leontiev’s framework, discussions, and even categorical labels as discussed in terms of activity theory to be informative of my work, especially in that I use the labels subject, object, activity, and outcome.

6. Core Dualities in Multiple Motivations Framework When the motivational categories in Multiple Motivations Framework are closely examined, it is evident that some of these categories make up an activity system, while some other categories directly pertain to the designed product. The activity system centers on the designed product and follows a historical route. Participation patterns of the high, medium, and low group members over time helped with revealing the historical activity system. Both the designed product and the activity system exist in a particular context. Dotted lines enclose the designed product and the activity system in Multiple Motivations Framework to reflect dissimilar contexts with differing cultures and norms. Dotted lines simulate the variability of the context; i.e., some other motivational categories might be introduced or some of the motivational categories might disappear in different contexts. However, I will argue that the basic activity system, including subject, activity, and outcome, will remain unchanged for most of the contexts. Below, components of the Multiple Motivations Framework, including subject, activity, outcome, and object are detailed. 6.1. Duality of Subject Identity presentation and social relations categories make up the “object” duality. In this sense an individual likes to be part of a culture or social structure, however the individual still would like to keep his identity through various means, a duality expressed by Jung (as cited in Stone, 1997) as individuation (our need to be unique) and integration (our need to be part of a community). Therefore, an individual can participate in activities alone or he can participate with others. Within QA context, identity found it’s meaning with avatars, usernames, and homepages. The existence of the individuality in the game starts with transforming into an avatar. Most starters of the game first customized and changed their avatars. A second indication of identity is usernames. Having a unique username is one of the ways the learners can represent their identity. Learners exposed more of their identities through their homepages. By revealing miscellaneous data about their selves (like what they liked, and things that they were good at) they made public what and who they were. Avatar, username, and homepage of a person were synthesized into a unified

H. Tüzün / Multiple Motivations Framework

79

identity. Having such an identity was a reason for them to play and continue to participate in this game. This identity appeared to be complemented through social relations as they interacted with others. Social relations happening in the game was one of the biggest motivators for the players. These social relationships included interacting with others, sharing, competing, forming groups, and showing off. These relations happened both within the online space and within the physical space where they connected to the game. In their relations, they interacted with various people through multiple communication modes. At times there was competition among the Questers but data showed that sharing dominated over competition. Although there were Questers who wanted to play the game individually, playing it as a group was more frequent. And security features within the game bettered the social relations. 6.2. Duality of Activity Playing and learning categories make up the “activity” duality. In Multiple Motivations Framework, activities are performed by the duality of subject and activities result in the duality of outcome. The combination of lower level actions produces higher level activities. For example, when playing a game called pushball, players travel to the healthy world, navigate to the pushball arena, follow the ball object, and attend to other players and the scoreboard. When completing a Quest, they travel to different game worlds, navigate in them, find Quests, read and listen to Quest descriptions, interact with other learners to get help, submit answers, and get feedback. Each of these steps is an action, and actions contribute to the higher level activities of playing and learning. Actions can be described as activities themselves, but one difference between activities and actions is that activities are relatively more meaningful than actions and therefore they have the potential to produce greater enjoyment. Perhaps the action of navigating in a 3D world is an amazing feeling for a learner who just starts the game (like for those in the low participating group). But over time this novelty wears off and users’ enjoyment decreases. On the other hand, activities give learners consistent, sustained enjoyment. This sustained enjoyment comes from the tight integration of actions that lead to an activity. Since there is a broad range of combinations of actions one can take while engaging with an activity, each instance of an activity becomes unique. As an example, a learner might logon to the game, meet with new players and talk to them, travel to “words of meaning” village in culture world, browse Quests in this village, choose a Quest, read the Quest description, browse online resources, discuss Quests with other players, and submit his answer in text format. Another player, on the other hand, might logon to the game, check to see if her previous response was approved, travel to “sound of music” village in culture world, find her assigned Quest, listen to a narrated Quest description, browse online resources, browse her textbook, create the artifact the Quest asks for, and submit this artifact as a response to Quest. Therefore, although these two learners can be said at the macro level to have simply completed a Quest, their engagement at the micro level differs greatly and reflects diversity. Interestingly, for some activities there were varying opinions on the type of that activity. For example, some kids characterized completing Quests more as learning while some other kids characterized the same activity more as playing. This difference comes from the diversity of combinations one can take while engaging with an activity. Kids who characterized doing the Quests more as learning were weak at integrating the

80

H. Tüzün / Multiple Motivations Framework

play elements (like those in the low and medium participating groups). On the other hand, kids who characterized doing Quests as fun tended to be good at integrating play elements into their learning (like those in the high participating group). Nonetheless, one should keep in mind that when the perception of an activity weighed more on playing or learning, the other part was still remembered and given importance. This supports the duality or inseparability of playing and learning. The playing and learning were so intertwined for the kids that they had difficulty distinguishing the two. For example, when I pointed to the education aspect of the game they reminded me about the playing aspect of it. Similarly, when I turned to the playing side they argued that one would also learn at the same time. Therefore, there was no point or need for them to separate the playing and learning. Overall, learners considered QA as a game and there was no question about this issue. Multimedia elements, points, and the pushball game contributed, to a great extent, to the game aspect of QA. Some other categories, like immersive context, creativity, or fantasy, were also considered as play elements. While these components, multimedia elements, points, and pushball, make QA game-like and make the learning fun, the fun part also comes from the learning itself. What makes learning fun are the features of it: meaningful learning and active learning. These two features of learning are highly interrelated in QA context. 6.3. Duality of Outcome Achievement and rewards categories make up the “outcome” duality. The duality of outcome is the result of the activity duality. Achievement refers to the enjoyment and recognition learners get after overcoming learning activities. Rewards refer to the extrinsic incentives they obtain. Based on my data, I can assert that most of the children interviewed in this study perceive their learning and education “as a job.” I speculate that the current status of the society, which has become increasingly materialistic, might have an impact towards this end. Considering their learning as a job, it is so natural for them to expect a return for their effort. Who in the world works in a job without payment? Therefore, although they like the achievement of overcoming the challenges and the recognition associated with it, obtaining some kind of extrinsic incentives is indispensable. Both the achievement and the rewards are the conditions for the present that will affect the future. In QA, the rewards are both materialistic and non-materialistic. Among the non-materialistic rewards are points, awards, and social approval. Among the materialistic rewards are items in the virtual trading post like trading cards, Internet time, pencils, rulers, and t-shirts. The availability of both kinds of rewards gives learners choice options for the outcome of their activities. 6.4. Duality of Object Six categories, immersive context, fantasy, uniqueness, creativity, curiosity, and control and ownership, make up the duality of object. Specifically immersive context, fantasy, and uniqueness categories contribute to the “identification,” while creativity, curiosity, and control and ownership categories constitute the “negotiability.” I adapted the terms identification and negotiability from Wenger (1998). He utilizes identification and negotiability duality to characterize identity in communities of practice. Identification refers to “the process through which modes of belonging become constitutive of our identities by creating bonds or distinctions in which we be-

H. Tüzün / Multiple Motivations Framework

81

come invested” (p. 191). Negotiability is “the ability, facility, and legitimacy to contribute to, take responsibility for, and shape the meanings that matter within a social configuration” (p. 197). In the framework of multiple motivations I use these two terms slightly differently. In my framework, identification refers to the overall reification of game material providing the experience. In a sense, identification is the learners’ identifying the game as something. Research participants identified QA as an immersive context with fantasy elements, and as completely different from other things, which made it unique. Negotiability on the other hand is the investment of learners in this reification. In a sense, it is learners’ impact or “mark” on the game. Research participants negotiated QA identity by building in it, by investing their curiosity in it, and by exercising control over it, which increased their ownership of the game. 6.5. Context of Support QA is implemented in multiple contexts, including schools and after-school environments. It is significant to put forward the contextual implementation differences in different contexts. Three of the interview participants also utilized the game in their schools and all three participants indicated differences in their schools implementations, which made the QA experience different from the after-school context. First of all, because of limited time at schools, it appears that teachers try to maintain the management of learners by enforcing control over them. For example, in one of the learner’s classroom the teacher synchronized Questers’ participation so that everybody was doing the same activity at the same time. Learners revealed a frustration over this control. They indicated that with increased control on the teachers’ part, and decreased control on their part their motivation to participate in the activities tended to decrease. A second contextual difference was in the assignment of usernames. At the club Questers were able to pick up any username they liked. At the schools teachers tend to assign usernames to Questers, usually in the combination of Questers’ names, last names, and some numbers, probably for easier management of their classes. However, this strategy may eliminate the identity of players and at the same time removes the empowerment from them. The third contextual difference was in the variety of rewards in the virtual trading post. The QA implementation at the club included many more materialistic items in the virtual trading post than the school implementations. As it was discussed under the duality of outcomes, having a diversity of materialistic rewards is a necessity for both the effort of Questers and also for the creation of a QA economy. Possibly, addition and distribution of these rewards by teachers has been neglected by time constraints or they were simply in conflict with their teaching beliefs. Apart from these differences, the “Internet time” emerged as a contextual item in the trading post of the club. The culture of the club was such that the use of Internet was tied to using educational software. When members used educational software for a specified amount of time, they then had the right to use the Internet for a specified amount of time. Having the Internet time as an item in the virtual trading post was extremely meaningful and valuable in this context. This reward item may not make sense and may not have a value in most other contexts, like schools. As these examples verify, the culture, values, and norms of the context of the game implementation can make a difference in providing the motivational categories. More research needs to be done in broader contexts in order to see the extent to which the implementation of motivational categories differs in other contexts.

82

H. Tüzün / Multiple Motivations Framework

7. Assertions of Multiple Motivations Framework In the following section, I present my assertions in the light of Multiple Motivations Framework. In this section my intention is to generalize my framework to the broader topic of motivation. Assertion 1: Motivation is distributed among many elements of the context Traditionally, theories of motivation have focused on just one, or a few, traits. My framework of multiple motivations includes multiple elements that contribute to one’s motivation and that collectively constitute the activity of motivation. Ignoring most of these elements, and the interactions between them, while focusing on just one or a few of them, could produce incomplete research results and possibly invalid conclusions. Motivation is dependent upon not just reinforcers as Skinner (1953) suggested, not just intrinsic reasons as Deci (1975) and Malone and Lepper (1987) suggested, not just modeling others as Bandura (1986) suggested, not just self-actualization as Maslow (1987) suggested, not just need for achievement as Atkinson (Atkinson & Feather, 1966) suggested, and not just the origin of people’s own actions as De Charms (1968) suggested. As a researcher in social sciences, I understand these researchers’ passion for explaining motivation with reduced variables just like physicists and astronomers have been struggling to come up with a theory of everything with a compact formula like e = mc2. However, I don’t see this as simply a case in motivation research, and instead posit that the most condensed form would have to include multiple motivations. Assertion 2: These elements are both intrinsic and extrinsic to the learners While browsing the literature on motivation and learning, it has been so typical to come up with a piece like the following for commenting on the decrease in motivation in learning: ... Before school age, learning seems clearly and universally intrinsically motivating for children. Few of us have ever seen, or even heard of, a three- or four-year-old with a “motivational deficit.” Instead, young children seem eager and excited about learning of all sorts, and the more typical parental complaints concern their children’s apparently insatiatable curiosity and boundless energy. Yet, by the time these same children have entered school, a sizeable fraction are quickly labeled as having motivational difficulties of one sort or another in learning (Lepper, Sethi, Dialdin, & Drake, 1997, p. 23).

The same ideas were expressed in the past by Cordova and Lepper (1996) and Brugman and Beem (1986). These kinds of ideas can be summarized in the following steps: 1) children are motivated to learn from their birth, 2) when children enter school their motivation to learn falls dramatically, and 3) what is the reason for this decreased motivation as the children grow up? Lepper et al. (1997) indicated that there was no single answer to this question, and worse there were no convincing data to help with choosing alternative explanations. One of their possible explanations was the heavy use of extrinsic rewards over time undermining children’s intrinsic motivation in the school. Actually there has been extensive literature that attempts to explain the decrease in motivation of learners as being the undermined intrinsic motivation. Three independent studies conducted almost at the same time by Deci (1971, 1972), Kruglanski, Friedman, and Zeevi (1971), and Lepper, Greene, and Nisbett (1973) showed the negative

H. Tüzün / Multiple Motivations Framework

83

effects of the extrinsic rewards on learners’ subsequent intrinsic interest in the activities, for which the extrinsic rewards were no longer available (as cited in Lepper & Henderlong, 2000). Since then, another 100 additional research studies have been conducted challenging the same issue; however, these follow-up studies came up with a similar conclusion (Lepper & Henderlong, 2000). Some other researchers on the other hand objected to the idea of the negative effects of rewards on intrinsic motivation. These various meta-analytical reviews of previous research on this issue revealed that negative effects of rewards occur under certain conditions, and rewards can be used to increase motivation when properly arranged (Cameron & Pierce, 1994; Eisenberger & Cameron, 1996; Cameron, Banko, & Pierce, 2001). Based on these conclusions, Cameron and Pierce (2002) stated that intrinsic motivation was a misguided construct. The opponents of these findings claimed that these meta-analyses were flawed and that their conclusions were incorrect and came up with their meta-analyses (Deci, Koestner, & Ryan, 1999; Deci, Koestner, & Ryan, 2001). They claimed again that extrinsic rewards undermined intrinsic motivation. While there are different viewpoints, my finding is that multiple motivations can exist simultaneously, including those that are intrinsic and extrinsic. Many reasons, both intrinsic and extrinsic, exist for learning. Moreover, both intrinsic and extrinsic reasons might be involved for a learning activity at the same time. As an example, some of the learners playing the Quest Atlantis educational computer game indicated that they completed Quests in the game both to get points and to help the fictitious Atlantian people. Likewise, they collected points to buy extrinsic items but at the same time mere availability of the points was a motivator as showing their development. Assuming the coexistence of intrinsic and extrinsic motivations is a very important theoretical standpoint; because it can change the scales used to measure motivation, and the conclusions based on data coming from these scales. As an example, Harter’s (1981) self-reporting scale, which is one of the most widely used scales for measuring motivation, assumes that intrinsic and extrinsic motivations are mutually exclusive. Therefore, while completing this scale a student has to be either intrinsically or extrinsically motivated for a learning activity. By using a modified version of this scale that allowed being intrinsically and extrinsically motivated at the same time Lepper, Sethi, Dialdin, and Drake (1997) found that both type of motivations could coexist. Eventually, even Lepper, whose taxonomy of motivation (Malone & Lepper, 1987) included just intrinsic factors, concluded that “… [S]uccess in school, as in many areas of life outside of school, may require us to attend simultaneously to both intrinsic and extrinsic sources of motivation” (Lepper & Henderlong, 2000, p. 295). This conclusion of Lepper and Henderlong (2000) came after a review of intrinsic and extrinsic motivation research within the past 25 years; however, this idea is not new. When the philosopher Plato conveyed the dialogues of Socrates and Glaucon about 25 hundred years ago, he mentioned the highest class where individuals would do tasks both for their own sake and for their results: Glaucon:

Socrates: Glaucon:

… How would you arrange goods – are there not some which we welcome for their own sakes, and independently of their consequences, as, for example, harmless pleasures and enjoyments, which delight us at the time, although nothing follows from them? I agree in thinking that there is such a class, I replied. Is there not also a second class of goods, such as knowledge, sight, health, which are desirable not only in themselves, but also for their results?

84

Socrates: Glaucon:

Socrates: Glaucon: Socrates:

H. Tüzün / Multiple Motivations Framework

Certainly, I said. And would you not recognize a third class, such as gymnastic, and the care of the sick, and the physician’s art; also the various ways of money-making –these do us good but we regard them as disagreeable; and no one would choose them for their own sakes, but only for the sake of some reward or result which flows from them? There is, I said, this third class also. But why do you ask? Because I want to know in which of the three classes you would place justice? In the highest class, I replied, –among those goods which he who would be happy desires both for their own sake and for the sake of their results. (Plato, The Republic, 357b-358c&d)

Our task then, as educators and researchers, is to utilize both intrinsic and extrinsic motivators to promote and support student learning. The framework of multiple motivations provides a useful framework for the coexistence of both kinds of motivations. Assertion 3: The use of playing and learning together is a strong motivator As Csikszentmihalyi (1990) points out, “One cannot enjoy doing the same thing at the same level for long. We grow either bored or frustrated …” (p. 75). It was discussed under the heading of “Duality of Activity” that when playing and learning elements are integrated, they produce unique activities. These unique activities eliminate or reduce the redundancy and the boredom in the learning process by providing sustained engagement. In the QA context, elements like back-story of the game, fictitious characters, use of points, multimedia elements, pushball game, immersive game context, and building in this context make it playful. It was found that learners in the high participating group, who engaged with most of these playful activities, undertook more Quests than the learners in the medium or low participating groups. After recognizing this assertion, one problem becomes that of figuring out what is play. Although there are many definitions of play, Fromberg’s (1992) characterization of play provides a useful explanation that includes all the play elements listed above. According to Fromberg (1992), play is symbolic, meaningful, active, pleasurable, voluntary, rule-governed, and episodic. One can notice that these characteristics of play also apply to the characteristics of learning in QA (unique, active, and meaningful), which suggests that even this kind of learner-centered learning can be playful without the play elements. When we examine traditional learning environments, we see a sharp distinction between playing and learning. As an example, school environments reflect a culture in which learning is treated as hard and serious. In such environments, play elements are excluded from learning, and most of the time playing is used as a separate and isolated reward after learning activities are completed (Silvern, 1998). Moreover, there is a concern among some educational researchers that when fun and entertainment are integrated into learning, learners will develop a new kind of attitude towards learning (see for example, Okan, 2003). These researchers fear that with this new kind of attitude, learners will despise the school and demand more enjoyable learning environments. Some other researchers on the other hand perceive this demand as a good thing. As an example, Prensky (2002) states that it is not the use of the Internet, distance learning, computers, wireless devices, computer-based learning, and e-learning that will revolutionize the learning in the 21st century. It is making learning fun and relevant, and therefore discarding the pain and suffering that accompanied it for so long, that will revolutionize it. Prensky (2002) predicts that after spending so much time playing with fun and engaging computer games, learners will demand these types of learning

H. Tüzün / Multiple Motivations Framework

85

environments, to the point that parents and teachers can no longer resist. Moreover, he envisions a future in which learners can get their degrees by choosing distributed accredited courses. Since the course content will be relatively the same among the courses with the same title, it will be the motivational elements of the course that will guide the learners towards choosing one of these courses. Although playing and learning together motivates learners and increases learner participation, providing a balance between playing and learning is crucial (Bergen & Fromberg, 1998). Besides, although this study frames how to increase learner motivation and participation, this might not be the ideal in every learning context. As an example, one of the teachers at a school where QA is implemented was concerned about the amount of time that is being spent by certain users on the bulletin boards (A mode of communication that enables asynchronous threaded discussions among Questers, it is similar to Usenet discussion groups) as opposed to educational activities. The whole purpose of Multiple Motivations Framework is to present ways for increasing such participation. Evidently, this is not preferred in all contexts and motivation should be considered with other factors of the learning context including learners, teachers, administrators, and parents. Assertion 4: Creativity is the new emerging “C” over traditional “4Cs” Traditionally, motivation in educational computer games and intrinsic motivation in general have been explained by 4Cs: challenge, curiosity, control, and context (Lepper & Henderlong, 2000). The results of this study showed that creativity is the new emerging “C” as a candidate for inclusion with the intrinsic motivators. Out of curiosity, when a Discriminant Function Analysis (DFA) was conducted to predict group membership from thirteen motivational categories, it was found that creativity was the only discriminant category that separated the high participating members of QA from the medium and low participating members. For that reason, creativity seems to be a very important construct in providing intrinsic motivation. There seem to be many definitions of creativity. For example, Sanders and Sanders (1984) cited various definitions of creativity given by leading educators and researchers. However, creativity defined in multiple motivations framework is closer to the spatial intelligence defined in Gardner’s theory of multiple intelligences (Armstrong, 1993; Gardner, 1993). This kind of intelligence includes perceiving the spatial and visual nature of the world, and the ability to perform transformations in it. A spatially intelligent person can shape and mold images in the world, either through physical means such as building, drawing, molding, sculpting, and inventing, or through mental means such as rotations and transformations (Armstrong, 1993). Children have used materials such as Lego bricks, wooden blocks, constructo straws, clay, pipe cleaners, and lasy blocks in the past to exercise their spatial intelligence (Forman, 1998). The 3D virtual worlds are the new frontiers for the utilization of digital objects for the same purpose. To understand the relationship between spatial intelligence and building activities, it is helpful to present some information from neurophysiology. The left side of the brain is responsible for analytical, logical, and verbal abilities. This side controls cognition and language in people. The right side is responsible for imagery, intuitive thinking, and spatial relationships. In the development of the right side of the brain and in fostering creativity, it is essential to practice imagery information (O’Neil, Abedi, & Spielberger, 1994). Building activities present such an opportunity toward practicing imagery and spatial information. While building, children participate in constructive

86

H. Tüzün / Multiple Motivations Framework

play in which they create symbolic patterns, real world objects, working systems, and sequences of actions (Forman, 1998). This kind of spatial intelligence requires a context that is conducive to creativity. In such a context, people first observe the aesthetics of the materials such as shape, line, space, volume, balance, light and shade, color, pattern, and harmony. Then they examine artifacts created by others. Eventually, they become artists themselves producing these artifacts (Armstrong, 1993). Two further examples illuminate the importance of creativity for sustained motivation in computer games. There was a time when a virtual world, called “Sandbox,” was created in QA per request of our remote collaborators in Denmark. These collaborators used the Sandbox world for building activities in alignment with their curriculum. Because of technical issues, the world was allowed to be entered and built in it by all Questers. It was assumed by QA designers that just Questers in Denmark would use this world. After the need of the Denmark collaborators was over, the Sandbox world was closed. However, it was apparent from many angry inquiries that this world was actually discovered by other Questers and used for building activities. The QA team members received many questions asking why the Sandbox world was not open for building anymore. Another example comes from the data collection site of this study. After my longitudinal daily observations were completed, I kept on visiting the club on different occasions. In one of these visits, I observed that the computer game “Roller Coaster Tycoon” was just installed on all lab computers. The purpose of this simulation game is to design and manage an amusement park, keep its guests happy, and increase the park profit. Most of the building phenomenon in this game is similar to building in QA. For example, while building rides and attractions players use pieces from the game’s library. Although the ideas that can be created are limited to just rides and attractions, the final completed product is a working system. For example, after building a roller coaster track, players can put a roller coaster on it, let the virtual guests ride it, and observe different data of the ride, like the speed of the ride and the thoughts and feelings of its riders. On the day of my visit, both boys and girls were playing this game with great engagement. To see if this interest was due to its novelty, I kept on visiting the club that week and on subsequent weeks. Not surprisingly, this interest has been high long after the game was introduced in the lab. Much of the interest towards this game came from the building activities in the game. Assertion 5: Choice is in the foundations of all motivators The curious reader might wonder about the core category found during the selective coding stage. The availability of choices in an educational computer game is the core category emerged in the study. Prior research supports that even a small amount of choice has the potential to motivate children (see for example, Cordova & Lepper, 1996). Interestingly, the “choice” code was not available after obtaining the codes at the end of the open coding process. The emergence of a core category from other categories and overall data is proof of the fact that I stood closer to the emergent nature (Glaser, 1992) of data analysis during the constant comparison method of grounded theory. When Papert (1980) talks about his LOGO programming language, designed for children, he conveys a personal story. Papert fell in love with car parts when he was two years old. His obsession was so high that he knew most of the concepts like the gearbox, the transmission system, and the differential. Later when he grew up, he prac-

H. Tüzün / Multiple Motivations Framework

87

ticed with these parts, and specifically with gears. He discovered the cause and effect relationships in the gear systems. He believes this experience with gears later helped him when he learned mathematics. For example, while solving equation problems with two variables (e.g., 5x + 3y = 12), he made a mental gear model of the relation between x and y. Overall, he had a love relationship in addition to a deep understanding of the gears. Therefore, his interest in gears cannot be reduced to just cognitive terms. This experience of Papert was a personal experience, and therefore it cannot be expected that many other children will like gears. Papert (1980) however, argues that computers have so much capability to simulate and are so flexible that they “can take on a thousand forms and can serve a thousand functions, [they] can appeal to a thousand tastes” (p. viii). Therefore, computers can be used as flexible instruments in which every child can find her/his gear, as long as the context does not stifle the child. The availability of choices in a computer game is what gives its flexibility. An individual has the best knowledge about the self; therefore, by using the choices in the game the individual has the ability to stretch the learning process based on her/his personal interest and taste. In the context of QA the choices are many, and the availability of choices in dualities and categories of framework of multiple motivations is a proof for this. For example, the introvert learners can participate in activities alone while the social learners can join the crowd. While participating in the learning activities, they can enrich the process with playful elements. If the learner finds these elements somewhat childish, she/he can trim, or minimize, these elements and focus entirely on learning. When doing a Quest they can read the Quest description and purposes, or they can listen to its narration. After completing the activities they can enjoy the achievement of overcoming these challenges, or they can get a reward for their effort. In choosing a reward, further options are available; materialistic souls can satisfy the cravings of a materialistic nature with trading cards, t-shirts, pencils, stickers, or other contextual items. In their social relations they can share information and activities with others; or they can compete over these activities. When interacting with others, they can choose different communication modes from among chat, e-mail, telegrams, and discussion boards. They can also use the immersive game context for exploring, for interacting with objects, for building, or for transactions. They can perceive the game points as an indicator of their development, or they can use them as an exchange currency in an open market environment. The back-story of the game can be learned through an animation, but further formats are available for different styles; in the form of a comic book for visual enthusiasts or in the form of a novel for those who like reading. These choice examples can go on for many other elements of the game. Furthermore, most of these choices do not have to be mutually exclusive. For example, while a learner may prefer handling the activities alone, the same learner can take on social relationships to overcome activities which are not possible, or are very difficult, to handle alone. This issue points to the fluidity of human nature. Human beings might be prone to changes in their preferences, interests, and tastes as the result of their physical, cognitive, and social development, conditions and constraints of the context, and by other factors. For this reason, multiple motivations framework contradicts the findings of Cordova and Lepper (1996) in which they found the personalization of the learning process motivational. Since human nature is changeable, so much personalization might create a state where old and new interests clash, which in turn might prevent learners’ coming back to the learning environment. The explicit availability of choices in the learning environment is the key for providing continued learner motivation.

88

H. Tüzün / Multiple Motivations Framework

8. Conclusion This study proposed an emergent explanation of motivation, “Multiple Motivations Framework.” The content and form of this framework were explicated by using the Quest Atlantis educational computer game. The framework needs to be amended and extended with further studies so that it can become more comprehensive. One way for doing this would be to replicate this study in different contexts. The data for this study was collected within an after-school context. In this sense, while playing with the educational computer game learners were not exposed to the limitations of a traditional school context, like a strict curriculum and deadlines. It would be fitting to replicate this study in a school context with such constraints to investigate the extent to which the results match or differ. This is one of the ways to improve and add to the framework. Other than by amendment and extension, future studies might verify the validity and preferability of the framework. For this purpose, horizontal studies involving other educational computer games need to be conducted. In addition, further vertical studies need to be conducted for different contexts, audiences, and conditions. Bandura (1986) acknowledged that “any theory of motivation must consider a large set of interactive processes if it is to provide an adequate explanation of human behavior” (p. 243). This statement has long been ignored in research regarding motivation, probably for the reason Bruner (1973) stated: “How one manages to time the steps in pedagogy to match unfolding capacities, how one manages to instruct without making the learner dependent, and how one manages to do both of these while keeping alive zest for further learning – these are very complicated questions that do not yield easy answers” (p. 122). As a result, motivation studies have focused on piecemeal factors to explain human motivation. On the other side, this study revealed the large set of interactive processes as a whole and proposed the Multiple Motivations Framework to provide an adequate explanation of human motivation. Traditional motivation studies have typically relied upon quantitative methods, including one time data collection through surveys. In addition, tasks whose meaning were not strategically aligned with the context were offered to measure motivation in most of these studies. The qualitative methods used in this study provided a very different perspective than what is available in understating motivation. I strongly recommend to future researchers of motivation the use of ethnographic methods, making prolonged observations in the research context, and observing learners in their naturalistic learning contexts. Sternberg, Kaufman, and Pretz (2002) have presented a descriptive taxonomy called “propulsion model of creative contributions.” They suggest that creative contributions propel a field in some way. They identified eight kinds of creative contributions which might propel a field: 1. 2. 3. 4.

5.

In replication a field stays where it is. In redefinition the current status of the field is seen from a new perspective. In forward incrementation the field is moved in the direction in which it is already moving. In advance forward incrementation the field is moved in the direction in which it is already moving, but beyond where others are ready for the field to move. In redirection the field is moved to a new direction.

H. Tüzün / Multiple Motivations Framework

6. 7. 8.

89

In reconstruction/redirection the field is moved back to where it was so that it can be moved to a new direction. In reinitiation the field is moved to a different starting point and then the field is moved in a different direction from that point. In integration many past contributions of the field, that were viewed as distinct, are put together.

This study replicated the conclusions of previous research on motivation in finding that the constructs of curiosity, control, choice, fantasy, achievement, and rewards motivated these learners. It advanced the field in finding that the availability of choice options to learners was more important than previously thought. It reinitiated the field in that creativity, identity of learners, social relations, and active learning were proposed as important constructs in providing motivation. And most importantly it integrated many past contributions in the field that were perceived as distinct, such as intrinsic and extrinsic motivators, playing and learning, and achievement and rewards into a coherent framework of motivation. I hope that these creative contributions move the conceptual understanding and practice of motivation positively. I also hope that the framework will be improved with progressive, analytical critiques by interested practitioners and scholars in the field.

Note This material is based upon author’s dissertation study, titled “Motivating Learners in Educational Computer Games,” Dissertation Abstracts International, 65(05), (Publication No. AAT 3134052). This material is based upon work supported by a CAREER Grant from the National Science Foundation REC-9980081 and by a National Science Foundation Grant #0092831. Correspondence about this article be addressed to Dr. Hakan Tüzün, Hacettepe Üniversitesi, Bilgisayar ve Öğretim Teknolojileri Eğitimi Bölümü (BÖTE), 06532, Beytepe/Ankara, Turkey. (90-312) 297-71 76.

References [1] Adler, P.A., & Adler, P. (1987). Membership roles in field research. Newbury Park, CA: Sage. [2] Armstrong, T. (1993). 7 Kinds of smart: Identifying and developing your many intelligences. New York, NY: Plume. [3] Atkinson, J.W., & Feather, N.T. (1966). A theory of achievement motivation. New York, NY: Wiley. [4] Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, IN: Prentice Hall. [5] Barab, S.A., Arici, A., & Jackson, C. (2005). Eat your vegetables and do your homework: A designbased investigation of enjoyment and meaning in learning. Educational Technology, 45(1), 15–21. [6] Barab, S.A., MaKinster, J.G., Scheckler, R. (2003). Designing system dualities: Characterizing an online professional development community. In S.A. Barab, R. Kling, & J. Gray (Eds.), Designing for virtual communities in the service of learning. Cambridge, MA: Cambridge University Press. [7] Barab, S.A., Thomas, M.K., Dodge, T., Carteaux, B., Tuzun, H., (2005). Making learning fun: Quest Atlantis, a game without guns. Educational Technology Research and Development, 53(1), 86–107. [8] Barab, S.A., Thomas, M.K., Dodge, T., Squire, K., & Newell, M. (2004). Critical design ethnography: Designing for change. Anthropology and Education Quarterly, 35(2), 254–268.

90

H. Tüzün / Multiple Motivations Framework

[9] Bergen D., & Fromberg, D.P. (1998). Emerging and future contexts, perspectives, and meanings of play. In D.P. Fromberg (Ed.), Play from birth to twelve and beyond: Contexts, perspectives, and meanings (pp. 537–541). New York, NY: Garland. [10] Bers, M. (2001). Identity construction environments: Developing personal and moral values through the design of a virtual city. The Journal of the Learning Sciences, 10(4), 365–415. [11] Brugman, D., & Beem, A.L. (1986). Classroom climate and continuing motivation. In J.H.L. Van Den Bercken, E.E.J. De Bruyn, & Th.C.M. Bergen (Eds.), Achievement and task motivation (pp. 147–165). Alblasserdam, Netherlands: Offsetdrukkerij Kanters B.V. [12] Bruner, J.S. (1973). The relevance of education. New York, NY: Norton. [13] Cameron, J., Banko, K.M., & Pierce, W.D. (2001). Pervasive negative effects of rewards on intrinsic motivation: The myth continues. The Behavior Analyst, 24, 1–44. [14] Cameron, J., & Pierce, W.D. (1994). Reinforcement, reward, and intrinsic motivation: A meta-analysis. Review of Educational Research, 64, 363–423. [15] Cameron, J. & Pierce, W.D. (2002). Rewards and intrinsic motivation: Resolving the controversy. Westport, CT: Bergin and Garvey. [16] Clark, R.E. (1983) Reconsidering research on learning from media. Review of Educational Research, 53(4), 445–459. [17] Clifford, J., & Marcus, G. (Eds.). (1986). Writing culture: The poetics and politics of ethnography. Berkeley, CA: University of California Press. [18] Cordova, D.I., & Lepper, M.R. (1996). Intrinsic motivation and the process of learning: Beneficial effects of contextualization, personalization, and choice. Journal of Educational Psychology, 88(4), 715–730. [19] Creswell, J.W. (1998). Qualitative inquiry and research design: Choosing among five traditions. Thousand Oaks, CA: Sage. [20] Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York, NY: Harper and Row. [21] De Charms, R. (1968). Personal causation. New York, NY: Academic Press. [22] Deci, E.L. (1971). Effects of externally mediated rewards on intrinsic motivation. Journal of Personality and Social Psychology, 18, 105–115. [23] Deci, E.L. (1972). The effects of contingent and non-contingent rewards and controls on intrinsic motivation. Organizational Behavior and Human Performance, 8, 217–229. [24] Deci, E.L. (1975). Intrinsic motivation. New York, NY: Plenum Press. [25] Deci, E.L., Koestner, R., & Ryan, R.M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125, 627–668. [26] Deci, E.L., Koestner, R., & Ryan, R.M. (2001). Extrinsic rewards and intrinsic motivation in education: Reconsidered once again. Review of Educational Research, 71(1), 1–27. [27] Eisenberger, R., & Cameron, J. (1996). Detrimental effects of reward: Reality or myth? American Psychologist, 51, 1153–1166. [28] Fetterman, D.M. (1998). Ethnography: Step by step (2nd ed.). Thousand Oaks, CA: Sage. [29] Fielding N.G., & Fielding, J.L. (1986). Linking data. Beverly Hills, CA: Sage. [30] Finn, J.D. (1989). Withdrawing from school. Review of Educational Research, 59, 117–142. [31] Finn, J.D., Folger, J., & Cox, D. (1991). Measuring participation among elementary grade students. Educational and Psychological Measurement, 51, 393–402. [32] Forman, G. (1998). Constructive play. In D.P. Fromberg (Ed.), Play from birth to twelve and beyond: Contexts, perspectives, and meanings (pp. 392–400). New York, NY: Garland. [33] Fromberg, D.P. (1992). Play. In C. Seefeldt (Ed.), The early childhood curriculum: A review of current research (pp. 42–84). New York, NY: Teachers College Press. [34] Gardner, H. (1993). Multiple intelligences: The theory in practice. New York, NY: BasicBooks. [35] Garris, R., Ahlers, R., & Driskell, J.E. (2002). Games, motivation, and learning: A research and practice model. Simulation & Gaming, 33(4), 441–467. [36] Gilbert, N. (Ed.). (1993). Researching social life. London: Sage. [37] Glaser, B.G. (1992). Emergence vs forcing: Basic of grounded theory analysis. Mill Valley, CA: Sociology Press. [38] Glaser, B.G., & Strauss, A.L. (1967). The discovery of grounded theory: Strategies for qualitative research. Chicago, IL: Aldine. [39] Good, T.L., & Brophy, J.E. (1997). Looking in classrooms (7th ed.). New York, NY: Longman. [40] Greenspan, R. (2002). Gaming industry serious business. Retrieved September 07, 2003, from http://cyberatlas.internet.com/big_picture/hardware/article/0,1323,5921_1492881,00.html. [41] Harter, S. (1981). A new self-report scale of intrinsic versus extrinsic orientation in the classroom: Motivational and informational components. Developmental Psychology, 17, 300–312.

H. Tüzün / Multiple Motivations Framework

91

[42] Hedegaard, M., Chaiklin, S., & Jensen, U.J. (1999). Activity theory and social practice: An introduction. In S. Chaiklin, M. Hedegaard, & U.J. Jensen (Eds.), Activity Theory and Social Practice (pp. 12–30). Aarhus: Aarhus University Press. [43] Keller, J.M. (1983). Motivational design of instruction. In C.M. Reigeluth (Ed.), Instructional design theories and models (pp. 383–433). Hillsdale, NJ: Lawrence Erlbaum. [44] Krendl, K.A., & Broihier, M. (1991). Student responses to computers: A longitudinal study. Chicago, IL: Annual Meeting of the International Communication Association. (ERIC Document Reproduction Service No. ED332250). [45] Kruglanski, A.W., Friedman, I., & Zeevi, G. (1971). The effects of extrinsic incentive on some qualitative aspects of task performance. Journal of Personality, 39, 606–617. [46] Lepper, M.R. & Henderlong, J. (2000). Turning “play” into “work” and “work” into “play”: 25 years of research on intrinsic and extrinsic motivation. In C. Sansone & J.M. Harackiewicz (Eds.), Intrinsic and extrinsic motivation: The search for optimal motivation and performance (pp. 257–307). San Diego, CA: Academic Press. [47] Lepper, M.R., Greene, D., & Nisbett, R.E. (1973). Undermining children’s intrinsic interest with extrinsic rewards: A test of the “overjustification” hypothesis. Journal of Personality and Social Psychology, 28, 129–137. [48] Lepper, M.R., & Malone, T.W. (1987). Intrinsic motivation and instructional effectiveness in computer-based education. In R.E. Snow & M.J. Farr (Eds.), Aptitude, learning and instruction: Conative and affective process analyses (pp. 255–286). Hillsdale, NJ: Lawrence Erlbaum Associates. [49] Lepper, M.R., Sethi, S., Dialdin, D., & Drake, M. (1997). Intrinsic and extrinsic motivation: A developmental perspective. In S.S. Luthar, J.A. Burack, D. Cicchetti, & J.R. Weisz (Eds.), Developmental psychopathology: Perspectives on adjustment, risk, and disorder (pp. 23–50). New York, NY: Cambridge University Press. [50] Lincoln, Y.S., & Guba E.G. (1985). Naturalistic inquiry. Newbury Park, CA: Sage. [51] Malone, T.W. (1980). What makes things fun to learn? A study of intrinsically motivating computer games. Dissertation Abstracts International, 41(05), 1955B. (UMI No. 8024707). [52] Malone, T.W., & Lepper, M.R. (1987). Making learning fun: A taxonomy of intrinsic motivations for learning. In R.E. Snow & M.J. Farr (Eds.), Aptitude, learning and instruction: Conative and affective process analyses (pp. 223–253). Hillsdale, NJ: Lawrence Erlbaum Associates. [53] Malouf, D.B. (1988). The effect of instructional computer games on continuing student motivation. Journal of Special Education, 21(4), 27–38. [54] Mark, R. (1996). Research made simple. Thousand Oaks, CA: Sage. [55] Martin, B.L., & Briggs, L.J. (1986). The affective and cognitive domains: Integration for instruction and research. Englewood Cliffs, NJ: Educational Technology Publications. [56] Maslow, A.H. (1987). Motivation and personality (3rd ed.). New York, NY: Longman. [57] Merriam, S.B. (1998). Qualitative research and case study applications in education: Revised and extended from case study research in education (2nd ed.). San Francisco, CA: Jossey-Bass Inc. [58] Mussino, A. (1999). Conceptual and operational problems in measuring participation in sports. Paper presented at the International Statistical Institute Conference, Helsinki, Finland. Retrieved September 30, 2003, from http://www.stat.fi/isi99/proceedings/arkisto/varasto/muss0846.pdf. [59] O’Neil, H.F., Jr., Abedi, J., & Spielberger, C.D. (1994). The measurement and teaching of creativity. In H.F. O’Neil Jr. & M. Drillings (Eds.), Motivation: Theory and research (pp. 245–263). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers. [60] Okan, Z. (2003). Edutainment: Is learning at risk?. British Journal of Educational Technology, 34(3), 255–264. [61] Papert, S. (1980). Mind-storms: Children, computers, and powerful ideas. New York, NY: Basic Books, Inc. Publishers. [62] Patton, M.Q. (1987). How to use qualitative methods in evaluation. Newbury Park, CA: Sage. [63] Plato (1901). The republic (B. Jowett, Trans.). New York, NY: P.F. Collier & Son. [64] Prensky, M. (2002). The motivation of gameplay or, the real 21st century learning revolution. On The Horizon, 10(1), 5–11. [65] Reigeluth, C.M. (1999). What is instructional-design theory and how is it changing? In C.M. Reigeluth (Ed.), Instructional-design theories and models, Volume 2: A new paradigm of instructional theory (pp. 5–29). Mahwah, NJ: Lawrence Erlbaum. [66] Rogers, E.M. (1995). Diffusion of innovations. New York, NY: Free Press. [67] Sanders, D.A., & Sanders, J.A. (1984). Teaching creativity through metaphor. An integrated brain approach. New York, NY: Longman. [68] Schwartz, D.L., Lin, X., Brophy, S., & Bransford, J.D. (1999). Toward the development of flexibly adaptive instructional designs. In C.M. Reigeluth (Ed.), Instructional-design theories and models, Volume 2: A new paradigm of instructional theory (pp. 183–214). Mahwah, NJ: Erlbaum.

92

H. Tüzün / Multiple Motivations Framework

[69] Silverman, D. (1993). Interpreting qualitative data: Methods for analyzing talk, text, and interaction. London, Sage. [70] Silvern, S.B. (1998). Educational implications of play with computers. In D.P. Fromberg (Ed.), Play from birth to twelve and beyond: Contexts, perspectives, and meanings (pp. 530–536). New York, NY: Garland. [71] Skinner, B.F. (1953). Science and human behavior. New York, NY: Macmillan. [72] Steiner, E. (1988). Methodology of theory building. Sydney, Australia: Educology Research Associates. [73] Sternberg, R.J., Kaufman, J.C., & Pretz, J.E. (2002). The creativity conundrum: A propulsion model of kinds of creative contributions. New York, NY: Psychology Press. [74] Stone, L. (1997). Virtually yours: The internet as a social medium. Vision, April 14, 1997. Retrieved December 11, 2003, from http://research.microsoft.com/vwg/papers/VISION.htm. [75] Strauss, A., & Corbin, J. (1994) Grounded theory methodology: An Overview. In N.K. Denzin, & Y.S. Lincoln (Eds.), Handbook of qualitative research (pp. 273–285). Thousand Oaks, CA: Sage. [76] Strauss, A., & Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Thousand Oaks, CA: Sage. [77] Turkle, S. (1995). Life on the screen: Identity in the age of the internet. New York, NY: Simon & Schuster. [78] Weiner, B. (1990). History of motivational research in motivation. Journal of Educational Psychology, 82, 616–622. [79] Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. Cambridge, Cambridge University Press.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

93

An Instructional Design/Development Model for the Creation of Game-Like Learning Environments: The FIDGE Model Göknur Kaplan AKILLI 1 and Kürsat CAGILTAY Department of Computer Education and Instructional Technology, Middle East Technical University, Turkey

Abstract. Computer games are considered as powerful tools to learning and they have a potential for educational use. However, the lack of available comprehensive design paradigms and well-designed research studies about the question of “how to” incorporate games into learning environments is still a question, despite more than thirty years’ existence of computer games and simulations in the instructional design movement. Setting off from these issues, a formative research study is designed to propose an instructional design/development model, which may be used for creation of game-like learning environments. Depending on the results and with the inspiration from fuzzy logic, an instructional design/development model for creating game-like environments, which is called as “FIDGE model” is proposed. Keywords. Games, simulations, game-like learning environments, instructional design/development (IDD), instructional design/development model (IDDM), formative research, fuzzy logic

1. Introduction One of the possible novelties regarding the methods of education, which should be discussed, is the use of games. As a matter of fact, games are not so much a novelty in this field, as young human beings, by nature, begin to learn through games and playing from their early childhood [1]. At the older ages, games are replaced by formal education. Nevertheless, the transition from informal games to formal education environment does not always, and especially nowadays, seem to be a sharp one as it is known that games are being used also in some educational environments, yet their success is questionable. When one looks deeper into the subject, it is understood that the use of games in education is not so much a novelty too, because its history may be traced back a few thousand years [2]. It is now known that even in times before history, games and dramatic performances as representations of real life were more effective as teaching tools than the presentations of life itself. In our modern day, with the new technological advancement of societies, traditional games of old times have been replaced by electronic games and in similar manner, dramatic representations of old have been transformed 1 Corresponding Author: Göknur Kaplan Akıllı, Department of Computer Education and Instructional Technologies, Middle East Technical University, İnönü Bulvarı 06531, Ankara, Turkey; E-mail: [email protected].

94

G.K. Akilli and K. Cagiltay / The FIDGE Model

into role-playing in simulation environments. Hence, electronic games and simulations have been parts of contemporary formal education. However, such transformation cannot be expected to take place quite smoothly and without its problems.

2. Problem Statement Traditional instructional design/development models have been criticized on the grounds that they hardly represent a variety of structure, although they are abundant in number. The procedural stratifications and time-consuming practices have constituted the main rationale of these criticisms. On the other hand, although computer games and simulations have a history of more than three decades in the instructional design movement and computer games are considered as powerful tools to increase learning [2,3]; there are two major problems that are encountered. One is that there are no available comprehensive design paradigms and the other is the lack of well-designed research studies [4]. While the literature is growing as time passes, by the carbon copy researches that report perceived student reactions preceded by vague description of games and simulations or comparative studies of simulations versus regular classroom instruction [4], the question of how to incorporate games into learning environments rather than, simply, to master the material, is much more frequently asked to the educational researchers [2,5].

3. Significance of the Study Computer games and simulations have appeared on the scene of instructional design/development (IDD) activities more than three decades ago; yet, the literature still lacks available comprehensive design paradigms and well-designed research studies [4]. Although there is vast number of instructional design/development models (IDDMs), which reveal answers of “how to” questions for various learning environments and situations and an accumulating mass of research about the perceptions of the students and their reactions, effects of games on learning and various skills, and even about the illustrations of such environments, the question of how to incorporate games into learning environments stays unresolved. Reigeluth and Frick mention that although the existing design theories have not reached perfection; there is need for more theories and models that will guide us through human development and related additional kinds of learning, where for different kinds of situations those utilize new information technologies as tools, via telling us “how to do” [6]. For this “how to do” question, the three studies within researchers’ reach were proposing basic design guidelines and principles for creating game-like learning environments rather than a model [7–9]. In conclusion, there is the apparent and urgent need for the introduction of an IDDM that will help and guide instructional designers and/or game designers for the efficient use of games and simulations in educational environments, more precisely to create game-like learning environments.

G.K. Akilli and K. Cagiltay / The FIDGE Model

95

4. General Overview and Current State of Knowledge Education has always been considered as one of the most productive breeding-grounds for technology where it would find its finest resonances and thus would have revolutionary effects. However, the use of technology in educational environments does not presently seem to have contributed significantly to the realization of these expectations [10,11]. The most important reason may be the insufficiency of current models and methods to meet the consequences of the paradigm shift from Industrial Age to Information Age [12,13]. The latter statement is also valid for the use of games and simulations in education. Games and simulations are often referred to as experiential exercises [4], in which there is ‘learning how to learn’ [14]. However, they have hardly become a part of IDD movement until early 1970s, despite their entrance to educational scene in the late 1950s [4]. There are two major problems that we faced with, when educational use of games and simulations is of concern, non-existence of available comprehensive design paradigms and the lack of well-designed research studies [4]. Despite existence of many IDDMs, which reveal answers of “how to” questions for various learning environments and situations, there is no trace about the presence of a such model with the exception of the model that the current study proposes, which can be used for creating game-like learning environments. 4.1. Instructional Design/Development During the review of the relevant literature, we faced with the interchangeable use of instructional design; instructional systems design (ISD); instructional development; and even instructional technology, which was also asserted by many researchers [15–18]. Even though several attempts have been made to derive standardized definitions and terms [17,16,19], the results have not been adopted and used in the literature. Thus, a unified term of “instructional design/ development (IDD)” will be used throughout the study. However, it would be better to assert that we considered the definition of instructional development given by Reigeluth, which is “concerned with understanding, improving and applying methods of creating [italics added] instruction” [18, p. 8], optimizes the process of developing the instruction and encompasses analysis, design, development, implementation, and formative and summative evaluation activities [18,17]. Winn and Jonassen et al. put criticisms about the positivist basis of ID models [20,21]. Both disapproved the linear design process assumes the predictability of human behavior, the closure and isolatedness of learning situations, responsibility of instructor than the learner for learning and ignores the dynamic, complex and non-linear nature of the design processes, contextualness of learning environments, differences among learners, metacognitive abilities of learners, unstable, elusive and complex nature of human consciousness. As alternative approaches that can be employed for the improvement of IDD process, various researchers offer various suggestions. Jonassen et al. suggest adapting new sciences, such as hermeneutics, fuzzy logic and chaos theory [21]. Reigeluth suggests customized, learner-centered and social-contextual design conducted by userdesigners [22,13], which is also articulated by Winn’s matched timing of design and use of instructional material [20] and Winn’s statement of necessity to get help from Human Computer Interaction discipline [24]. Lastly, Hoffman offers plasticity and

96

G.K. Akilli and K. Cagiltay / The FIDGE Model

modularity [23], as a result of linking Reigeluth ‘s Elaboration Theory and hypermedia [18]. There are further suggestions, however, among these issues only fuzzy logic which is first suggested by Jonassen et al. will be explained within the scope of the current study [21]. 4.2. Fuzzy Logic Fuzzy logic is based on the idea that reality can rarely be represented accurately in a bivalent manner. Rather, it is multivalent, having many in-between values, which do not have to belong to mutually exclusive sets. It implies for IDD that behavior could only be understood probabilistically, using continua, rather than binary measures and integration of problematic areas such as student perceptions of the efficacy of the educational program into the design. More specifically, set-theoretic facet of fuzzy logic also implies the non-linear, dynamic IDD phases, which has “fuzzy” rather than strict boundaries. Moreover, since the sequence of events within a project depends on human decisions, which are based on approximate reasoning of human beings, fuzzy logic can be well-applied to IDD process. Instead of having strictly bounded and sequenced phases, having intertwined phases, which have flexible and fuzzy boundaries would be more advantageous in that it would allow designers to move freely in between phases throughout the entire IDD process. Jonassen et al. state that the more one moves away from deterministic approaches to thinking and designing toward more probabilistic ways of thinking, the more useful it becomes in providing methods for assessing in “real-life” issues, where things are not black-and-white, but rather any number of different shades of color across the spectrum [21]. Jonassen et al. further state that it is impossible to predict, let alone describe, what will happen in learning situations due to elusive and complex nature of human consciousness [21], which is also consistent with Winn’s opinion that although instructional designers would like them to do otherwise, people think ‘irrationally,’ and reason ‘implausibly’ [24]. Both of these statements support the main definition of fuzzy logic. However, both researchers’ studies lack more specific facets of fuzzy logic, which had been coined first by Lotfi A. Zadeh in 1960s, but remained concealed until it was discovered in the late 1980s [25]. For the current study, the researchers were especially interested in the set-theoretic facet of fuzzy logic, which is concerned with fuzzy sets, whose boundaries are not sharply defined [26,27, ¶5]. The set-theoretic facet of fuzzy logic is also the initial focus of the development of fuzzy logic, which gave birth to applications such as fuzzy arithmetic (also known as “computing with words” [26, p. 1]) fuzzy topology, etc. Moreover, by fuzzification, which is the process of replacing the concept of a set with that of a fuzzy set, it becomes possible to provide a way of constructing models or theories that are more general and more reflective of the imprecision of the real world than the models or theories in which the sets are assumed to be sharply bounded and definitely limited. Briefly, any concept, method or theory can be generalized to a reflection of the real world via fuzzification. This was exactly what the researchers wanted, i.e. the proposal of an alternatively structured instructional design/development model against the traditional instructional development models that have been criticized for their linear structures, procedural stratifications and time-consuming practices.

G.K. Akilli and K. Cagiltay / The FIDGE Model

97

4.3. Games and Simulations Games and simulations are often referred to as experiential exercises [4], in which there is ‘learning how to learn’ [14]. Turkle further contended that it provides more than thinking; beyond thinking [14]. Specifically, Prensky defines games as “organized play” [8, p. 119]. Heinich, Molenda, Russell, and Smaldino define game as “an activity, in which participants follow prescribed rules that differ from those of real life as striving to attain a challenging goal” [28, p. 10]. Dempsey, Rasmussen and Lucassen define gaming in a basic sense as “any overt instructional or learning format that involves competition and is rule-guided” [3, p. 4]. According to Prensky simulations and games differ in that, “simulations are not, in and of themselves games. They need all the additional structural elements – fun, play, rules, a goal, winning, competition, etc.” [8, p. 212]. Depending on these definitions and characteristics, as an attempt to derive a general term, the researchers will use game-like learning environments, which is defined as ‘authentic or simulated places, where learning is fostered and supported especially by seamless integration of motivating game elements. As for theories that inspired the game design, “Flow Theory of Optimal Experience” developed by Mihaly Csikszentmihalyi [29] and “Activity Theory” developed by Alexey Leontiev, a student of Lev Vygotsky could be recognized [30]. Moreover, there are some principles to be taken into consideration proposed by Cerny and John [31]. Yet, there seems to be hardly any design models except for the instructional design/development model tailored for the creation of game-like learning environments, which is called the FIDGE model [32]. 4.3.1. Effects of Games and Simulations on Learning Although the literature about games and simulations is accumulating day by day, the issue of whether games influence students’ learning in a positive way is still vague. For instance, Molenda and Sullivan state that among problem solving and integrated learning systems, games and simulations were the least used technology applications in education [10]. However, there exist studies that put forth effects of games and simulations on discovery learning strategies, problem solving skills and computer using skills; their effects on students’ intellectual, visual, motor skills and about the engagement and interactivity which are important for learning environments. 4.3.2. Educational Use of Games and Simulations There is evidence that the use of games as instructional tools dates back to 3000 B.C. in China [2]. Nevertheless, games and simulations have hardly become a part of instructional design movement until early 1970s, despite their entrance to educational scene in the late 1950s [4]. Seels and Richie report that in those times audio-visual specialist saw the potential of games and simulations but not of video (or likewise electronic) games [16]. Rieber argues that growing technological innovations provide opportunities of interactive learning environments that can be integrated with the theories of learning [1]. However, Prensky further claims that, instruction through neither CAI, nor web based technologies contributes to learning, rather they subtract. People do not want to be included in such learning “opportunities” offered via innovative technologies, but they have to, since these learning “opportunities” possess still the same boring content and

98

G.K. Akilli and K. Cagiltay / The FIDGE Model

same old fashioned strategy as traditional education [8, p. 92–93]. Prensky puts forth that learning can best take place when there is high engagement and he proposes “digital-game-based learning,” which has potential for achievement of the necessary “high learning” through “high engagement” [8, p. 149]. He states that high engagement, interactive learning process and the way the two are put together will guarantee the sound working of digital game-based learning works [8]. 4.3.3. Design Models for Educational Use of Games and Simulations The literature is reviewed to search for design models that will help and guide instructional designers especially to design game-like learning environments, “which requires the ability to step outside of a traditional, linear approach to content creation—a process that is counter-intuitive to many teachers.” [33, ¶15]. However, it seems there is lack of IDD models, in turn there are various design principles and lessons learned from commercial game designs. For instance, Amory, et al. identified game elements that students found interesting or useful within different game types, which were the most suitable for their teaching environment and presented a model that links pedagogical issues with these identified game elements [7]. Furthermore, Prensky presents various principles for good computer game design and other important digital game design elements [8]. The other recent study on the subject that the design and research team currently works on is the “Games-to-Teach” project carried by Massachusetts Institute of Technology (MIT). This study also proposes design principles for successful games design [9].

5. Methodology This study is designed as a special methodology that is similar to that of a case study method of qualitative research [34], which is referred as “formative research” by Reigeluth and Frick [6, p. 633]. Formative research methodology fits best, while the researchers were interested in the design and development process rather than the product or outcomes, in contextual structure rather than specific variables and in discovery of the underlying elements rather than conformation [35]. Moreover, the reason that the case, which is selected by the researchers, is not especially and intentionally designed according to a specific model, but fulfils the same goals and provides the same context as researchers intention, leads the design of the study towards naturalistic case of formative research. Participants of the study were selected by using convenience and purposive sampling and consisted of 18 out of 56 senior undergraduate students of the Computer Education and Instructional Technologies (CEIT) Department of the Middle East Technical University in Turkey, with the age range of 20 to 24. Participants, who were fourth year students, enrolled in an educational game design course in their department, were invited to participate in a study that was designed to suggest a new IDDM that might be used for creation of game-like learning environments, utilizing a 3D virtual world tool on the Internet. The researchers collected data during this IDD process, duration of which was three months. As for the data collection and analysis techniques, observations, documents (such as weekly reports, peer evaluations, e-mail logs and reflective papers about the specific aspects of the selected case written by the participants), and semi-structured interviews

G.K. Akilli and K. Cagiltay / The FIDGE Model

99

with the participants (students), instructors (of the selected case); and content analysis were used. These data sources are the ones that formative research implies.

6. Findings The findings of the research is summarized under three groups, which are completely invented and were determined by the researchers themselves, depending upon the data, to scrutinize the phenomena much more easily and comprehensively. In reality, the collected data showed an inter-relational and fuzzily-bounded nature. The findings were grouped as the soft issues, process-related issues, and hard issues related to creation of game-like learning environments. Soft issues are the peopleware part, or in other words, the human relations and social/organizational issues of an instructional design/development (IDD) process. The findings related with these soft issues can be summarized as the following: 1. 2.

3. 4.

IDD requires teamwork. Team members’ characteristics and qualifications are important (i.e. field knowledgeableness, proficiency in technology, strategic, holistic and especially creative thinking abilities, project management skills, leader qualifications, communication skills, responsibility, honesty, empathy, professionality, high-level programming knowledge and advanced coding skills are essentially required). Quality and qualifications of the team members affect the quality of an instructional system. Beforehand determined cautions and ways for resolutions to possible conflicts are essential.

As its name suggests, process-related issues are the findings related to analysis, design, development and evaluation steps of an IDD process, i.e. the findings that formed the core of the model. These can be summarized as: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

It is impossible to omit or ignore any of the ID phases. Do not isolate ID phases from each other strictly. Do not conduct ID phases in a linear sequence. Conduct ID phases parallel to and within each other. Transitions between phases should be flexible. Turning back to a phase should be possible. Tendency to think of and create a utopia in the design phase. Flexibility and modularity of the product is important. Progressing via prototypes is very effective. Evaluation should be fused into the whole ID process. Formative evaluations are of great importance. Guidance is needed in each step.

Lastly, hard issues are the technical aspects of this IDD process, which can be summarized as the following: 1.

Technical problems, which are mostly originated from the tool 1.1. usability (hard-to-use, hard-to-adapt)

100

G.K. Akilli and K. Cagiltay / The FIDGE Model

PRE-ANALYSIS

EVALUATION

ANALYSIS

DESIGN AND DEVELOPMENT

Figure 1. General structure of the FIDGE model.

1.2. manipulation (even the administrator of tool has limited permission, log on option for one of the group members at a time) 1.3. limitations/restrictions (mostly related to 3D component of the tool) 1.4. location (difficulty of carrying the program to another location other than the computer labs in the department) 1.5. interaction features of the tool (difficulty of finding putting or exporting appropriate objects or animations to 3D area) 1.6. server-related problems (difficulty of uploading their works to the server via FTP) 1.7. speed 1.8. similarity (possibility to find out the corresponding features and similar sides among the tools that were previously used) 2.

Location and communication problems in reaching target audience 2.1. lack of permission from the school administration 2.2. difficulties to conduct evaluations 2.3. difficulty in describing the project to the other people

3.

Time limitations

7. The FIDGE Model 7.1. General Overview of the FIDGE Model In the current study, the researchers suggest an appropriate and evolving IDDM for the creation of game-like learning environments based on the findings. In accordance with many of the traditional models, this model also consists of four parts, which are analysis, design, development and evaluation. Additionally, it possesses an additional phase; the “pre-analysis” phase. However, the components of these parts and the way they are structured are different from the traditional models. It consists of dynamic phases, which have fuzzy boundaries and through which the instructional designers progressed in a non-linear manner. Indeed, the reason of these characteristics is the basis of the model that is cultivated on the fuzzy logic concept, which also leads to a different visualization of the model rather than the traditional “boxes-and-arrows” approach (see Fig. 1). Another reason is that the model is proposed depending directly on actual and

G.K. Akilli and K. Cagiltay / The FIDGE Model

101

concrete data collected from real-life practices. What is more, the researchers also coined the below presented model’s name regarding these attributes, especially the non-linearity and fuzziness emerged from the findings. It is called as “FIDGE” model, which is the acronym that stands for “Fuzzified Instructional Design Development of Game-like Environments” for learning. According to the Oxford English Dictionary [36], “fidge” as a verb means “to be eager and restless; to express pleasurable eagerness by restless movements,” which is also consistent with the impatience that anybody shows when playing a game is of concern. Before scrutinizing the details of the proposed model, it would be better to emphasize the two general patterns that were dominant within the findings. First one is the contexts, in which IDD takes place, and, in which the product attained at the end of IDD process will be used. The second one is the attributes of this IDD process. The first pattern is both a contributor and the by-product of the IDD process. It seriously affects the quality of the product and IDD process itself, and is in turn affected by the socio-organizational needs and cultural issues, which appeared during the IDD process, such as the need for a leader who will lead the rest of the team throughout the process, and the necessity to avoid acting with their emotions and feelings. As for the context, in which the product attained at the end of IDD process will be used, the findings revealed the importance of the appropriateness of the product regarding the socioeconomic status and the abilities of the learner with consideration of the cultural issues. For the second general pattern, the “must-be” non-linearity and dynamism throughout the IDD process; the fuzziness among each step of IDD process; and lastly, features originated and inherited from games and simulations were asserted by the participants. Throughout the entire process, all the participants had to make modifications and revisions in their plans and actions that they took to overcome the problems and obstacles, by means of continuous evaluation. Findings have indicated that it is impossible to omit or ignore any of the analysis, design, and development phases and instead of isolating these phases from each other strictly and conducting them in a linear sequence, it would be better to conduct them parallel to each other and in an intertwined manner. Lastly, findings of study revealed that all of the participants used some features peculiar to games and simulations. For instance, findings of the study provided traces of unique features peculiar to simulations, such as non-linear event sequence, intertwined consequences of action-reaction chains, and dynamic set(s) of relationships changing with respect to the actions that the user took. As for the games, findings of the study indicated the use of game characteristics, especially in the design phase of IDD process, such as challenge, fantasy, curiosity and control given to the learners that contribute specifically to motivation and thus eager learning. Moreover the findings also pointed out other features of games and simulations, such as engagement, interactivity and active participation. The use of popular culture elements with the above mentioned elements was another issue revealed by the findings of the study. There are two sets of principles that underlie the model, which are related to some socio-organizational issues for the design team and to the instructional design/development process itself. The first set of principles, which is related to peopleware (soft) and technical (hard) issues, is as follows: 1.

Form a multidisciplinary and multi-skilled team including an experienced game-player.

102

G.K. Akilli and K. Cagiltay / The FIDGE Model

Formative evaluations with subject-matter experts

Selection of a tentative target group PRE-ANALYSIS

EVALUATION

Formative evaluations with tentative group representatives

Selection of a tentative subject

LR

Selection of tentative goals Tool Analysis Game Analysis

Figure 2. The visualization of pre-analysis phase of the FIDGE model.

2. 3. 4. 5. 6.

Provide common standards about the work done. Identify and develop awareness and need for an instructional system, and create mechanisms for motivation. Meet the need for a leader and a guide. Establish good communication strategies and create active involvement. Manage, plan and schedule time.

The second set of principles, which is related to the whole instructional design/development process itself, is as follows: 1. 2. 3. 4. 5. 6. 7.

Dynamic, non-linear and fuzzy phases. Early decisions about the utilities and restraints of the technology, which will be used throughout the project. Analogous, participatory design and prototypes. Support from the literature. Continuous and iterative evaluation and synthesis. Focus on the modularity and flexibility of the product. Creativity.

Lastly, before moving on the details of the model, it would be better to emphasize that the components of the model are intertwined through each other and sometimes conducted in parallel among the phases, despite their exhibition as a listing. 7.2. Pre-Analysis Phase of the FIDGE Model The reason for the existence of this phase is to provide a starting point for the instructional designers (see Fig. 2). However, if there is no need for such a warm-up period, this phase could be skipped. All of the following issues are tentative and could be easily changed when the instructional designers begin to conduct analysis phase: 1. 2.

Determine and specify a tentative target group. Select a tentative subject regarding your target group and depending on your previous experiences.

G.K. Akilli and K. Cagiltay / The FIDGE Model

103

EVALUATION

Design foundations of game-like learning

LR

Learner Analysis

LR

Content Analysis

Tool Analysis

LR

Instructional Approach

Needs Analysis

General goals

Context Analysis Risk Analysis

Cost Analysis (if needed)

Time Planning

Self-Analysis

Formative evaluations

Scenario Preparation

Updating- and Maintenance-related

Game Analysis Tentative sketches illustrating an example product User guide / Help / Technical support ANALYSIS

Figure 3. The visualization of analysis phase of the FIDGE model.

3.

4. 5. 6. 7.

Conduct a small literature review (LR) to find evidence, whether your tentative subject fits or is likely to be fit for creation of game-like learning environment, or not. Specify the tentative goals of your design according to the selected subject and regarding your target group. Take the opinions and recommendations of the subject-matter experts, and a representative group of the tentative target group via interviews. Begin to explore and analyze the development tool/software. Begin to analyze different games to: 7.1. Differentiate different game genres such as strategy, adventure, sports, etc. 7.2. Find out game utilities such as multiplayer, collaboration, communities etc. 7.3. Find out which game genre is appropriate for which subject matter e.g. strategy games are appropriate for social sciences. 7.4. Find out game elements such as the use of pirates, magic, history, etc. 7.5. Find out the appropriateness of the instructional approach in relation to game genre, game utilities, and game elements.

7.3. Analysis Phase of the FIDGE Model In this phase, needs analysis, learner analysis, context analysis, content (or task) analysis, cost analysis (if needed), risk analysis, an analysis to adjust the duration and the frequency of the system for effective use, and a self-analysis should be conducted, while the tool and game analyses which began in the previous phase will continue (see Fig. 3). Moreover, instructional approach and its implications should be specified and a time planning activity should be done.

104

G.K. Akilli and K. Cagiltay / The FIDGE Model

The components are as follows: 1.

2.

3.

4.

5.

Needs analysis: 1.1. potential stakeholders’ (e.g. teachers’, students’, parents’, etc.) attitudes toward computers 1.2. potential stakeholders’ opinions about computer use 1.3. potential stakeholders’ expectations from simulations and games 1.4. why the target group should use simulations and games 1.5. the insufficiencies and gaps in the topic/ subject stated by the target group (continued in content analysis) 1.6. support from literature 1.7. obtain or create sketches illustrating a completed product should to be shown to the target group or experts as an example, in order to be able to transform an abstract concept into a tangible one for them. According to the conducted needs analysis, the general goals of the project will be constituted; more precisely, the needs will be transformed into the general goals of the project. Learner analysis (conducted parallel to needs analysis and includes real observations, surveys, structured or semi-structured interviews with the actual target group and time schedule for all these): 3.1. Actual target group’s background information, i.e. characteristics, attributes, skills, prior knowledge, and specific entry competencies 3.2. Support and help from the literature Context analysis (conducted parallel to needs and learner analyses): 4.1. Actual learners’ perspectives about the attributes of a game-like learning environment, in which they would learn the designated content. 4.2. The role of the teacher or instructor. 4.3. The amount of the learner control. 4.4. Examination of the computer infrastructure (fulfillment of the necessary and sufficient conditions, specification of minimum system requirements to work out the prepared program, identification of the hardwarerelated issues). 4.5. Specification of the socio-economic status of the learners (in relation with their computer literacy, in order to determine at which grade the program will be used). 4.6. Begin to lay design foundations. Content (task) analysis (conducted parallel to needs, learner and context analyses; and affected by tool analysis): 5.1. Efficiency assessment that stood for the optimum amount of content in a limited amount of time. 5.2. Checking the accuracy of the content (formative evaluation with experts and learners). 5.3. Verifying the topics included in the content through various resources (literature review). 5.4. Taking students’, subject matter experts’ and experienced instructional designers’ opinions (formative evaluation). 5.5. Carrying out step-by-step reduction (regarding the limitations and boundaries of the tool and the structure (or nature) of the tool) (iterative cycles of formative evaluation).

G.K. Akilli and K. Cagiltay / The FIDGE Model

5.6.

105

Setting the structure of the content (regarding the limitations and boundaries of the tool and the structure/nature of the tool). 5.7. Creating the main elements of the scenario. 5.8. Synthesis of the collected opinions of the students and the experts; the elements included in the content; and the instructional designers’ own opinions (to provide intact objectivity). 6. Tool analysis (began in the pre-analysis and will be continued in analysis). 6.1. The tool’s structure/nature: 6.2. What its uses are. 6.3. How it is used. 6.4. What its limitations and utilities are. 6.5. Students’ or learners’ viewpoints and reactions to the tool. 6.6. Investigation for alternatives to the selected tool/ technology. 6.7. Thinking of suggestions about updating and maintenance of the system regarding tool analysis. 6.8. Thinking of the issues concerning the guidance for and support to the user, such as ‘help,’ or ‘technical support.’ 7. Game analysis (began in the pre-analysis and will be continued in analysis. it is conducted parallel to tool analysis and in relation to learner, content, context analyses and instructional approach): 7.1. Think of game utilities such as multiplayer options, collaboration, online virtual communities. 7.2. Think of which game genre is appropriate for your instructional design/development. 7.3. Think of the game elements for your instructional design/development. 7.4. Specify the appropriateness of the instructional approach in relation to game genre, game utilities, and game elements of your instructional design/development. 8. Instructional approach (selected regarding the structure of the content, tool’s nature/structure): 8.1. Selection of instructional approach (e.g. discovery learning, scenariobased learning, problem-based learning or a hybrid approaches, which are offspring of two or more different approaches). 8.2. The implications of the selected instructional approach to the design. 8.3. Assurance of the selected instructional approach’s capacity and aptitude for the application of game-elements. 8.4. Adjustment of the instruction’s duration and frequency for effective use. 8.5. Formative evaluation to take the opinions of the learners and the experts. 9. Self-analysis for each instructional designer in the design team: 9.1. Find out the needs, characteristics and skills that are lacking, but should be possessed by the members in the design team. 9.2. Specify strategies to gain them. 10. Risk analysis: 10.1. Envision of potential risks. 10.2. Outline of a “panic room” plan against these foreseen risks. 10.3. Cautions both to avoid and to solve possible problems.

106

G.K. Akilli and K. Cagiltay / The FIDGE Model

EVALUATION Synthesis Formative evaluations M o t i v a t i o n Challenge

Curiosity

Fantasy

Content Analysis

LR

Assessment

Feedback

Attention

Scenario Preparation

Updating- and Maintenance-related Issues

Prototypes

General goals

Context Analysis

Instructional Approach

Design foundations of game-like learning environment

Interaction & Engagement

Control

Tentative sketches illustrating an example product User guide / Help / Technical support DESIGN/DEVELOPMENT

Figure 4. The visualization of design-development phase of the FIDGE model.

11. Time schedule: 11.1. Time arrangement for group meetings. 11.2. Time arrangement for meetings with the designated experts. 11.3. Time arrangement for meetings with learners from the target group. 12. Preparation for design: 12.1. Specifying design foundations of game-like learning environment 12.2. Writing down the tentative design decisions as the design team envisions them at the moment, i.e. a very general overview of design. 12.3. Writing down the ‘scenario bits,’ such as the main theme of the scenario, the characters, etc. 12.4. Drawing or providing tentative sketches illustrating an example product (The design team could find an example or create one to be used in formative evaluations mainly to give an idea to the interviewees about how the completed product would look like). 7.4. Design-Development Phase of the FIDGE Model In this phase, scenario preparation; content clear-cuts; specification of motivation, attention, feedback, and learning assessment elements; preparation of user-help; creation of prototypes; preparation of rating scales, checklists and interview guides for formative evaluations; design of orientations; and insurance of usability issues, product’s modularity and flexibility will be conducted (see Fig. 4).

G.K. Akilli and K. Cagiltay / The FIDGE Model

107

It would be better to emphasize that the implementation phase of the traditional models is contained in the intertwined design and development phase of the FIDGE model. This phase encompasses: 1.

Preparation of more than one scenario, namely alternative scenarios for the game-like learning environment and selection of the most appropriate one: 1.1. Regarding the content analysis. 1.2. Regarding the selected instructional approach. 1.3. Regarding the boundaries of the tool. 1.4. Utilizing the team member’s wide experiences as a game player.

This gives an opportunity to switch to another scenario, in case the scenario prepared at the beginning failed to be implemented. 2.

3.

4.

Preparation of the scenario-related components: 2.1. Setting of the scenario (prototypes should be prepared, to be used to take feedback from the learners and ID experts continuously (formative evaluation)). 2.2. Plot structure of the scenario (a typical use case should be written, which also provides guidance for the usage of the prepared program). 2.3. A flowchart regarding the scenario (the flowchart should inherit the content’s structure and should be framed by the tool’s limitations). Content clear-cuts (as an extension of the content analysis in the analysis phase): 3.1. Clarify the content in its brief, intertwined and clear-cut form using the continuing step-by-step reductions and modifications via iterations and feedbacks from the team members; subject matter experts, ID experts; and learners (formative evaluation). 3.2. Make necessary changes in the goals, content analysis and the flowchart according to these modifications and reductions. Specification of motivation, attention, feedback, and learning assessment elements for creation of game-like learning environments: 4.1. Utilize the essential elements of many commercial games possess, such as curiosity, challenge, fantasy and control given to the learner to give his/her own decisions. 4.2. Utilize elements from popular culture. 4.3. Pay attention to the relatedness of motivation elements with the feedback and attention components. 4.4. Feel free to combine feedback and the learning assessment elements. 4.5. Pay attention to interaction and engagement elements, which are also peculiar to games, related to the feedback, motivation and content. 4.6. Include activities to provide learners’ active engagement. 4.7. Enrich the social aspect of the interaction provided via the program you are designing, to help your learners to establish a virtual community, or to give the feeling of togetherness. 4.8. Specify collaboration utility and climate of your game-like learning environment by describing how the users will collaborate, what utilities they will have when collaborating each other (e.g. will they share the tasks, complete only their own part of the task and merge each completed part at the end or each of them will complete the task on their own and use collaboration utility of your program to get help from each other?).

108

G.K. Akilli and K. Cagiltay / The FIDGE Model

The employment of formative evaluations to specify the details and components of the above is also important and helpful. 5.

6.

Preparation of user-help which should be conducted parallel to tool analysis including: 5.1. The issues concerning guidance for user such as ‘help’. 5.2. The issues concerning support for user, such as ‘technical support’. Creation of prototypes (in relation to the above mentioned analyses and scenario) 6.1. Paper-based prototypes preparation. 6.2. Computer-based prototypes preparation.

These prototypes are used to take feedback from the learners, experts and team members, about both the user-interface design and the overall design itself. They are also likely to reveal the above-given issues about the motivation, attention, feedback and the learning assessment elements of the design. All the way through these feedbacks, it is likely that the details of the ‘user-help,’ or ‘technical support will also emerge. 7.

Preparation of rating scales, checklists and interview guides for formative evaluations by: 7.1. Including items about the arrangement, presentation, appropriateness, consistency of the content. 7.2. Including items about the general appearance, appropriateness and consistency of the user-interface. 7.3. Including items about the extent to which the program appeals to the user. 8. Design of orientations, such as: 8.1. An orientation about the program to avoid misunderstandings. 8.2. A more general orientation to acquire target audience with the basic computer literacy and game-related skills (such as adjustment of an environment, in which the learners could play a simple game to acquire the gamelogic and gain basic eye-hand coordination). 9. Insurance of usability issues throughout the whole phase by: 9.1. Keeping in mind that everything should be user-centered and check the program’s status accordingly, since users would be the ones who would use the product. 9.2. Being aware of usability issues and employing them in the first place. 10. Assurance of product’s modularity and flexibility by providing as much flexibility and modularity as possible for the final product, so that the need of a radical change, which might emerge following the formative evaluations, could easily be conducted. Lastly, each of the above-mentioned elements should be supported by the literature. It should also be emphasized that final user interface could not even bear a resemblance to the initial one envisioned at the beginning. What is more, it is a possibility to be confronted with a very different version of the program, compared to the previously visualized design at the beginning.

G.K. Akilli and K. Cagiltay / The FIDGE Model

109

Figure 5. The visualization of evaluation phase of the FIDGE model.

7.5. Evaluation Phase of the FIDGE Model Although the related elements of the evaluation phase were presented in the above phases, it would be better to give the general structure and main elements of the evaluation phase. Evaluation phase has three main elements, which are formative evaluation; summative evaluation and the synthesis (see Fig. 5). Instructional designers should never forget that evaluations and feedback taken from the learners should be continuous and should begin as soon as they started with the pre-analysis (or alternatively, analysis) phase. Before conducting evaluations, the instructional designers should clarify the issues, such as, by whom the product would be evaluated, how they would be reached, where the evaluations would be conducted. For data collection, instructional designers should employ rating scales, checklists and the interview guides that would be already prepared in the design/development phase. Instructional designers should conduct the formative evaluations frequently and with shorter intervals throughout the instructional design/development process and should employ them while determining the foundation stones of the instructional design/development process. Instructional designers should conduct formative evaluations with the team members, their peers, learners in their target group and various experts of various professions; however, as stated in the previous parts, the learners representing the variance of the target group should be in the first place. This also puts forth the usability test that should be conducted within the evaluation phase. In the synthesis part, as its name implies, instructional designers should make a synthesis and interpretation of all the data collected from the evaluations, related literature and their own comments, when making the final decisions about the project. For these purposes, after each evaluation, collected data should be analyzed; common points should be noted and discussed with the other instructional designers in the team. As for the summative evaluation, it will be used to evaluate the instructional system as a whole. However, summative evaluation is not as critical as it was for the traditional model, since there will not be much left, due to the continual formative evaluations conducted throughout the design-development phase.

110

G.K. Akilli and K. Cagiltay / The FIDGE Model

Table 1. Summary of the FIDGE Model. Issue

Its Property

Participants

All of actively participating learners and experts

Team

Multidisciplinary, multi-skilled, game-player experience

Environment

Socio-organizational, cultural

Process

Dynamic, non-linear, fuzzy, creative, enriched by games’ and simulations’ elements (fantasy, challenge, etc.)

Change

Continuous, evaluation-based

Evaluation

Continuous, iterative, formative and summative, fused into each phase

Management

Need for a leader and a well-planned and scheduled time management

Technology

Suitable, compatible

Use

By (novice /expert) instructional designers and educational game designers for game-like learning environments and educational games

7.6. Summary of the FIDGE Model To summarize, the FIDGE model is a real-life originated model that has a dynamic, non-linear and fuzzy structure and is enriched by unique features of games and simulation, which combines the context with peopleware throughout the instructional designdevelopment process. The proposed model might be used in creation of educational games as well as in creation of game-like learning environments. The researchers think that the proposed model might be appropriate to be used by both novice and expert users. The existence of the “pre-analysis” phase of the model is the most apparent evidence that this model addresses the novice instructional designers’ needs. Other evidence is that they are the affiliates of the so called “game generation,” who are different from many of us in various aspects and possess differentiating characteristics and skills resulting from different experiences and the “new media society” [8, p. 65; 37]. However, they also lack sufficient instructional design experience, which would have impact on their use of this model, both positively and negatively. The probable positive effects are their untouched creativity and ingenious design habits. The probable negative effects would be the difficulty in understanding the model or misinterpretations, which would result in void and ineffective design practices. As for the expert instructional designers, the researchers believe that this model might widen their visions and help them catch up with the current trends and changes of the coming generation. Lastly, among the limitations of the model, the probable inheritance from the selected case and the complex and complicated nature of the model could be asserted. The elements that constituted the model seemed to be affected by the IDD model used in the case. For instance, the use of prototypes was inherited from the “rapid prototyping” model [38]. However, it would be meaningless to strive naming this concept in another way or to eliminate it, since it was found to be very useful. As for the complex and complicated nature of the model, it could be said that this is the first impression

G.K. Akilli and K. Cagiltay / The FIDGE Model

111

and it would be much easier to use this model due to its more flexible structure and fitting nature to human reasoning than other traditional models. In conclusion, it should be kept in mind that all these issues should be further verified by the follow-up studies. It would be hardly possible to clarify the uses, users and the limitations of the model, earlier than the conduction of such follow-up studies. References [1] Rieber, L.P. (1996). Seriously considering play: Designing interactive learning environments based on the blending of microworlds, simulations, and games. Educational Technology Research and Development 44(2) pp. 43–58. [2] Dempsey, J.V., Lucassen, B.A., Haynes, L.L., & Casey, M.S. (1996). Instructional applications of computer games. In J.J. Hirschbuhl & D. Bishop (Eds.), Computer Studies: Computers in education (8th ed., pp. 85–91). Guilford: Annual Editions. [3] Dempsey, J.V., Rasmussen, K. & Lucassen, B. (1996). The Instructional Gaming Literature: Implications and 99 Sources (Tech. Rep. No: 96–1). USA, Alabama: University of South Alabama, College of Education. Retrieved May 30, 2002, from http://www.coe.usouthal.edu/TechReports/ TR96_1.PDF. [4] Gredler, M.E. (1996). Educational games and simulations: a technology in search of a (research) paradigm. In D.H. Jonassen (Ed.) Handbook of Research for Educational Communications and Technology. Missouri: Simon and Schuster. [5] Dede, C. (1996). The evolution of constructivist learning environments: immersion in distributed, virtual worlds. In B.G. Wilson (Ed.), Constructivist learning environments: Case studies in instructional design (pp. 165–175). New Jersey: Educational Technology Publications. [6] Reigeluth, C.M. & Frick, T.W. (1999). Formative research: methodology for creating and improving design theories. In C. M. Reigeluth (Ed.) Instructional-design theories and models (Volume II) A new paradigm of instructional theory (pp. 633–652). Mahwah, NJ: Lawrence Erlbaum Associates. [7] Amory, A., Naicker, K., Vincent, J. & Adams, C. (1999). The use of computer games as an educational tool: Identification of appropriate game types and game elements. BJET, 30(4), 311–321. [8] Prensky, M. (2001). Digital game-based learning. New York: McGraw-Hill. [9] Massachusetts Institute of Technology (MIT), (2003). Design principles of next-generation digital gaming for education. Educational Technology, September-October 2003, 17–22. [10] Molenda, M. & Sullivan, M. (2003). Issues and trends in instructional technology: Treading water. In Fitzgerald, M.A., Orey, M., and Branch, R.M. (Eds.) Educational Media and Technology Yearbook 2003. Englewood, CO: Libraries Unlimited. [11] Russell, T. (2003). The “No Significant Difference Phenomenon”. Available at: http://teleeducation.nb. ca/nosignificantdifference (Retrieved on January 10, 2003). [12] Bates, A.W. (2000). Managing technological change. Strategies for college and university leaders. San Francisco: Jossey-Bass. [13] Reigeluth, C.M. (1999). What is instructional-design theory and how is it changing?. In C.M. Reigeluth (Ed.) Instructional-design theories and models (Volume II) A new paradigm of instructional theory (pp. 5–29). Mahwah, NJ: Lawrence Erlbaum Associates. [14] Turkle, S. (1984). Video Games and Computer Holding Power. In The Second Self: Computers and the Human Spirit. New York: Simon & Schuster, Inc. [15] Shrock, S.A. (1995). A brief history of instructional development. In G. Anglin (Ed.), Instructional technology: Past, present, and future (2nd ed. pp. 11–19). Engelwood, CO: Libraries Unlimited. [16] Seels, B. & Richie, R. (1994) Instructional technology: The definitions and domains of the field. Washington: AECT. [17] Gustafson, K.L., & Branch, R.M. (1997). Survey of instructional development models (3rd Ed.). Syracuse, NY: ERIC Clearinghouse on Information Resources. (ED 411 780). [18] Reigeluth, C.M. (1983). Instructional design: What is it and why is it? In C.M. Reigeluth, (Ed.) Instructional-design theories and models: An overview of their current status (pp. 3–36). New Jersey: Lawrence Erlbaum Associates. [19] Schiffman (1995). Instructional systems design: Five views of the field. In G. Anglin (Ed.), Instructional Technology: Past, present, and future (2nd ed. pp. 131–144). Engelwood, CO: Libraries Unlimited. [20] Winn, W. (1997). Advantages of a Theory-based Curriculum in Instructional Technology. Educational Technology, (37)1, 34–41.

112

G.K. Akilli and K. Cagiltay / The FIDGE Model

[21] Jonassen, D.H., Hennon, R.J., Ondrusek, A., Samouilova, M., Spaulding, K.L., Yueh, H.P., et al. (1997). Certainty, determinism, and predictability in theories of instructional design: Lessons from science. Educational Technology, 37(1), 27–33. [22] Reigeluth, C.M. (1996). A new paradigm of ISD? Educational Technology, (36)3, 13–20. [23] Hoffman, S. (1997). Elaboration theory and hypermedia: Is there a link? Educational Technology, 37(1), 57–64. [24] Winn, W. (1996). Cognitive perspectives in psychology. In D.H. Jonassen (Ed.) Handbook of research for educational communications and technology: A project of the Association for Educational Communications and Technology (pp. 79–112). New York: Macmillan Library Reference. [25] Dubois, D., Foulloy, L., Galichet, S. & Prade, H. (1999). Performing approximate reasoning with words?. In L.A. Zadeh & J. Kacprzyk (Eds.), Computing with words and information/intelligent systems 1 (pp. 24–49). Heidelberg: Physica-Verlag. [26] Zadeh, L.A. (1999). Fuzzy logic = Computing with words. In L.A. Zadeh & J. Kacprzyk (Eds.), Computing with words and information/intelligent systems 1 (pp. 1–23). Heidelberg: Physica-Verlag. [27] MIT Encyclopedia of Cognitive Science. (2003). Fuzzy Logic. Retrieved June 25, 2003, from http://cognet.mit.edu/MITECS/Entry/zadeh.html. [28] Heinich, R., Molenda, M., Russell, J.D. & Smaldino, S.E. (2002). Instructional media and technologies for learning (7th ed.). New Jersey: Merrill Prentice Hall. [29] Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York: Harper Perennial. [30] Kaptelinin, V. & Nardi, B.A. (1997). Activity theory: Basic concepts and applications. Retrieved October 11, 2003, from http://www.acm.org/sigchi/chi97/proceedings/tutorial/bn.htm/. [31] Cerny, M. & John, M. (2002, June). Game development. Myth vs. method. Game Developer. [32] Akilli, G.K. (2004). A proposal of instructional design/development model for game-like learning environments: The FID2GE model. Unpublished master’s thesis, Middle East Technical University, Ankara. [33] Morrison, J.L. & Aldrich, C. (2003, September/October). Simulations and the learning revolution: An interview with Clark Aldrich. The Technology Source. Retrieved from August 11, 2003, from http://64.124.14.173/default.asp?show=article&id=2032 on. [34] Yin, R.K. (1996). Case study research: design and methods. (2nd ed.) Thousand Oaks, CA: Sage Publications. [35] Merriam, S.B. (1998). Qualitative research and case study applications in education. (2nd ed.) San Francisco: Jossey-Bass. [36] Oxford English Dictionary (2nd edition) [Online version] (1989). Retrieved January 9, 2003 from http://dictionary.oed.com/cgi/entry/00084385/. [37] Calvert, S.L. & Jordan, A.B. (2001). Children in the digital age. Applied Developmental Psychology, 22, 3–5. [38] Tripp, S., & Bichelmeyer, B. (1990). Rapid prototyping: An alternative instructional design strategy. Educational Technology Research and Development, 38 (1), 31–44.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

113

Learning when Using Commercial Computer Games as Simulations: A Case Study Using a Simulation Game Preston P. PARKER Utah State University, College of Education and Human Services Department of Instructional Technology, Logan, UT 84321 [email protected]

Abstract. As educators and trainers continue to turn to technology, they are constantly seeking to find ways to make interventions quicker, better, and cheaper. This case study looks at using an off-the-shelf commercial computer game, Age of Empires II, as a simulation to facilitate learning Multimedia Production Management. Age of Empires II was not created for this purpose. The possibilities of implementing a structure—mapping elements of the game to elements in real-life— are explored. The participants of this case study felt that the intervention helped them learn elements of Multimedia Production Management. Keywords. Computer games, simulations, simulation games, learning, training

1. Introduction In recent years, as technological advances have proliferated the amount of media involved in peoples’ daily lives, there has been a great focus on understanding its effects. Parents, politicians, and scholars have questioned, investigated, and discovered a wide variety of media effects, from increased aggression, desensitivity, and time displacement to improved skills, understanding, and communication. One effect of great interest and debate is the discussion of using media to improve learning [1]. Since the days of World War II—when using audio/visual media to facilitate learning really began—educators, trainers, and instructional designers have sought to improve upon their theories, methods, and applications regarding facilitating learning involving media [2,3]. In the last decade, an increasing number of researchers have recognized the social impact of pre-existing, off-the-shelf, computer-mediated games [4–6]. One exciting possibility, gaining interest among scholars and practitioners, is using these types of media for the purpose of facilitating learning [7]. These off-the-shelf, prepackaged, “video games” (hereafter referred to as “commercial computer games”) are meant for entertainment. These are the games that are commonly found in an electronics section of a distribution outlet or retail vendor. Using these commercial computer games as part of a learning intervention is appealing because it is believed that doing so might be cheaper, require less time, and be more motivating than self-produced, “in-house” computerized (artificial, virtual, and augmented) learning environments which are geared toward achieving a specific learning objective.

114

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

Commercial computer games are now readily accessible and commonplace, which leads to the comparatively low cost of running an activity over and over. They can be quickly accessed and “mapped” to real life situations. In essence, a single game could be used for many different situations and learning objectives. Organizations desiring to train personnel might not need to invest the time and money in creating specific computer-mediated learning environments. Instead, they would be able to take an environment, such as a commercial computer game, and build a structure around it to fit their specific learning objective. We now live in a world full of “digital natives”—those who grew up in a technology rich environment—who are ready and, in many ways, expecting learning interventions to be highly technical and game-like [8]. At an increasing pace, researchers and practitioners are investigating how a commercial computer game might be used in a learning intervention. The U.S. Defense Department is investigating how war games and flight simulating games might be used [9]. The Navy, in particular, is intrigued by games because it is recognizing that sailors are accustomed to “vivid, paced digital experiences” [10]. City planners are interested in the ability of the game SimCity to represent a real city planning environment [11,12]. Teachers are trying ways to incorporate historical games, like Civilization III, when teaching social studies [13]. And, business professors are attempting to teach management techniques using war games [14]. There is much to explore in this area.

2. Background Theorists, researchers, and practitioners disagree as to the effectiveness of using computer-mediated games (commercial or otherwise) in an intervention for enhancing learning. One of the understood benefits is that a computer-mediated game allows the learner to go through a trial and error process. As Winn and Windschitl [15] put it, “the identification of errors, and the opportunity to correct them, are necessary strategies in complex learning environments.” (p. 12). In games, learners are engaged in an environment where they are allowed to go beyond the boundaries of real-life with few, if any, negative consequences. Games “are contrived and controlled situations, extensions of group awareness that permit a respite from customary patterns.” (p. 243) [16]. By doing activities outside of normal patterns, learners can discover where the boundaries are in a certain concept and when the concept falls apart. An example is pilots learning to fly in flight simulation games where they are allowed to tilt the rudders too high or too low and discover the consequences. Thus, learners are more likely to try out new ideas and push boundaries that they may not otherwise pursue. There are those who believe that using a game adds to engagement, involvement, and motivation and therefore, by extension, improves the learning process [8]. Other investigators believe that when a game is used as a theme (say, playing a golfing game for learning accounting) that the game is actually a distraction from learning the content. In this sense, the learners’ cognitive load is at least doubled by having to learn the game, learn the content, and how the one maps to the other. These investigators believe that having a game just as a theme, or illustration, in the hopes that it will motivate someone, and thereby improving the learning process, is ludicrous and should be avoided [17]. This is because such “gratuitous illustrations make little or no instructional contribution and are often ignored or may interfere with efficient learning.” Merrill [17] states that the greatest motivation for learning is learning itself and that

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

115

once a person knows what they will be able to do after the intervention (and has an opportunity to demonstrate this new knowledge), there will be sufficient motivation to improve the learning process. However, Merrill [18] also argues that there can be a need for teaching abstractions out of context—when students are learning general principles and adapting these principles to individual cases, this can increase overall efficiency and uniformity. Other investigators believe that giving learners similar experiences to that found in real life, improves learning results without increasing cognitive load [19]. This is because learners are engaged in an activity similar to real life where they can construct their own knowledge and skills, while gaining a “feel” for what they are learning [20]. Learners can create meaning while involved in the practices and contexts that are similar to the real life practices and contexts [21]. Indeed, some claim that knowing and doing are one and the same, that learning is situated and developed through activity [22]. Therefore, learning is more than just gaining an understanding. It is creating an “increasingly rich implicit understanding of the world in which [learners] use the tools and of the tools themselves (p. 33) [22]. Computer-mediated simulations and games can lend themselves to such contextual-based learning experiences. 2.1. Gaming, Interactivity, and Learning Commercial computer gaming is an $18 billion worldwide enterprise that many scholars see as being a dominant form of entertainment in the upcoming decades [23]. Jim Gee [4] states that games are not only pushing the creative boundaries of interactive digital media but also suggesting powerful models of next-generation interactive learning environments. There is a basic, perhaps erroneous, assumption that interactivity is uniformly desirable and associated with positive outcomes. This assumption exists on a social level, psychological level, public-policy level, and cognitive level. Furthermore, an assumption exists that, in many respects, a learner interacting with a live person can be replicated using technology. Moreover, it is thought that technology and media can offer many advantages that a live interaction cannot, such as exact repeatability (being able to perform the same action over and over), ability to review (a recording can be made that can be viewed later), and the ability to handle much more content and data than a human can. Perhaps if a learning plan incorporating media offered many choices, frequently enough, which mattered or made a difference [24], learners would be more successful. Perhaps if a learning plan incorporating media provided a means of two-way communication whereby the responses between the learner and the computer built off of the previous responses [25] learners would be more motivated. For trainers interested in harnessing the power of computer games to support learning [8], there is the challenge of how to account for both the person-tool interaction and the person-person interaction—the broader social contexts in which gaming is situated and where game meanings are created [26]. This study is an attempt to offer qualitative data in support of and/or refuting some of these assumptions about learning and interactivity. 2.2. Computer-Mediated Environments There are three different categories of computer-mediated environments that are used in learning interventions: those that are simulations, those that are games, and those that are used as simulation games. A simulation game is not an actual environment, like

116

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

a game or simulation is. According to Thiagarajan [27], the only difference between a game and a simulation game is the mapping (or as he calls it correspondence) that takes place between elements of the game and elements of reality. These correspondences (and, frequently, the learning objectives) are known to the learners, therefore they are aware that they are doing—or have done—more than just playing a game. Most researchers investigating how to use a commercial computer game in a learning intervention have looked at learning plans where the learning objective can be achieved by utilizing the setting of the game, i.e. using a game that takes place in space to teach astronomy, or a game set within an historical period to teach history [13]. In contrast, this study pushes boundaries and investigates an area that has been largely ignored—using a commercial computer game as an instructional simulation which utilizes the environment but not the setting of the game, requiring both obvious and analogous mappings to reality. An analogous mapping is a correspondence from an element in the environment to a different element in reality (baseball team management mapping to business management, or resources mapping to a budget), whereas an obvious mapping refers to a correspondence from an element in the environment to a similar element in reality (flight simulator control panel mapping to a plane’s control panel, or managing armies mapping to managing armies). The more obvious the mappings are to reality (such as a flight simulator) the more authentic, more lifelike, and higher in fidelity, the simulation and/or game is [10]. It is a common understanding that the more fidelity one desires in a computer-mediated environment, the more it will cost. And, it is a common assumption that, in general, the more authentic a computer-mediated environment is, the better the learning experience will be. This study challenges this assumption. Instead of the conclusion that to better achieve a specific learning objective, one must increase the authenticity and thereby increase the cost of the activity this study attempts to show that it might be possible to have low fidelity (analogous mappings), low cost, and still successfully facilitate achieving a learning objective. 2.2.1. Advantages to Using Computer-Mediated Environments There are several possible advantages to using a computer-mediated simulation and/or game in an intervention designed to facilitate learning. Using this study as an example to illustrate these advantages, Multimedia Production Managers can go through the entire production process in a matter of hours, instead of months. They can do it repeatedly, either with the exact same scenario or a different scenario each time. Because they are doing the activity, they can get the “feel” for their task, recognizing and dealing with problems that arise. Time and money can be saved by not producing an environment specifically designed to train multimedia production management. Instead, a commercial computer game is simply used in a learning intervention that makes use of analogous mappings. Finally, motivation for learning can increase when using an environment recognized as a game.

3. Methodology This study is an exploratory case study, as opposed to a descriptive or explanatory case study [28,29]. It is a single case using a pre-existing Real-Time Strategy (RTS) commercial computer game (Age of Empires II) as the computer-mediated environment in

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

117

a training session designed to facilitate learning for a project manager. The specific type of project manager chosen for this study is a Multimedia Production Manager. This study is not a controlled experimental study, comparing how learning occurs using a RTS game with how learning occurs using other training techniques. Instead, the purpose of this study is to explore what happens when a multimedia production team goes through a training intervention using an RTS game as a simulation for learning. The researcher recognized similarities between an RTS game and multimedia production management, i.e. managing resources, managing personnel, and timing critical decisions. It is intended to investigate how a production team feels about using a commercial computer game to improve a Multimedia Production Manager’s skills by helping him to get the “feel” for his job [20]. This includes recognizing when alterations to the production plan need to be made and when to report these to the Producer. Additionally, a goal of this research is to add to the emerging body of theory on designing learning interventions that make use of computer-mediated learning environments. 3.1. Real-Time Strategy Games and Multimedia Production Management A real-time strategy game lends itself to training a Multimedia Production Manager. It has obvious mappings, such as working with a group of people to achieve a common goal, and it has analogous mappings, such as “creating a character” mapping to “hiring a person” and “gathering resources” mapping to “securing funds for a project”. Using this game as a simulation represented a real scenario of a production team working through the production process and the Multimedia Production Manager learning how to handle the team, the resources, working toward the Producer’s vision, and learning when and how to report alterations of the production path to the Producer. The researcher took a participatory role in the study. He served as the Producer to whom the Multimedia Production Manager reported. He shared his “vision” with the team (Multimedia Production Manager and subordinates) which, analogously, was the goal to accomplish within the RTS game. As the simulation progressed, the path to achieving the vision was altered when necessary depending on how the game panned out, and depending on the reaction of the Producer. This simulated real life when a Multimedia Production Manager approaches the Producer with updates, alterations, and problems. Aside from the participant researcher, there were four participants in this study (see Fig. 1): one who was the Multimedia Production Manager (at the strategic level of management) [30], and three who served under the Multimedia Production Manager as Supervisors (at the tactical level of supervision) who were over various aspects of production: visual production team, audio production team, and equipment team (at the operational level of production). The game served as the operational level personnel. The objects/roles/actions of each participant had corresponding objects/roles/ actions in the learning activity (see Table 1). For example, the Visual Content Supervisor oversaw the military buildings and people within the RTS game; the Audio Content Supervisor oversaw the miscellaneous buildings and people within the RTS game; and the Equipment Supervisor oversaw the technology improvements within the RTS game. Additional correspondences include: managing resources within the game mapping to managing cash flow, pausing the game mapping to having a meeting, and creating characters in the game mapping to hiring new employees.

118

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

Producer Strategic Level

Multimedia Production Manager

Visual Content Supervisor

Audio Content Supervisor

Equipment Supervisor

Visual Production Team

Audio Production Team

Equipment Team

Tactical Level

Operational Level

Figure 1. Levels of Organization.

Each Supervisor had their own computer monitor and was in a sound-proof room with a two-way radio so they could communicate with the Multimedia Production Manager (but, as the radios were tuned in to different frequencies, the Supervisors could not communicate with one another). There were also windows between the Supervisors and the Multimedia Production Manager. Hence, the Multimedia Production Manager could see the individual Supervisors, but they could not see each other (see Fig. 2). The Supervisors had the portion of their computer screen that displayed the game’s resources covered up. Only the Multimedia Production Manager had full view of the resources available. All of this was structured to simulate real life production management where one team can communicate with the Manager, yet they may not, as easily, be able to communicate with other teams. In the activity, just as in real life, it was the responsibility of the Multimedia Production Manager to gather all the necessary information from all of the team Supervisors and make decisions throughout the process. The Producer was in the same room as the Multimedia Production Manager. This allowed easy access for questions of clarification and defining of vision. The researcher provided a Gantt chart to each participant, which had the breakdown of activities and order that they were to occur within the game, Age of Empires II (see Fig. 3). A corresponding Gantt chart was also provided which represented the Producer’s vision and contained the real life milestones to which the game activities map (see Fig. 4).

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

Table 1. Correspondences (Mappings). Object/Role/Action Element in the Activity/RTS Game

Correspondence to Element in Real-Life

Participant Researcher

→

Producer

RTS Game Player

→

Multimedia Production Manager

Military Buildings and People Supervisor

→

Visual Supervisor

Misc. Buildings and People Supervisor

→

Audio Supervisor

Researching Technologies Supervisor

→

Equipment Supervisor

Military Buildings and People within RTS Game

→

Visual Production Team

Misc. Buildings and People within RTS Game

→

Audio Production Team

Technology Improvements within RTS Game

→

Equipment Team

Dark Age

→

Beginning the Project

Feudal Age

→

Rapid Paper Prototype

Castle Age

→

Beta Version

Imperial Age

→

Final Product

Villagers

→

Employees Capable of Multiple Tasks

Creating Characters

→

Hiring Employees

Characters Dying

→

Employees Leaving the Project

Building House

→

Adding Office Space

Building Buildings

→

Adding Equipment and Capabilities

Upgrading People

→

Personnel Training

Researching Buildings

→

Equipment Upgrades

Enemies Attacking

→

Unanticipated Problems Arising

Pausing Game

→

Group Planning Meetings

Managing Resources

→

Managing the Cash Flow

Visual Content Supervisor

Audio Content Supervisor

Monitor

Equipment Supervisor

Monitor

Monitor

Monitor Multimedia Production Manager Producer

Figure 2. Layout of the Learning Activity.

119

120

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

Dark Age

Feudal Age

Castle Age

Imperial Age

Begin in Dark Age with three villagers and a scout Build Barracks Build Mill Build Mining Camp Build Lumber Camp Upgrade to Feudal Age Build Blacksmith Build Archery Range Build Stable Upgrade to Castle Age Build Monastery Build University Upgrade to Imperial Age Build Castle End

Figure 3. Age of Empires II Gantt Chart.

Rapid Paper Prototype

1st Iteration

Beta Version

Final Version

Begin Multimedia Production with a Production Manager and Three Multimedia Content Leaders Write Script Do Storyboard Produce Paper Prototype Edit Paper Prototype Prepare for 1st Iteration Produce 1st Iteration Do Usability Testing Edit 1st Iteration Prepare for Beta Version Produce Beta Version Edit Beta Version Prepare for Final Version Produce Final Version End

Figure 4. Real Life Multimedia Gantt Chart.

Before the activity, there was a briefing period, where the researcher explained the study, paying special attention to clarify the analogous mappings and the goal within

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

121

the game. After the learning activity, there was a debriefing period where the activity was discussed. Concepts learned by the participants were a main focus. 3.2. Participant Selection All four participants had past multimedia production experience. In fact, that was the only requirement for the participants chosen. The researcher selected the team members based on preliminary interviews. Since this study was designed to gather data from a single case, there was no need for random selection. The participants needed only to have some multimedia production experience, so as to be able to understand the analogous mappings to reality. They were selected after having responded to a soliciting email that the researcher sent out and after a subsequent interview with the researcher. Coincidentally, the participants had interacted with each other somewhat in previous multimedia productions, but that played no role in their selection. Using pseudonyms, Kirk was selected to be the Multimedia Production Manager based on his past experience of being in production management positions. The researcher felt this experience would add to the realism of the study. Bob was chosen to serve as the audio content supervisor because of his extensive experience in this area. Troy was chosen as the visual content supervisor because of his extensive experience in this area. And, Nick was chosen as the equipment supervisor because he had general overall production experience. The researcher served a dual role of a Participant-Observer and a Multimedia Producer. It was more realistic for the researcher, who would know the most about the game and the learning activity, to serve as the overseer of the entire process (just as the Producer in real-life would do). 3.3. Data Collection In gathering data, the researcher employed a combination of qualitative methods. Stake’s case study methodology [29] is particularly useful when studying an exploring context, such as the one in this study. Entering an exploring context with a detailed and inflexible agenda, he argues, does not work because the researcher would tend to overlook emergent patterns, behaviors, activities, needs, and issues. Simply put, his methodology is responsive to emergent findings. He states, “The aim is to thoroughly understand [the case]. If early [research] questions are not working, if new issues become apparent, the design is changed [by the case researcher].” (p. 9). Such a methodology affords researchers the ability to catch the unintended consequences of an educational intervention. They are able to focus on “the real business of the case study [which] is particularization, not generalization.” (p. 8). Stake’s [29] case study methodology calls for a narrative description of the case in order to illustrate how events unfolded. In this study, this description was achieved with data gathered through observations, interviews, document analysis (the recorded computer game scenario), reactionnaires (see Appendix A), and survey questions (see Appendix B). Qualitative researchers, like Stake, argue that multiple data sources should be used to generate more credible and trustworthy data, which then lead to more valid assertions [31]. This is the process frequently referred to as triangulation [32]. Along with the group debriefing session, which doubly served as a group interview, the researcher conducted a post-interview with each of the participants individually. The interviews focused mainly on attitudes towards using an RTS game as a simulation tool for learning multimedia production management and why participants acted

122

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

in certain ways during the activity. Most questions were open-ended and the researcher refrained from leading the participants to certain answers. It was understood that the interviews were used as an opportunity for the researcher to gain understanding, not as a quasi-assessment of the participants’ actions. All interviews were videotaped. Observations were made during the briefing, activity, and the debriefing. During the briefing and debriefing session the researchers served as the activity facilitator and made field notes as the sessions unfolded. During the activity, the research, serving as a participant-observer, made field notes when he was not directly involved in the intervention. As a participant-observer, he took a risk of missing important interactions occurring away from him. To alleviate this risk, he video recorded each participant individually in their sound proof room. From the interviews, it became clear that the participants did not feel the video recordings were intrusive or altered their behavior during the activity. The researcher explicitly made two types of observations: one, of actions that involved interacting with the game, including when participants recommended certain courses of action; and, two, of actions that involved interacting with other participants, including how and when they chose to communicate with others. 3.4. Data Analysis The researcher synthesized data from observations, interviews, document analysis, reactionnaires, and survey questions into a case document. Data was compared from one source against that of another source, thus engaging in triangulation. First, the researcher reviewed his field notes of his initial thoughts and observations. Next, the videotapes were reviewed of the briefing, activity, and debriefing. The particularly interesting moments were flagged with time code identifiers. Then, the researcher reviewed the videotaped post-interviews. Emergent themes and patterns were identified and analyzed. From these themes and patterns, the researcher developed conclusions. As these conclusions developed, he iteratively reviewed the case document to identify strengths and possible weaknesses of his conclusions. Part of this study aimed to explore the feasibility of using a game that requires many analogous mappings to achieve a learning objective: in this case, of training competencies of a good production manager. It became necessary, then to identify what those competencies were. The researcher referred to the findings of a recent study regarding project manager competencies: An Investigation into the Competencies Required of an Effective Project Manager: A Web-Based Delphi Study [33]. In this study, the researchers surveyed 598 project managers about what the project managers felt were the main factors for success in their positions. The researchers identified the top 25 competencies of an effective and successful project manager (see Table 2). The researcher compared these competencies with the qualitative data from the multimedia project manager case to form additional conclusions regarding how feasible it is to use a game for training in this manner. All of these conclusions, as part of a qualitative case study, do not have inherent generalizability to other populations [34]. Their power, instead, lies in providing others with narrative descriptions which allow for making inferences about other possible situations. Since the uniqueness of the case always exists, case studies are not meant for producing generalizable results. As Merriam [35] (p. 19) puts it, “A case study design is employed to gain an in-depth understanding of the situation and meaning for those involved. The interest is in process rather than outcomes, in context rather than a specific variable, in discovery rather than confirmation. Insights gleaned from case studies

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

123

Table 2. Competencies Required of an Effective Project Manager. Order of Importance 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Competency know the goals, scope, and mission of the project conduct business ethically know how project success is measured listen effectively share credit for successes know available resources (funds, equipment, people, etc.) have strong verbal communication skills be able to recognize a problem/crisis prioritize manage crises make time-sensitive decisions effectively delegate and follow-up effectively execute project plan develop/create a project plan set milestones/deadlines know oneself take responsibility for failures know the team members manage a budget manage expectations understand the decision-making process in the organization resolve conflicts establish mutual trust align/focus team members know when to take control/back off

can directly influence policy, practice, and future research.” Additionally, as Stake [29] suggests, case studies can be used to suggest problems with, or modifications to, broad theories, perhaps identifying avenues to pursue and holes in theories that need exploring.

4. Case Description The researcher began thinking of the possibilities of using a commercial computer game in a learning intervention where it would be used as a simulation with many analogous mappings to reality. Then he worked on detailing the study, including: conducting a literature review, identifying existing holes in past studies, deciding on the particular knowledge and skills to be learned (those associated with multimedia production management), locating a commercial computer game with elements that could

124

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

be mapped to this knowledge and skill set, designing the learning activity, finding a suitable location to conduct the intervention, and selecting the participants. 4.1. The Case The learning intervention (briefing session, learning activity, and debriefing session), filling out the reactionaires, and conducting the post-interviews occurred over the space of about three hours. In describing the case, information from these events will be intertwined to form a clear narrative of what transpired. 4.1.1. Briefing All four participants arrived individually to the prearranged location. After initial introductions, the researcher began the briefing session. This took place in the Multimedia Production Manager’s location of the activity layout (see Fig. 2). From this location there was a clear view of the other participants’ locations. Therefore, it was unnecessary to use diagrams to explain the activity because the researcher could simply point to the locations. He did, however, use a white board when explaining the levels of organization (see Fig. 1). Once again, he explained the study to the participants, but this time in more detail (having already explained it in both the original soliciting email and the subsequent interview before participant selection occurred). Each participant received an explanation and a handout of his respective role and responsibilities. The researcher handed out the Gantt charts (see Figs 3 and 4) to all participants and explained how their individual roles and responsibilities fit into the big picture. He also made a specific and deliberate focus of the correspondences that their objects, roles, and actions had with real life multimedia production (see Table 1). Communication procedures were then explained (using the two-way radios). And finally, the researcher fielded any questions, which turned out to simply be clarifying questions. Once all questions were answered, each of the participants read through the mandatory informed consent form and signed it so that they were aware of the study in which they would be participating. They then took their places and the activity commenced. 4.1.2. Commencing Kirk reacted much the same way as a newly-assigned, real-life Multimedia Production Manager would. He had been given a new project, with new goals, with a new team of supervisors with whom to work. He was apprehensive and spent a little time getting to know his surrounding, tools, and team members. He probed each supervisor individually, saying, “Okay, Bob, what do you need to reach your goals?” or “Troy, is there anything you suggest?” There was then a short period of radio silence while Kirk tested out his familiarity with the game play. The actions were very structured with Kirk asking and receiving information, then acting upon that advice. Resources began plentiful, so Kirk was able to do nearly any request made by his supervisors. Rather uneventfully, the team progressed to the Feudal Age. When this happened, Kirk gave praise to each supervisor and attempted to make it individualized like, “Good job Bob, we’re now in the Feudal Age. Thanks for telling me what to build.” Once in the Feudal Age, team dynamics and individual personalities began to unfold. Resources became scarce and Kirk’s overall knowledge of the project was limited so he began to rely on his supervisors more. He asked Bob, “Hey, should we build a farm? What does it give us? More food? Is this our best source of food right now, be-

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

125

cause we are getting low on food?” Bob answered but his answer was a bunch of rambling and obviously not satisfactory to Kirk. In Bob’s post-interview, he revealed that he had no clue what a farm did, but he felt since he was supposed to be the expert in that area that he needed to at least act like he knew what was going on. He stated, “The little time [Kirk] did talk with me I thought it was imperative for me to give a response… I don’t think I really went out on a limb saying, ‘the farm produced food.’ … but for me to say, ‘I don’t know’ … would be like if I was in charge of running a thousand dollar mixing board and then I was calling him to say, ‘what does this knob do?’ [It] would be out of place.” So, Kirk did not respond (audibly, over the radio) to what he felt was Bob’s useless information and felt on his own to make a decision regarding the farm. He decided to build one anyway. At this time, Bob silenced some on his requests while Troy began requesting his requisite Spearman, Tower, Archery Range, and Skirmisher. Nick, though he should have began requesting many Researches, made a simple, weak attempt at requesting to research a Wheelbarrow. When he was told the resources were not available to do this, he did not offer up any more recommendations. Nick never asked about Researching a Loom, even after the resources were available. Kirk ended up Researching a Loom on his own accord, without the recommendation of Nick. In a way this resembled real-life, in that the Multimedia Production Manager would normally request an upgrade before one in actually offered. 4.1.3. Stabilizing During the Feudal Age, Kirk seemed to act upon whatever Troy was recommending. In his interview, Troy stated, “I suggested a lot of stuff and, I think, in general, [Kirk] trusts my opinion on things… I think that comes from our working relationship from outside the experiment… We’ve worked on projects together and we work well together… We are used to how one another works. One of the reasons I wasn’t questioning what he was doing was… I didn’t want to step on his toes. The one call I almost called him on the radio, but didn’t, was the placement of that second tower. It was a waste of resources and didn’t move our interests ahead… [Trust went both ways though] because there were some bad calls on my part that contributed to the situation at the end.” So, Kirk and Troy drew upon relationships from outside of the activity to determine how their relationship within the activity would be structured. In other words, they were already aware of working styles of one another and that came through in the training intervention. Kirk began to ignore requests of Nick and Bob, mostly because the two of them appeared unknowledgeable from the start and because Troy was more forthright in his requests. Kirk stated, “At the beginning of the [activity], I was really trying to accommodate everybody else’s goals. And then, about half way through, I decided that I was going to try and focus on what my task sheet said. And, I started ignoring more requests. So, I think that in the second half of the [activity], if they [the supervisors] weren’t vocal about what they wanted, they were definitely going to get ignored… So, if you were talking, you were going to be heard. But otherwise, if you weren’t going to answer a question unless I asked one, I wasn’t going to ask you one anymore… I think, like with Nick and Bob, who were more set back, that Troy, who was more vocal, definitely got more of his requests met than the other two simply because he was more demanding… During the [activity], I felt like Troy and I were having a conversation and for the other two, I really felt like it was more of a question and answer session… It

126

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

was pretty interesting to see how [communication] worked. Some people would just sit back and not say much and sort of only answer questions when asked them, where others would be more demanding of certain things. So, if this were the same individuals you were working with [in the real-world] … you’d be able to figure out individual working styles and address that directly to the real work place. It was interesting to sort of get the feel for that… You could definitely pick up on how willing people were to participate and get a feel of individuals’ communication styles.” To back this up, Kirk said in his survey that, “learning communication styles of the team was the most effective result of the simulation.” 4.1.4. Reacting and Managing Then the team progressed into the Castle Age. This is where Kirk had to react to many conflicting demands and his priorities had to be established and followed. A practical issue came up where Kirk realized that the radio frequency being used to communicate with Nick was no longer functioning and he had to use hand gestures to get Nick to change to a working frequency. Kirk then made an observation: it seemed as though many villagers were standing around doing nothing. “Bob, are those villagers doing anything? Can they all be harvesting from the same farm?” Again, Bob had no idea, but this time he did not respond over the radio, instead opting for a simple communicating facial gesture. Kirk, visibly bothered by Bob’s lack of input, decided to discover for himself that only one villager could harvest from a farm at a given time. He was visibly disturbed that these villagers had been sitting around doing nothing—much in the same way that a manager would view unproductive workers. Kirk then paused the game, calling for a meeting with the Producer, and asked for training. He said, “Is there a way I can find out if there are any idle villagers or idle military characters?” The Producer taught him how to accomplish this task and Kirk unpaused the game and continued. Fire then erupted on several building throughout the game. Troy asked about the fire and if they were getting attacked, but Kirk didn’t really respond because he was looking over some other responsibilities. This had the effect of appeasing Troy and he did not push the issue any further. Unfortunately, Troy was right. Their buildings were being attacked by the enemy and they were eventually destroyed. Bob then weighed in, saying, “Looks like we have a small fire in the southern quadrant. Looks like something might be on fire.” Kirk reluctantly responded, saying, “Yes, that’s okay, because unlike the previous fires, the fire you’re seeing now is the enemy building that we are attacking. It’s our land.” “Oh,” Bob responded, “Burn the f**ckers.” Kirk did not respond to this. All the while, in the Castle Age, resources were amassing. It seemed as if Kirk did not know what needed to be done with the resources (Nick had discontinued any communication, which might have prompted Kirk to invest resources in research or upgrades). Also, there was an obvious need to find more stone. In fact, Kirk was aware he could not reach his goal, the Producer’s vision of completing a Castle, unless more stone mines were discovered. He seemed focused on this one objective. So focused was Kirk that, “Oh, sh!#... Invading army on the South,” Troy’s voice echoed over the radio. Kirk replied, “We’ve got the reds, the greens, and the yellows all coming after us.” Even Bob and Nick weighed in, “Are we being attacked?—Is that the enemy?” Bob was a little reluctant to ask because of feeling a little dumb from asking the last time. “Yes,” replied Kirk, “A lot of trouble.” Then Kirk paused the game to

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

127

have a meeting with Troy…a long meeting. Everyone was surprised at how sudden the attack was. During this meeting, out of nowhere, unsolicited, and for no apparent reason, Bob yelled out, “Protect the stable, protect the stable!!” It was as if he wanted to appear productive in some way. There was no response from Kirk. Then Bob let out a long sigh and kicked back in his chair. Again, after several minutes of waiting, as if wanting to feel productive, Bob called out, “Looks like there’s a fire in the northern quadrant. Looks like we might have a small fire.” Then he began to laugh to himself. “Thanks, Bob,” came the sarcastic reply from Kirk. Bob then really began to laugh. But soon, Kirk’s meeting dragged on, and Bob got bored again. He started interjecting seemingly useless questions—”Anything I can help with?... How are we doing with that whole attack thing?”—to which Kirk never replied. Meanwhile during the meeting, Troy made a suggestion, “This is what I think you should do. Get everybody—I mean, all the villagers, all the military units and move them against the greens, because they’re trying to attack a tower on the South. So they’re going to take casualties attacking that tower. We’re going to lose villagers any way we cut it. But, the reds and the yellow army are probably going to waste their time on the building up there.” Kirk followed Troy’s advice. Then Troy responded, “Can you upgrade the towers?—Holy crap, they’re just…” And Kirk cut him off, “We have, like…all the military guys, I think, are dead. And they’re after our buildings right now.” Troy responded, “It looks like they’ve upgraded all of their archers too so that they can fire very rapidly. We aren’t going to be able to rebuild [the towers] fast enough. I don’t know if we can recover from this, to be perfectly honest.” Then, jokingly, but yet tying it back to the real-world, Troy said, “I think we just lost our grant funding.” Kirk then paused the game again to have a meeting with the Producer. “I need some clarification on your vision,” he said. “I’ve got good news and bad news—we have made it to the Castle Age and are doing well, but we can’t find any stone, so we can’t build a castle. So, we won’t be able to reach your goal,” Kirk continued, “Okay, that’s the good news. The bad news is our competition is closing in hard and we probably aren’t going to survive this shake-out. What do you want us to do? Should we aim for your goal, which is still possible, but very unlikely—or should we build some other buildings, like a University or a Monastery, which would increase our value to a potential buyer?” The Producer told Kirk to gather his information and present an alternative plan, if the main goal could not be reached. Kirk then turned to his visual content/military supervisor, asking, “Troy, what do you advise, militarily, to hold these guys off? I think these guys are pretty much going to wipe us out. So, in terms of corporate gains, we’re kind of looking at a sell-out to our competition—a buy-out.” Troy responded, “Well, this is more of a question for [the Producer]. If they destroy our Town Center, is the game over?” Kirk conferred with the Producer and replied, “Yes, if the Town Center is gone, you can’t—game’s over.” Troy then let out a long sigh and said, “I don’t know if there’s a defense, because we’ve got no military forces anywhere near there. Our best bet would be to try and rush their archers while they’re working on [destroying] the stable. But my bet is once they finish the stable, they’re going to start hacking on [our] Town Center. We might buy ourselves some time and they might try and take out [our] archery range, but I think we’re done for.” Kirk responded to Troy, saying, “Then, I think that’s it. We’re pulling out. This game’s—we’re selling out to them.” He then, on his own, chooses to build a University and a Monastery, while simultaneously defending the Town Center with every charac-

128

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

ter in the game. Later, in the interview, Troy admitted, “I had no clue until after the game, during the debriefing, why he went and built the University and a Monastery. It seemed to conflict with everything we were trying to do. It made sense afterward when I realized he was trying to add value to our company before the sell-out.” So, once the two final buildings were completed, Kirk paused the game one last time and turned to the Producer and said, “I recommend we cut our loses and sell-out to our competition.” The Producer agreed. Then Kirk announced, individually, to each team member, “We have liquidated our company to the red army corporation.” And, with that, the activity was completed. 4.1.5. Debriefing Once again, the five team members (including the Producer/participant-researcher) found themselves in the room where they began. The researcher expected the debriefing to be led by him, with key points of the activity being drawn out and the mappings explained. But, this is not what occurred. Instead, the team members led the discussion themselves, each having questions and need for clarifications from the other team members. “Oh, that’s why you did that,” and “That makes more sense now,” were commonly heard phrases. The researcher hardly had to emphasize the learning objectives and mappings, because the self-guided discussion went well beyond the learning objectives of the researcher and they had been referring to the mappings throughout the entire activity. This continued until one team member asked, “So, where was the stone mine anyway?” They decided to all investigate. When they unpaused the game and explored, they discovered that a stone mine was right near the Town Center and had been the whole time. It was just inconveniently placed so that it was not easily noticed. The participants all groaned in agony, “Oh, had we known it was right there, it would’ve changed everything.” And such is life.

5. Discussion It appears to be feasible to use a commercial computer game in a learning activity that requires many analogous mappings in order to achieve a learning objective. The Multimedia Production Manager certainly was able to gain a “feel” for the job. Even more impressive were the participants who, after just a short briefing session, were creating their own mappings to reality and developing their own meanings. 5.1. What Was Learned At the beginning of the activity, it appeared as though the increase in cognitive load was going to hinder the Multimedia Production Manager’s learning. However, this initial period of “accustomization to the people and tools” was soon superceded by a comfort-level with the game and participants, so that a meaningful learning session could continue. However, all four participants, as mentioned in their surveys, had at least some experience playing RTS games like Age of Empires II. Again, this was not intended, planned for, nor a criterion for their selection.

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

129

5.1.1. Twenty-Five Core Competencies It was interesting to note that every single one of the twenty-five core competencies for Project Managers, mentioned earlier [33], were afforded during the learning activity in this study. The Multimedia Production Manager was able to practice and improve upon each of these competencies, whether it was calling a meeting and talking through things with team members or simply attempting to achieve the goal and handling the process along the way. Kirk was able to know the goals of the project and work toward achieving them. He was able to assess small failures and problems along the way, take responsibility for these failures, and practice reacting to these failures to solve the problems. When there was success, Kirk shared credit for that success with others by verbally telling them so over the radios. He could also practice when to take control of a situation and when to back off and let a supervisor handle it. Verbal skills could be practiced and improved upon, both on the part of the supervisors and Kirk. Listening skills could likewise be practiced. This aided in helping the participants to learn about themselves and about team members. Once they all knew each other, and mutual trust was established, they could better work toward achieving the mission of the Producer. Kirk could practice aligning the team members so that everyone understood the major goal to achieve and how their part fit in it. He could gather information from all the participants and practice making decisions based upon the gathered information. A major affordance of this activity was the ability for Kirk to practice assessing the resources available to him, prioritizing the needs of his subordinate supervisors, and then allocating the resources to achieve the goals of the team. Kirk was able to practice the balancing act of resource management, gaining a “feel” for how to do this in reallife. 5.1.2. Working, Communication, and Personality Styles Certainly a recurrent theme in the post-interviews, surveys, and reactionnaires was the ability to discern the other participants’ styles of working, communicating, and personality throughout the intervention. They felt like there were opportunities to interact with the other team members in a way that increased the ability to understand personalities and work styles. Kirk learned that he had to handle his team members differently. With Bob, he needed to be specific in his questioning, or Bob’s response could seem irrelevant. Also with Bob, Kirk learned that he needed to take Bob’s comments not too seriously, as Bob might be making up a response just to appease his Manager. With Troy, Kirk learned to rely on someone’s input who appeared to have more experience. However, this appearance of experience could prove detrimental, when Troy really was not sure about what action to take and just guessed. With Nick, Kirk realized that he needed to return to him frequently, even if he had asked his questions several times before. Nick would not volunteer information over and over, especially if it was not solicited. And, Kirk learned about himself that he was more likely to pay attention to the most vocal members of his team, even if what they had to say was not that crucial. Just because a team member was talking more, they would get listened to more frequently than not. Something interesting that was not anticipated, but was revealed during Bob’s post interview was that he had been diagnosed with and takes medication for attentiondeficit/hyperactivity disorder (ADHD). Kirk was able to pick up on this during the

130

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

learning activity based on Bob’s actions. It explained why he was frequently bored, had a hard time paying attention, began the activity very excitedly and was involved, but then became distracted and uninterested. In fact, he said in his post-interview, had it not been for the cameras, he would have surfed the Internet during the activity. Bob admitted that in real-life he needs specific tasks to work on with deadlines and ability to be creative and original. He also needs to stay busy and might try to do several things at once or might impulsively choose to do things that have an immediate but small payoff rather than engage in activities that may take more time and effort yet provide greater rewards. Amazingly enough, all of these traits were exhibited during the activity, and the Multimedia Production Manager picked up on them. 5.1.3. Understanding Knowledge Management The learning activity in this study was deliberately set up to parallel a real-world situation with a multimedia production team. One of the choices that the researcher made was to have communication between the supervisors and the Multimedia Production Manager, but not between the supervisors themselves. This is a possible set-up for a real-world production team: vertical communication. It has its strengths and weaknesses, which became all too apparent during the intervention in this study. These were a main focus during the debriefing session. “If I would have known that…” was a common phrase in the debriefing. What this phrase represented was a recognition of a weakness that the vertical communication structure has: it does not provide for communication to travel horizontally, from supervisor to supervisor. Instead all information must flow up to the top manager and then be proliferated down again to whomever necessary to achieve a task. This make the top manager in charge of many decisions that the supervisors could handle. And, it does not allow the supervisors to share information with one another, information that might be very beneficial for another supervisor in handling an assignment. It became clear that the members of the production team in this study would have preferred a more horizontal flow of knowledge. 5.2. Implications A goal of this study was to examine some of the issues involved in using a commercial computer game as a simulation to facilitate learning. By employing the tactic of creating mappings around such a game that correspond to elements in the real-world, money, time, and resources could indeed be saved. A training institution, though averse to investing the resources into creating a specific, computer-mediated, instructional simulation to achieve particular learning objectives, might likely use a similar mapping strategy that was employed in this study. Additionally, using a commercial computer game appears to engage learners, facilitating complex negotiations between the game and between the learners. Learners can effectively gain the “feel” for their assigned task, even when that task is not explicitly laid out in the game, but instead mapped to through correspondences. Motivation for learning can increase, in part, because the learning is hidden within the confines of a game and does not appear as content to be learned.

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

131

5.3. Limitations A qualitative case study approach, such as in this study, comes with certain limitations. Indeed the features which provide for its rationale for selection also present limitations in its usage [35]. Though a thick description of the case may be desired, there is a balance that must be negotiated between the availability of resources (time, money, personnel) to the researcher gathering data and the amount of description and analysis the researcher feels is necessary. The researcher may oversimplify or exaggerate a situation [34], perhaps in hopes of offering practitioners a product that is not overly detailed, lengthy, or involved [35]. Since the researcher is the main instrument of data collection—through observations, interviews, and content analysis—the integrity and experience of the researcher can affect the qualitative case study. Oftentimes, the researcher is left to rely on instincts and abilities (and, as in this study, a researcher may be at it alone, with no assistants to offer checks and balances). Biases may creep into the qualitative study even more so than a quantitative study. This is because the researcher is making judgments and adjustments throughout the study. Unethical case study researchers have the opportunity to paint whatever picture they desire based on the data. Even further, they could present the data as accounts representative of the big picture, when, in fact, case studies are only a slice of the whole [34]. And, qualitative case studies, as a whole have been faulted for their lack of representiveness. Their generalizability is a limitation. It is not possible to make a claim that the findings of a particular case reflect a population at large. Though some researchers erroneously attempt to employ such tactics, it must be remembered that aiming for generalizability is not the impetus for this methodology. The main goal, instead, is to thoroughly understand the case [29]. This understanding, then, is meant to guide future investigators and practitioners in the actions that they take. This study, in particular, has specific limitations. For example, all four participants had RTS game experience. It can be inferred that the study would have been very different had each participant needed to become accustomed to the elements of the game environment. Had participants not been tech savvy and not had gaming experience, there might have been an increase in their cognitive load. This may have changed the learning experience for the multimedia production manager. Additionally, the debriefing that took place before the post-interviews could have affected individual interview responses and the answers to the reactionnaires and surveys. Group think may have caused participants to make statements that they might not have otherwise made had the interviews, filling out the reactionnaires, and answering the survey questions taken place before the debriefing session. 5.4. Future Research It is recommended that future studies investigate further the use of off-the-shelf computer-mediated games in achieving specific predetermined learning objectives that lie outside the obvious context of the game. These types of studies would focus on understanding the obvious and analogous mappings of elements in the game to the corresponding real world elements (the shell that is built around the game). Ultimately, experimental studies could be undertaken, comparing a group of trainees doing traditional training versus trainees using an off-the-shelf computer-mediated game as a simulation. Ideally, such studies would investigate the effectiveness of such

132

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

interventions at all four of Kirkpatrick’s [36] levels: reactions, learning, transfer, and results. In this study, learners seem to like how they are learning and, indeed, they seem to be learning; so further studies would be conducted to discover whether the new learning is being transferred to the real-world workplace and what the effects are on profits, or the return on investment. Perhaps, even, a cost-effectiveness analysis could compare which intervention actually achieves better outcome cost-wise. 5.5. Summary It is possible that people can learn by using off-the-shelf commercial computer games. Many studies look at designing simulations to achieve a certain learning objective. Some studies look at using computer-mediated games as simulations with mostly obvious mappings to reality where the learning plan utilizes the setting of the game. However, few studies investigate how feasible it is to use computer-mediated games as simulation games to achieve a learning objective that requires many analogous mappings. This study shows that it is possible that learning objectives can be reached when using a commercial computer game with some obvious mappings, but mostly analogous mappings, to reality. It is important to further pursue studies on how researchers can use off-the-shelf commercial computer games to achieve specific predetermined learning objectives. By studying how to build an intervention around the game that can facilitate achieving a specific learning objective, researchers will further develop theories about the best ways to use these games in learning interventions.

References [1] Hoffman, B., & Ritchie, D. (1997). Using multimedia to overcome the problems with problem based learning. Instructional Science, 25(2), 97–115. [2] Blanchard, R., & Christ, W., (1993). Media Education and the Liberal Arts. [3] Heinich, R., Molenda, M., & Russel, J.D. (1993). Instructional Media and the New Technologies of Instruction (4th ed.). New York: Macmillan Publishing Company. [4] Gee, J.P. (2003). What video games have to teach us about learning. New York: Palgrave. [5] King, & Borland, (2003). Dungeons and Dreamers: The Rise of Computer Game Culture from Geek to Chic. [6] Poole, S. (2001). Trigger Happy: Videogames and the entertainment revolution. London: 4th Estate. [7] Squire, K. (2005). Resuscitating Research in Educational Technology: Using Game-Based Learning Research as a Lens for Looking at Design-Based Research. Educational Technology, Jan-Feb. [8] Prensky, M. (2001). Digital Game-Based Learning. New York: McGraw Hill. [9] Macedonia, M. (2001). Games, Simulation, and the Military Education Dilemma. Paper presented at the Forum on the Internet and the University. [10] Rossett, A. (2004). Simulations and Games: Revisiting Their Strategic Value. The eLearning Developers’ Journal. October 4, 2004. Available: http://www.elearningguild.com. [11] Friedman, T. (1999). The Semiotics of SimCity. First Monday. Available: http://www.firstmonday.dk/ issues/issue4_4/friedman/. [12] Kolson, K. (1994). The Politics of City Planning Simulations. Paper presented at the Annual Meeting of the American Political Science Association, New York, NY. September 1–4, 1994. [13] Squire, K. (2004). Replaying history: learning world history through playing Civilization III. Dissertation, Indiana University, Education Library. [14] Walker, K. (1997). War-gaming Moves From War Room to Board Room. Available: http://www. purdue.edu/UNS/html4ever/9710.Chaturvedi.wargame.html Purdue News. [15] Winn, W.D. & Windschitl, M. (2001). Learning in Artificial Environments. Cybernetics and Human Knowing. 8(4), 5–23. [16] McLuhan, M. (1964). Understanding Media. (6th ed.). McGraw-Hill.

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

133

[17] Merrill, M.D. (2002). The First Principles of Instruction. Educational Technology Research & Development. 50(3), 43–59. [18] Merrill, M.D. (1992). Constructivism and Instructional Design. In Duffy, T. & Jonassen, D., (Eds.), Constructivism and the Technology of Instruction: A Conversation. Hillside, NJ: Earlbaum. [19] Duffy, T.M., & Cunningham, D.J. (1996). Constructivism: Implications for the design and delivery of instruction. In Jonassen, D.J. (Ed.), Handbook of research for Educational Communications and Technology. 170–198. New York: McMillan Library Reference USA. [20] Cook, D.N., & Brown, J.S. (1999). Bridging Epistemologies: The Generative Dance Between Organizational Knowledge and Organizational Knowing. Organization Science. 10(1). [21] Barab, S.A. & Duffy, T.M. (in press). From Practice Fields to Communities of Practice. In Jonassen, D.J. & Land, S. (Eds.), Theoretical Foundation of Learning Environments. Mahwah, NJ: Lawrence Erlbaum Associates. [22] Brown, J.S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of learning. Educational Researcher. 18, 32–42. [23] Jenkins, H. Squire, K. & Tan, P. (in press). You can’t bring that game to school! Design of Supecharged. To appear in Laurel, B. (Ed.) Design research. Cambridge, MA: MIT Press. [24] Laurel, B. (1991). Computers as theatre. Reading, MA: Addison-Wesley. [25] Rafaeli, S. (1988). Interactivity: From new media to communication. In R. Hawkins, J. Wiemann, & S. Pingree (Eds.), Advancing communication science: Merging mass and interpersonal processes, 110–134. Newbury Park, CA: Sage. [26] Squire, K. & Jenkins, H. (2002). The art of contested spaces. In Game on!, King, L. (Ed.) London, England: Barbican. [27] Thiagarajan, S. (1996). Instructional games, simulations, and role-plays. Chapter 25 in Robert L. Craig (Ed.). The ASTD Training and Development Handbook, 4th ed. [28] Yin, R.K. (1989). Case Study Research. Design and Methods. Revised Edition. Volume 5. Beverley Hills: Sage Publications. [29] Stake, R.E. (1995). The art of case study research. Thousand Oaks, CA: Sage Publications. [30] Solomon, M.R. & Stuart, E.W. (2000). Marketing, Real People, Real Choices. (2nd ed.). Prentice Hall. Upper Saddle River, NJ. [31] Fetterman, D.M. (1998). Ethnography. (2nd ed.). Sage Publications, Inc. [32] Denzin, N.K. (1989). The research act. (3rd ed.). Englewood Cliffs, NJ: Prentice Hall. [33] Brill, J.M., Bishop, M.J., & Walker, A. (in press). An Investigation into the Competencies Required of an Effective Project Manager: A Web-Based Delphi Study. Educational Technology Research & Development. Paper presented at the 2003 annual international conference of the Association for Educational Communications and Technology. [34] Lincoln, Y.S. & Guba, E.G. (1985). Naturalistic inquiry. Beverly Hills, CA: SAGE. [35] Merriam, S.B. (1998). Qualitative Research and Case Study Applications in Education. (2nd ed.). San Francisco, CA: Jossey-Bass Publishers. [36] Kirkpatrick, D. (1994). Evaluating Training Programs. San Francisco, CA: Berrett-Koehler Publishers, Inc.

134

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

Appendix A: Reactionnaire Simulated Multimedia Production Position: _________________ Are you interested in multimedia production management? If so, in what areas specifically; if not, why not? What was your favorite part of using Age of Empires II to simulate multimedia production management? What was your least favorite part of using Age of Empires II to simulate multimedia production management? Do you feel that using Age of Empires II can teach multimedia production management? What would you say you learned by using Age of Empires II to simulate multimedia production management? What aspects of multimedia production management do you feel cannot be learned when using Age of Empires II to simulate multimedia production management? If you could change anything about using Age of Empires II to simulate multimedia production management what would it be?

P.P. Parker / Learning when Using Commercial Computer Games as Simulations

135

Appendix B: Survey Questions Simulated Multimedia Production Position: _________________ Please describe what multimedia production management is to you. Please describe what a simulation is to you. What parts of using Age of Empires II to simulate multimedia production management do you feel most simulated reality? What parts of using Age of Empires II to simulate multimedia production management do you feel least simulated reality? In your opinion, what were some of the memorable points that stand out after having used Age of Empires II to simulate multimedia production management? What did you learn (about the game, about production management, about using simulations, etc.) after having used Age of Empires II to simulate multimedia production management? How often do you play video games? How much do you enjoy playing video games? What are your favorite video games? Had you ever played any real-time strategy game (like Age of Empires II) before today? If so, how much and which ones?

136

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

Serious Games and ‘Simulation Based E-Learning’ for Infrastructure Management Igor MAYER and Geertje BEKEBREDE Faculty of Technology, Policy and Management, Delft University of Technology, PO Box 5015, 2600 GA Delft, the Netherlands [email protected]

Abstract. The authors review the use and usefulness of digital games and simulations for (e-) learning, training and decision & policy support of technological infrastructures, such as ports, container terminals, off shore wind farms etc. The recently established ‘serious games initiative’ promotes and explores the use of the concepts and technologies of the entertainment, video-gaming and e-learning industries for serious purposes, i.e. higher education, professional and corporate training, policy support and management. Several examples of such applications are presented and three cases that bear relevance for infrastructures, are discussed in more detail: CONTAINERS ADRIFT is a computer-supported simulation-game revolving around the planning and design of an inland container terminal. VENTUM ON LINE is a multi-user on-line role playing game that revolves around the planning and design of an off shore wind farm. SIM MV2 is an animated and network based simulation game commissioned by the Port of Rotterdam to explore and support the planning and design of its second harbour area (2nd Maasvlakte). Keywords. Simulation- games, complex decision making, infrastructures

Educational simulations will be in widespread use by leading instructors within 5 years and will eventually change education as much as textbooks and motion pictures. Clark Aldrich, In: Simulation and the e-learning Revolution, Vision, Sept/Oct. 2003

1. Introduction The world of infrastructures includes some of the key determinants for the economic and social performance of industrialized societies. Yet this world is becoming more and more difficult to understand and manage. Business managers in infrastructure-based industries, such as utility companies, (air) ports and container terminal operators, – are facing enormously complex strategic and operational problems with a long-term decision horizon. Public policymakers and politicians who want to privatise, liberalise andor regulate infrastructure-based industries are confronted with many unexpected consequences of their policies and institutional designs. Students and young professionals who want to become either managers or regulators of such infrastructures certainly have a tough job getting the big picture and acquiring the necessary professional skills. How can we best prepare them? In this paper, we explore the possible contribution of recent innovations in the field of ‘games and simulations’ for such serious learning objectives. We focus on advanced

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

137

learning to support strategic decision-making about the design, planning and management of complex infrastructures. We will explain why and how games and simulations are powerful learning methods – and focus on significant and very recent innovations in the field. We will introduce some underlying concepts and give a few examples of serious ‘digital’ games and simulations that are currently being explored, developed and-or used by others. To substantiate and clarify our argument, we will discuss three cases where we have used simulation-games for infrastructure management: 1. A computer supported simulation game CONTAINERS ADRIFT that revolves around the planning of an inland container terminal and is part of our Master program ‘Systems Engineering, Policy Analysis and Management’. 2. A web-based simulation game, VENTUM ON LINE that we will soon be implementing. This game lets students and young professionals experience the pitfalls of project management when trying to realize an offshore wind farm and 3. A first prototype of the simulation-game SIM MV2, which will be used to support the infrastructure planning and land allocation in the second Harbour area in the port of Rotterdam. We round up the paper with some conclusions and observations about the future of games and simulations for infrastructure management. We will not attempt to be comprehensive, but will hopefully provide enough background and practical illustrations for educators and trainers to start exploring the possible use of games and simulations for their own purposes and fields of interest. We will mainly draw upon our own body of knowledge and line of work – decision-making about infrastructures in complex multi actor environments (see www.nginfra.nl). This does not imply that the message is not meaningful for other learning and training contexts for instance operations, finance or HRM. Irrespective of background or line of work, for most trainers and trainees there is a world of gaming to discover – our present and future students and trainees are already discovering that world everyday!

2. Serious Games and Learning I can’t wait to see what happens when online-learning collides with online games and simulation. I am looking for a big bang! Paul Stacey. In E-learning, Sept 5, 2003 [1].

2.1. How Serious Are Games? For many people, the word ‘game’ will refer to an entertainment artefact or event. Far from that narrow perspective, games are not limited to entertainment purposes but can be serious business – in more than one respect. As early as the 18th century, (board) games were used for serious, often political or pedagogical purposes. Since then, this practice has reached significant higher levels. Some interesting 18 and 19 th century examples have recently been presented at the ‘Past times and Paradigms’ exhibition at Cornell University (from February 2004, http://rmc.library.cornell.edu/games/intro. html). The exhibition shows how games were thought to improve work ethics, provide moral instruction, and instil social and cultural values and to develop particular skills such as manual dexterity, deductive reasoning and memory. Among the examples is a board game titled PANK-a-SQUITH (around 1910) – a suffrage game that garnered support in the battle for women’s votes.

138

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

Text 1. The intricacies of a war-game.

In 2002, when the United States were preparing for a possible war with Iraq, the Pentagon arranged a major war-game. A retired American general was asked to play the role of the Iraqi military commander. Much to the dismay of the US army command, their opponent decided not to play the war-game according to the Pentagon set of rules. By using any unconventional, untimely and unexpected means of war, such as ‘pre-emptively’ sending kamikaze pilots to attack American aircraft carriers and using minarets for military communications, this American ‘Saddam’ forced the war-game operators to turn back the clock several times in order to revive their military losses. Finally, realizing that he would not be allowed to win, the American general decided to step out of the game. What was the point, testing the robustness of military strategies if the Commanders in Chief were not prepared to play the game to the full? Source: ‘And then Saddam Hussein won’, Volkskrant, 21-09-2002.

2.2. Simulation or Game? A pure sim focuses on the thing or process being simulated – A pure game focuses on the user experience. A pure sim doesn’t reward or promote, have levels of difficulty – A pure game doesn’t include any ‘boring bits’ or bother being real. Marc Prensky, Source: www.marcprensky.com/writing/nasaga/MPrensky-nasaga.html

As yet, corporate and public managers and even most of trainees are no Digi-kids or Digital Natives – at least not in their professional time! That is why we believe that for advanced learning and policy support, we need the best of gaming and simulation. For advanced learning we need good games that also convey a sense of reality and urgency to professionals. Just as we need good simulations that are interactive, entertaining and fun! In order to keep away from a semantic discussion about what games and simulations really are – or what the difference between them is – we simply posit simulation-games as a safe (physical or virtual) environment, based on reality, in which participants can experiment with decisions and negotiations. This environment allows them to both experience and analyse the consequences of those decisions over time. Such experiences are relevant for a better understanding of how complex socialtechnological systems work and how decisions can be made about them [2: 27–35]. The use of gaming-simulation for learning, decision-making and management of complex infrastructures is relatively new, given that gaming-simulation has a long and established tradition that began with the use of these methods for warfare [see 3]. The brief example of a war-game below, indicates that simulation-games are more than just entertaining; they are relevant and are taken seriously enough to involve top-level decision-makers such as military commanders. In many ways, the war-game is a powerful metaphor for the non-military use of simulation-games – in our case with the objective to learn master students, young professionals and top decision makers about infrastructure design and management. Infrastructures are highly complex, multi-actor systems, involving strategic decision-making, competition and conflict, negotiation and diplomacy, tactics, logistics, operational planning and much more. Like military analysts, business consultants and policy analysts can use simulation-games to foresee and assess future scenarios, draw lessons from them and communicate strategic recommendations. These lessons can and should be translated into tactics and decisions, skills and personnel training and operational exercises. In all of these dimensions, simulations and games can be effective-learning methods.

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

139

For business and public managers, the rapidly changing world of infrastructures resembles a war in many regards. While industries are being liberalized and privatised, strategic alliances between companies are formed and broken up. The relevant parties make strategic and tactical manoeuvres to advance their positions. Opponents may lie and cheat. We would therefore do better to teach our students and train our professionals in a realistic, but safe and experiential environment before we send them ‘into the field’. By just once in a while playing out different strategic or operational scenarios in a war room-like setting, a management team may be better prepared for the unlikely or even the unthinkable. Due to the enormous advances in computer and intelligent communication systems, the military as well as the non-military ‘war-games’ of our time can be played in highly realistic and sophisticated environments and virtual worlds. Modern information, communication and simulation technology allows us to use and manipulate the real and actual data, to built simulation models and computer animations with which players can interact and can give impressions of a thinkable or future state of affairs. If not a discipline, S&G at least can be considered an established field of practice with proven value in most areas of public policy analysis, business consultancy, learning and research [4]. The practice and underlying theories, concepts and tools of simulation and gaming have been institutionalised in an International Gaming and Simulation Association – ISAGA (www.isaga.net). For 35 years, ISAGA organizes an annual international conference, attended by hundreds of game and simulation designers, consultants and policy analysts, teachers, computer scientists etc. from all over the world. In addition, several national or regional gaming and simulation associations have been founded – such as NASAGA (North America), ABSEL, JASAG (Japan) and most recently established in May 2004, the Netherlands Simulation and Gaming Network (SAGANET) (for a complete list of all affiliates see: http://www.isaga.net/affiliates. htm). Since a couple of years however, the interest in using games and simulations for learning and policy support is coming from a somewhat unexpected side: the entertainment and video gaming industry. 2.3. Digital Games and Simulation Based E-Learning According to an increasing number of authors who find their roots in computer-mediasciences and the gaming-entertainment industry, the use of simulation and gaming for learning will soon be taken to a next level [5]. For many years, board games, roleplaying games and computer-supported simulations have been used for K12 and vocational education and for higher education (graduate and post graduate) and professional learning. But there is now growing support for the idea that the technology and concepts used by the entertainment and video-gaming industry can be used to revolutionize-learning [6: see also www.seriousgames.org, 7, 8]. Before we can turn to the questions if, why and how such visions and technologies will impact infrastructure management, let us first consider some of the underlying concepts, implications and best prototypes of what is presently known as the serious games initiative.

140

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

2.4. Who Are Behind It? The opportunities and benefits of using digital games and simulations for ‘serious purposes’ are very recently being explored in partnerships between industry, universities and public organizations. ‘Serious Games’ is a relatively new, but sensitising concept which we will use here simply to denote a variety of loosely coupled initiatives that are presented under the same or other headings: social impact games, digital game based learning or simulation based e-learning. In March 2004, a 2-day Serious Games summit attracted over 300 participants mostly game developers but also potential users, scientists, business consultants, software companies etc. Serious games and simulation based e-learning initiatives try to use the technology and concepts of the game (and e-learning) industry for nonentertainment purposes in particular (higher and life-long) education and public policy (support) in particular for Healthcare/Hospital Management, Education/High-School Leadership; Public Lands Management and the military/Navy. A website (www. seriousgames.org) is dedicated as a portal for initiatives on serious games in particular to enhance productive links between the electronic game industry and projects. Like the entertainment industry, the United States seems to be leading in the field of ‘serious games’, Various programs and networks have been initiated by leading game-designers such as Ben Sawyer (from Digital Mill, www.dmill.com) [9], the Woodrow Wilson International Center (www.wilsoncenter.org), MIT (‘Games to Teach project’) and are supported by institutions such as Alfred P. Sloan foundation (www.sloan.org). But there are also a couple of interesting examples and initiatives of serious games and simulation based e-learning for higher education and policy support in Europe (Maharg, 2001 & 2004; Dziabenko et al., 2003; www.unigame.net; www. sig-glue.net). Marc Prensky (Prensky, 2004) and others like the Education Arcade consortium (i.e. MIT, and several game designers funded by Microsoft) (www.educationarcade. org) mainly focus on serious games as an education tool – elementary school to university. Other institutions such as www.watercoolergames.com focus on games that go beyond entertainment and education purposes and have introduced new gaming concepts such political games (e.g. terrorism) and ‘advergaming’ (for advertisement, recruitment etc.). The serious game initiative has now been acknowledged by North American Simulation and Gaming Association (NASAGA, www.nasaga.org) and the Digital Games Research Association (DiGRA, www.digra.org). 2.5. Why Serious Games? Kids want games not because they are games but because they are the most engaging intellectual things they have. Marc Prensky, Source: www.marcprensky.com/writing/nasaga/MPrensky-nasaga.html

According to proponents of the idea there are several reasons why serious games will (and should) change, even revolutionize, the practice of higher and professional learning (and training). First and on the demand side, game developers and users such as teachers and corporate trainers are now starting to serve a generation of students and young professionals that have grown up with advanced digital games and simulations. Our students and trainees possess the practical knowledge and gaming skills that far exceed their teach-

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

141

ers’. To quote Marc Prensky ‘technology is everything that did not exist when you were born’ [8]. Digital games are ‘technology’ for most teachers and trainers, but not for their students and/or trainees. Although for many purposes, board games and social simulations will continue to have didactical power, it becomes increasingly difficult to persuade and motivate students and young professionals to play and learn from them. Their expectations regarding a game – in terms of speed, fun, gains, looks and forms of discovery – let alone how and what they learn from them! – are to a large extent mediated by their experiences with video games or massively multiplayer on line role playing games (MMORPG). Second, during the previous decades our perspective about learning and education has undergone radical changes. The dominant view on teaching and learning by class room lectures and literature study, has gradually been replaced, or supplemented by ‘constructivist’ learning concepts – ‘authentic learning’, ‘active-learning’, ‘self directive-learning’, ‘open learning communities’ and ‘life long learning’ have become prevalent notions in (post) graduate programs. By and large, these concepts imply that students take up an active role in, and are responsible for, their own learning process – inside and outside the classroom, during their formal education and after (life long). Students are increasingly provided with a learning context and subsequently guided and facilitated by their teachers and instructors in their learning process. As many will know from experience, this often implies that traditional ways of teaching are supplemented with case based project work by student groups. Games and simulations have a definite advantage over traditional project work – business cases – because only the first provide dynamic and experiential feedback from a simulated world (i.e. by other students, AI or a simulation model). In sharp contrast to business cases and even most e-learning modules, simulation-games do bounce back. They provide us with the most authentic learning experiences – next to the real world of course. Third and much related to the above, the constructivist-learning paradigm has been accompanied by the implementation of e-learning tools and technologies – e.g. learning content management systems (LCMS) of which Blackboard and WB-CT are commonly known. Experiences however with many linear e-learning courses and applications, in particular those for corporate training, have led to some discontentment. Some disregard such e-learning systems as ‘click and fall asleep’ [8]. The response may be due to the poor quality of interactivity – among students, between students and teachers but most importantly between the students’ products and decisions and the world ‘out there’. This is one of the underlying reasons why there is now a growing interest in the combination of e-learning systems with simulation and gaming – both for corporate training, business consultancy and higher education [see for example 10: www.sigglue.net]. Fourth and on the supply side, the entertainment and video game industry has induced rapid and major innovations in gaming and simulation technology, concepts and applications. The power of the ‘push and pull factors’ that trigger innovation in this industry can probably only be matched by the military industry – and it is therefore no surprise that both industries increasingly seem to join forces. The game American Army, which is used to recruit army personnel, is a huge success in terms of downloads, number of players and new recruits (www.americasarmy.com). According to Learning Circuits, a journal on e-learning: ‘Modelling and simulation are leading the assault of new learning technologies that are winning favour with the US military.’ ‘… We stand on the verge of potential

142

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

Text 2. Marc Prensky on games.

According to Mark Prensky [8], one of the leading proponents of digital game based learning, innovations in gaming technology will have the following consequences: – Games will be much more realistic, experimental and immerse – Games will be fully online, wireless, and massive multiplayer – Games will be including more and better storytelling and characters – Games will be more about people and human interaction – Communication and cooperation will become more important elements – We will create the games we want – We will have new games forms and subject matters – Mass entertainment types of games will become common – Games will be higher quality – Games will become even more engaging

training revolution’. ‘US Defense is tapping the expertise of the entertainment and gaming industries to improve simulator based instructions’ and ‘the latest in video games is being used to entice and teach computer savvy soldiers’.

Paul Harris in Learning Circuits, October 6, 2003, Source: www.learningcircuits.org/ 2003. [11]

It is no wonder therefore, that ‘the corporate sector is keeping an eye out for new techniques suitable for corporate training’. However, the budgets available for higher education or even corporate training/policy analysis, are nowhere in the range of the budgets involved in the video gaming or the military industry. Yet, if we are smart we might be able to use elements and components from the entertainment and digital gaming industry efficiently for non-entertainment purposes such as learning, corporate training and policy support – even if the simulation-games need to be tailor-made, thelearning that needs to take place is advanced and the budget does not involve a million or more. Of course, the big question is whether these visions and expectations will materialize. And whether games and simulations will truly be able to enhance-learning – not only for recruitment, edutainment, K12 education or relatively simple instructional objectives, but also for higher forms of open-ended learning and support of complex decision-making. In other words, can we learn about the management of ports and other complex infrastructures through digital simulation-games and simulation-based elearning? As stated before, our CEO’s and top decision-makers and even most of our trainees are no Digikids or Digital Natives (yet!). Let us take a brief look at some examples and see just how far we are with serious games and simulations (for an extensive overview see www.socialimpactgames.com/). A few years ago and with sponsorship by the Sloan foundation, Ben Sawyer and some distinguished game designers, developed ‘VIRTUAL UNIVERSITY’ – a game about university management. The game became a hit with 1000 copies sold and more than 15,000 downloads (www.virtual-u.org). Another game, VIRTUAL LEADER www.simulearn.net/SimuLearn/simulearn_home_page.htm) designed by Aldrich and others is a sophisticated and widely distributed game that teaches leadership skills. The player is in a meeting room discussing issues with simulated characters whose words

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

143

and behaviour are controlled through Artificial Intelligence (AI) [7, see also an interview by Morrison with Aldrich, 2003 in Vision, 12]. VIRTUAL-U and VIRTUAL LEADER are interesting examples of single user digital games with a serious subject and learning purposes. They have serious learning objectives but are still entertaining. To some extent the games are quite complex, use state of the art gaming technology and have characteristics of an open ended game; On the other hand, in order to be useful for professional learning or real policy support they may not adequately reflect the social and strategic complexity of real world decision making. Another major disadvantage is that they are quite costly to develop – a million US dollars or more – while the content is not very case specific or adaptable to other contexts. Some great and ambitious serious game concepts such as DAEDALUS END therefore never materialized. FIX YOUR COMMUTE (FYC) (fyc.heraldnet.com) and US OIL POLICY SIMULATION (US OIL) (broadcast.forio.com/pro/oil/index.htm?FD_rand=1659) are interesting examples of web based single user games with serious content and objectives. The advantage of both games is that anyone with Internet connection can play them. The current state of the web-based game only requires about 10 or 20 minutes to play one cycle of the game. In other words, web technology reduces the resources needed to develop and play the game and can therefore diffuse serious games to massive number of players. The downside is that as long as they do not include elements of multi user role play, the social and strategic complexity of these games is even lower than the above examples and the opportunity for debriefing and evaluation are even less. There are many examples of business games, based upon running mock companies, that for instance teach management or business skills (supply chain management, HRM, finance etc.). More and more of these games are now Internet mediated – but experience has shown that designing and running a web based and multi player business simulation can be extremely hard and requires a variety of expertise. Most business games are entertaining mainly because they are competitive but many of them are also rather inflexible, with straightforward messages build in from the onset. A web-based simulation E-GLOBE (www.simenco.nl) on business and technology management, and the web-based GLOBAL SUPPLY CHAIN GAME designed at TU Delft (www.gscg.org) are good examples of what can be done with current technology in simulation-based e-learning. Both games are flexible, adaptable and open and suitable for more advanced learning objectives. A completely different and even more flexible approach to gaming is taken up in role playing environments such as UNIGAME (www.unigame.net), ARDCALLOCH (www.ardcalloch.ggsl.strath.ac.uk) and FABLUSI (http://www.fablusi.com/). These are nice examples of open simulation based e-learning systems that provide a re-usable simulation and gaming architecture that is relatively independent from the game content. They are used for e-learning in higher education. UNIGAME is an Internet based learning content management system that allows university teachers to develop and manage their own simulation, i.e. develop their own game scenario [13]. Such flexibility of e-learning systems of course would also be very helpful for corporate training. It would allow trainers to develop their own simulation based e-learning modules for specific clients, topics and objectives. The underlying architecture provides chat, e-mail and video conferencing services as well as features to make the game motivational. ARDCALLOCH has a similar structure but also includes a map of a virtual town and virtual companies [14,15]. ARDCALLOCH is used for law education at the university of Edinburgh. A Dutch version SIEBERDAM to be used also for other areas in particu-

144

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

lar public administration is currently implemented and tested in several Dutch universities (www.kodos.nl). The advantage of the aforementioned systems is that the technical architecture and the content are separated so that new games can be developed and implemented relatively easily. Teachers or trainers can be involved or even develop and manage the games themselves. Furthermore, these systems do reflect the strategic and social complexity of the real world system. The downside of these systems however is that they do not include simulators, so there is a risk that students will reach ‘negotiated nonsense’ because there is (no long) term and dynamic feedback of (quantitative) consequences. How and in what sense are the aforementioned illustrations relevant for trainers and trainees of infrastructure management? For infrastructure management the subjects and objectives may not seem that serious yet – but the underlying visions, technology, concepts and ideas can be (re-) used! In the next section we will give three examples of ‘simulation-games’ for higher and professional learning and policy support that revolve around the management of infrastructures. 1. 2. 3.

CONTAINERS ADRIFT is a computer-supported simulation-game revolving around the planning and design of an inland container terminal. VENTUM ON LINE is a multi-user on-line role playing game (MORPG) that revolves around the planning and design of an off shore wind farm. SIM MV2 is an animated and network based simulation game commissioned by the Port of Rotterdam to explore and support the planning and design of its second harbour area (2nd Maasvlakte).

3. Serious Games and Simulation Based E-Learning for Ports and Infrastructures 3.1. What Is Behind the Games? From a decision-making perspective, the planning and design of infrastructures can be characterized on two dimensions: 1. the degree of consensus on norms and values; and 2. the degree of consensus on facts and causal relations [16]. Most commonly, the planning and design of complex infrastructures will score low on both dimensions, i.e., a strong disagreement between stakeholders on values and norms in combination with a great many technological, economic and logistical uncertainties. These situations are usually characterized as ill-structured, wicked or messy problems. The decision-making process under such conditions is known to have a number of characteristics among others: 1. 2.

3.

The planning and design process will be pluricentric in the sense that no single stakeholder can dominate or monopolize the decision-making. It will be dynamic in the sense that the perceptions of problems and solutions will change over time. Stakeholders will enter and exit the decision-making arena and the process will progress by fits and starts. Stakeholders will behave strategically in order to optimise their own interests and values.

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

145

Table 1. Characteristics of the three simulation-games. CONTAINERS ADRIFT

VENTUM ON LINE

SIM MV 2

Realism of subject

Planning and design an inland container terminal

Building an Off shore wind farm

New Harbor area Port of Rotterdam

Actuality in real world

Realistic, but no real decision-making process or problem

Political decision making – consortium forming

Master plan infrastructure and land designation

Long term horizon

Months

2005–2020

2005–2030

Advanced learning

Individual – post graduate – professional – life long learning

Individual – post graduate – professional – life long learning

Individual and organizational learning

Decision support

Low

Indirect influence – Project managers

Direct influence – Port of Rotterdam

Professional users

Limited

(Future) project managers

Real decision-makers port of Rotterdam

Openness in the game

High

Ambiguity of information; interdependency

Uncertainty and path dependency

Degree of realism

High

Fictitious data and context

Real and fictitious data and context

Interaction

Personal

Various forms of social & computer interaction

Various forms of social & computer interaction

Number of users

10–40 in one session

Multi user (8–100 simultaneous.)

Multi user (4–16 simultaneous)

Debriefing

Personal

Evaluation, assignments, on line, personal

Personal

Flexibility

Face to face;

(a-)synchronous, co-located and distributed formats

Co-located and synchronous

Platform

PC

Internet

Internet – LAN – PC

Visual aspects

Low quality graphics (simple drawing tool in visio, GIS map)

High quality user interfaces (flash); animations;

User interfaces (flash); 2,5 D animations; GIS;

Communication

Personal, not through computer

Chat, message exchange (on line)

Personal

Seriousness

Game design

Technical design

Architecture

Simulated players (AI)

Simulated players (AI)

Flexible and re-usable

Flexible and re-usable

Java, flash, D-sol

Java, flash, D-sol

Organization issue Partnerships

Faculty departments (systems eng., public admin etc.)

Commercial game designers; interaction designers;

Client; Commercial game designers; interaction designers;

Development time

1 year

1 year

1 year

Design team

Systems engineers, public administration and organization management, students

Information sc., game and interaction designers; public admin. sc.

Harbour and port experts; Information sc., game and interaction designers; public admin. Scientist.

146

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

Whereas stringent project management may become crucial in later (operational) stages of infrastructure planning and design, the management of relations between stakeholders and the stakeholder negotiation process itself are essential during the initial (strategic) stages of the project [17]. The underlying objective of the three simulation-games discussed further below, is to let participants experience and understand how uncertainty, ambiguity and strategic behaviour interfere in such projects and how to manage such projects in interaction with other stakeholders. How do the simulation-games achieve this? 3.2. CONTAINERS ADRIFT CONTAINERS ADRIFT is a computer-supported simulation-game that revolves around the planning and design of an inland container terminal to be located near the fictitious provincial town of Maaswijk. The simulation-game is played with 15–35 participants divided over 10 roles such as the terminal operator Mega Container Terminal (MCT), the municipality of Maaswijk, a commercial bank, potential customers of the terminal such as Holland Food, an association of local residents, a local environmental group and various companies involved in shipping, logistics and transportation. Two to four participants play each role as a team. In general terms, the plot is as follows. In its national freight policy, the Ministry of Transport enhances the development of inter-modal means of transportation. In the Netherlands this implies making optimal use of innovative combinations of waterway, railway and road transportation. Inland container terminals are an inevitable link in these inter-modal transportation chains. A policy scan has shown that it is worth exploring the feasibility of an inland container terminal situated in a neglected harbour area near a major waterway and industrial zone in the town of Maaswijk. MCT, a large German container terminal operator, has shown interest in developing the project commercially. However, many local and regional stakeholders are needed to realize the terminal. The terminal will have to be financed, it will have to meet customers’ demands, the municipality and province will have to approve and co-operate, and local residents and environmentalists may obstruct and delay the project by starting legal procedures. The Ministry of Transport and MCT have set up a joint project group. The process managers of this project group are made responsible for the initial negotiation and information process with all important stakeholders in the region. The Ministry of Transport will only subsidize the container terminal if a number of social, financial, economic and environmental criteria are met. At the end of the game therefore, the participants are required to present the Director General of the Ministry of Transport with a (few) design(s) of a container terminal, indicating its financial, logistic and environmental feasibility. The designs of the container terminal are the result of an interactive or participatory design process, which takes place during the game using an interactive simulation tool. The level of support for the container terminal and the continuation of the process have to be expressed by the stakeholders in a covenant. This covenant specifies the agreements on procedure and project and sets the agenda for further negotiation. Over the past four years we have held about 20 sessions of CONTAINERS ADRIFT. Sessions ranged from two-hour demo versions to a one and a half month version. Besides for experimentation and testing the participatory decision-making process and the interactive simulation tool, we used the game as a teaching environment, integrated into the Master’s curriculum of the faculty of Technology, Policy and Management of the Delft University of Technology and professional trainees [for a

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

147

Figure 1. Master students at work with the participatory modelling tool during the game.

detailed description and evaluation of the game, see 18]. At present, we are working on a new and improved version – with high quality animations and underlying technology that allows wide distribution and use. 3.3. VENTUM ON LINE VENTUM ON LINE (www.ventum.nl) is an adapted on line version of an original role-playing simulation-game, VENTUM, that has no computer support. VENTUM was used for many years within our master program and for external in-company training sessions with young professionals. It was therefore taken as the starting point to develop an on line version in a partnership between two universities and IJsfontein interaction design – one of the leading Dutch companies in development of educational games. The original game as well as the on line version revolves around the projectmanagement for building an off shore wind farm. Both versions are modeled after actual developments in the Netherlands, i.e. the decision to experiment with offshore wind energy winning. In the game, groups of students form a consortium by setting up initial agreements among the various companies involved. Next they negotiate and agree upon a tender proposal to submit to the national government. After the tender has been granted, the consortium members have to actually design and build the off shore wind farm according to their own specifications. Each company can perform R&D, make decisions for design, can buy information about aspects such as wind speed or wave heights from consultants. Overall, the game is characterized by a high level of uncertainty, distributed information and interconnected decisions. The main challenge for the consortium members is to co-ordinate their activities and to manage the tensions and conflicts that will certainly arise among them. When all partners have agreed upon their final design they can agree to build. A scenario then shows a video of the actual building of the wind farm and the teams’ final performance indicators. During the game, social interaction among the players can be arranged through the Internet (chat,

148

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

Figure 2. User interface of VENTUM ON LINE.

Figure 3. Building the harbor in the game SIM MV 2.

e-mail) or by arranging face-to-face meeting (in settings where this is allowed and possible). Important parts of the game are modules for registering, information, communication and decision-making. At the time of writing, the on-line version is in a test phase. Information about its actual state of development can best be found at the website (www.ventum.nl).

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

149

3.4. SIM MV 2 SIM MV2 is a multi player computer based simulation-game that revolves around the infrastructure planning and land designation in the second harbour area (2 nd Maasvlakte) of the Port of Rotterdam. After a lengthy and highly controversial public decision making process, the Dutch national government has recently decided to reclaim from the sea some 1000 ha of new land in the Port of Rotterdam area. During the coming decades, this new land has to be suppleted in several phases. The infrastructure (energy, roads, but also docks, jetty etc.) has to be built and future clients have to be found. The planning and decision making process is therefore characterized by a high level of uncertainty, path dependency and strategic stakeholder behaviour. Technical and political aspects of the decision-making are highly interwoven and a number of pitfalls are foreseen. Commercial and infrastructure decisions for a period of several decades need to be coordinated between different departments of the Port of Rotterdam. Moreover, exogenous uncertainties such as the development of the global and national economy, the relative economic development of the various industrial sectors, future innovations in containerships (their depth) and logistics need to be taken into account. The simulation-game was commissioned by the Port of Rotterdam to support its actual decision-making process. SIM MV 2 uses advanced simulation and gaming techniques to: 1. Reach better short and long-term commercial results for the 2 nd Maasvlakte; 2. Increase the insights and knowledge about exogenous and endogenous uncertainties related to infrastructure planning and land designation strategies 3. Improve the communication and co-ordination of different departments of the Port of Rotterdam. The simulation-game, which is currently underdevelopment, will be played with 8–16 players. Players will be the real decision-makers of the Port of Rotterdam. Social interaction can be arranged among the players – while various decision makers insert decisions in a distributed fashion (LAN). The underlying simulation model contains a Geographical Information System, a design tool and a 2,5 D visualization of the results. This game furthers the underlying ideas and technology of CONTAINERS ADRIFT but in this game the situation is real, the data are realistic, the players are the real decision-makers and the simulation and design tool is advanced.

4. Conclusion Digital games and on line simulations will increasingly be designed and used for higher learning and professional training. The world of infrastructures – (air)ports, utilities etc. – may become an early adopter of such applications. One of the reasons is that due to the characteristics of these industries – e.g. the operational and logistic aspects and the spatial and physical dimensions – modelling, simulation and virtual worlds are already an integrated part of that world. Gaming can and will be used to add user experience, motivation and creativity in learning or training contexts. There is however one point we would like to stress. Simulation-games will never be able to meet the aforementioned challenge if they are used separately and/or in isolation from other approaches and methods. They work best when they are embedded into a broader research, learning or intervention process in which a number of complementary methods and activities are used. In a learning environment, this implies that simulation-games should be integrated into a well-designed educational or training program, where lectures on theory, project work and internships are (at least) equally important. In an em-

150

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

bedded environment, simulation-games gain in integrative power. One of the definite pitfalls of simulation-games is the fact that we often falsely assume that the game in itself will be powerful enough to cause change or learning that the outcomes will be used ‘automatically’ for decision-making. This is seldom the case. In our experience, the game often ends where it ends. Getting the best out of simulation-games implies that careful attention should be paid to the preamble, the debriefing and follow-up stages.

Acknowledgements This article is based on several games and publications published previously with a number of co-authors. We wish to acknowledge the invaluable contributions to those games and publications of Wieke Bockstael, Roy Chin, Stijn-Pieter van Houten, Joost van Kempen, Armand Moeis, Richard Scalzo, Edwin Valentin, Wijnand Veeneman, Alexander Verbraeck and others. The games CONTAINERS ADRIFT, VENTUM ON LINE and SIM-MV2 are products of the Faculty of TPM / Delft Rotterdam Centre for Process Management and Simulation (CPS, www.cps.tbm.tudelft.nl). We are obliged to the Port of Rotterdam for financing SIM MV2 and to IJsfontein interaction design (www.ijsfontein.nl) for their participation in the VENTUM ON LINE project. This publication has been made possible by the co-financing of the Next Generation Infrastructures program (NGI, www.nginfra.nl). References [1] Stacey, P., Digital game based learning & simulation, in E-Learning. 2003. [2] Mayer, I. and W. Veeneman, Games in a world of infrastructures simulation-games for research, learning and intervention. 2002, Delft: Eburon. 268 blz. [3] Brewer, G. and M. Shubik, The war-game. 1979, Cambridge: Harvard University Press. [4] Crookall, D., A guide to the literature on simulation/gaming, in Simulation and gaming across disciplines and cultures, D. Crookall and A. Kiyoshi, Editors. 1995, Thousand Oaks: Sage. p. 151–177. [5] Davies, P., Simulation: bringing e-learning to a new level, in Computeruser. 2003. [6] Crawford, C., Serious games: Improving Public Policy through Game-Based Learning and Simulation. no date. [7] Aldrich, C., Simulation and the Future of Learning: an Innovative (and perhaps revolutionary) Approach to E-learning. 2004, San Francisco: Pfeiffer. [8] Prensky, M., Digital game-based learning. e-book ed. 2001: McGraw-Hill. 442. [9] Sawyer, B., Serious Games: Improving Public Policy through Game-based Learning and Simulation. 2002. [10] Dziabenko, O., et al. A web-based game for supporting game-based learning. in GAME-ON Conference. 2003. London, United Kingdom. [11] Harris, P.I., Retrieved August 30, 2004 from, Sims, Sims, Sims. Learning Circuits, 2003. [12] Morrison, J.A., C. (2003) In: Retrieved 30th August 2004 from: http://ts.mivu.org/default.asp?show= article&id=2032, Simulations and the-learning Revolution: an Interview with Clark Aldrich, in Vision. 2003. [13] Pivec, M., O. Dziabenko, and I. Schinnerl. Aspects of game-based leanring. in I-KNOW 03, the Third International Conference on Knowledge Management. 2003. Graz, Austria. [14] Maharg, P. Authenticity in Learning: transactional learning in Virtual Communities. in Bridging the Gap: transforming knowledge into action through gaming and simulation. 2004. Munchen: SAGSAGA. [15] Maharg, P., Virtual Communities on the web: Legal Skills in a Virtual Community. International Review of Law Computers & Technology, 2001. 15(3): p. 345–360. [16] Hisschemoller, M. and R. Hoppe, Coping with untractable controversies. The case for problem structuring in policy design and analysis. Knowledge and policy. The international journal of knowledge transfer and utilization, 1996. 4(8): p. 40–60.

I. Mayer and G. Bekebrede / Serious Games and ‘Simulation Based E-Learning’

151

[17] Bruijn, J.A.d., E.F.t. Heuvelhof, and R.J.i.t. Veld, Procesmanagement over procesontwerp en besluitvorming. 2e dr. ed. Bedrijfskundige signalementen. 2002, Schoonhoven: Academic Service. 199 blz. [18] Mayer, I.S., W. Bockstael-Block, and E.C. Valentin, A Building Block Approach to Simulation: An Evaluation Using Containers Adrift. Simulation and Gaming, 2004. 35(1): p. 29–52.

On the Web sg.comp.nus.edu.sg www.fablusi.com www.socialimpactgames.com rmc.library.cornell.edu/games/intro.html www.absel.org www.ardcalloch.ggsl.strath.ac.uk/introduction/index.htm www.cps.tbm.tudelft.nl www.digra.org www.gsgc.org www.ijsfontein.nl www.isaga.net www.kodos.nl www.marcprensky.com/writing/nasaga/MPrensky-nasaga.html www.nasaga.org www.nginfra.nl www.seriousgames.org www.sig-glue.net www.simenco.nl www.simulearn.net/SimuLearn/simulearn_home_page.htm www.ventum.nl

This page intentionally left blank

Motivation and Learning

This page intentionally left blank

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

155

Learning and Motivation with Virtual Tutors. Does It Matter if the Tutor Is Visible on the Net? Manuela PAECHTER a and Karin SCHWEIZER b a Karl-Franzens-University, Graz, Austria b University of Mannheim, Germany

1. Introduction A growing number of research projects are focusing in the use of the Internet in university courses for distance and on-campus education. The convergence of technical and institutional factors is contributing to the importance of these projects. At the technological level, telecommunication services enable to link students with courses, tutors, and educational institutions in a rapid and efficient way. At the institutional level, institutions such as universities are faced with the necessity to provide students with online services reigning from mere access to online libraries to e-learning seminars and lectures. Yet, many questions of how to develop and implement pedagogically effective frameworks of internet-based distance education still remain unsolved. In this paper, a research questions which is concerned with social processes in the classroom was investigated: “How important is information about a tutor or lecturer in an online seminar?” Is social information about the tutor’s appearance or his or her voice important for learning? These questions concerning the role of a tutor and the form of communication between a tutor and students was investigated in a university seminar. It was analyzed whether the absence or presence of social and personal cues in the communication between a tutor and his or her students influence students’ learning and their satisfaction with the tutor and the course. 2. The Tutor’s Role in Offline- and Online Learning Processes Nearly twenty years ago, the German Educational Scientist Hartmut von Hentig published a book with the title “The gradual disappearance of reality” (this is the translation of the German title “Das allmähliche Verschwinden der Wirklichkeit”, Hartmut von Hentig, 1987). In his book von Hentig assumes a harmful influence of digital media, i.e. of computers and the internet. According to his opinion, media and especially digital media can always convey only a semblance of reality. Hence, in a world of media it becomes more and more difficult for humans to gain authentic “real” experiences. In a research project, we took on this idea and asked ourselves what would happen if the persons involved in learning processes “disappear” in a virtual learning environment. How will it affect learning if the tutor and the learners become “invisible” on the internet in the sense that they can only communicate by written messages and can no

156

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

longer use nonverbal, pictorial or audible information such as mimics, gesture, or voice intonation to convey information about their mood, their opinions, or their attitudes? In our project, we focused on the role of the tutor and investigated whether students’ achievement and their satisfaction are influenced by the type and the amount of personal information they receive about their tutor. 2.1. The Tutor’s Tasks in Instructional Design Theories When designing a virtual seminar one may ask oneself whether a human tutor has to look after the students, to instruct, and to advise them. Tasks or instructional events such as presenting the learning contents, offering exercises, or giving feedback probably can be carried out by a computer system and need not to be carried out by a human person. However, is it really that easy? Besides, instructional design theories assign more than only these tasks to a tutor. In their instructional design theory, Aronson and Briggs (1983) list several instructional events that occur in a learning situation and that are to be provided for by the tutor: Tutors, on the one hand, are faced with tasks referring to the learners’ cognitive processes. They have to present and to explain the learning material, to stimulate the recall of prerequisite knowledge, to enhance retention and transfer, and to assess the learners’ performance. On the other hand, they also are faced with tasks which aim at the learners’ interest and motivation. They have to gain the learners’ attention, to provide guidance, to arouse curiosity, and to stimulate motivation. Besides, tutors should give feedback, a task which aims at cognitive as well as at motivational processes. Feedback does not only provide information about learning achievements but also serves to stimulate and to motivate a learner. As this description shows tutors are faced with various tasks that do not only address cognitive but also emotional, motivational issues of learning. Besides, these tasks need an interaction between a tutor and the learners in which the tutor explains, answers questions, gives feedback etc. 2.2. Influence of the Tutor-Student Interaction on Learning Processes The most famous investigation of the interaction between students and a teacher is certainly Rosenthal and Jacobson’s “Pygmalion in the classroom” study (1968) which took part in a public elementary school. At the beginning of the school year, Rosenthal and Jacobson gave an intelligence test to the students. They told the teachers that this test could identify those students who would make rapid, above-average intellectual progress in the coming year, whether or not they were currently “good” students. The teachers received the names of those students who could be expected to perform well in the next school year. Actually, however, Rosenthal and Jacobson had randomly selected the students. The test was one of the usual intelligence tests and did not identify academic potentials as the teachers had been led to believe. A second intelligence test was administered at the end of the year. Those students who had been identified as academic potentials showed, on average, a higher increase on their IQ scores than the rest of the students. Rosenthal and Jacobson concluded that a self-fulfilling prophecy was at work. The teachers had subtly and unconsciously encouraged the performance they expected to see. Not only did they spend more time with these students, they were also more enthusiastic about teaching them and unintentionally showed more warmth to them than to the other students.

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

157

While the Pygmalion study shows how powerful teacher expectations can be it does not reveal how (positive or negative) expectations are conveyed in the interaction between a teacher and a student. In a subsequent study, Brophy and Good (1986) observed teacher-student interaction more closely and concluded that teachers may unconsciously send different messages to low achievers than to high achievers. Low achievers often receive shallow praise, less feedback, and more criticism. In addition, these students tend to be called on less often and they are given less time to respond. The results from the Brophy and Good study indicate that teachers may be sending subtle, nonverbal cues that they expect less of certain students. Students do not miss these cues and they react accordingly. Coates and Smidchens (1966) also investigated the interaction between teachers and students. They observed that students learn more successfully and develop more interest if the teachers show their enthusiasm in the classroom. The researchers identified a teaching style which they called dynamic vigor. Teachers with dynamic vigor show more emotion, they express their own attitudes and opinions in the classroom. Dynamic vigor is mainly expressed by nonverbal cues such as the voice intonation. The three studies emphasize the importance of the interaction between tutors and students. The personality of the teacher, the emotions, and the interest he or she communicates in the classroom influence the students’ motivation and achievements. What, however, happens if such information cannot be conveyed in an internet-based seminar? What happens if the tutor and the students can only exchange written messages and cannot send nonverbal information? 2.3. Online Communication and Interaction In virtual seminars, communication is reduced to visual text when internet services such as newsgroups, e-mail, or chats are used for the interaction between a tutor and the learners or between groups of learners. Daft and Lengel (1984) distinguish communication across media according to the richness of the communication. Face-to-face communication can be considered as the richest as it allows verbal, paraverbal, visual channelling and backchannelling cues, and immediate feedback. Compared to face-to-face interaction, videoconferencing is still moderately rich and certainly richer than telephone interaction. Communication based on visual text messages without pictorial (facial expression, gestures) and paraverbal (intonation of the voice) information about the speakers can be regarded as being rather lean (McGrath & Hollingshead, 1993; Paechter, Schweizer & Weidenmann, 2001). Closely related to the richness of media is the concept of “social presence” (Short, Williams, & Christie, 1976). A medium with a high degree of social presence allows a personal communication in which participants feel that they are jointly involved in the interaction and in which personal attitudes of the communication partners can be perceived. In such a communication pictorial, verbal, and paraverbal information of the communication partners are conveyed. When cues about the interaction partners are filtered out social presence declines. Social presence theory as well as related theories as the “theory of reduced context cues” (Kiesler, Siegel, & McGuire, 1984) are characterized as a “cues-filtered-out-perspective” (Culnan & Marcus, 1987). Theories within this framework assume very stable and distinct consequences of a communication in which social cues are omitted, e.g. disturbing behaviour, greater selfabsorption versus other-orientation, and a negative group atmosphere (Dubrovsky &

158

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

Sethna, 1991). Lean communication also poses problems to the coordination of communication when visual and auditory signals of the speakers cannot be conveyed. Sproull and Kiesler (1986) accentuate the lack of cues referring to a person’s social context, her or his personality, status, etc. The consequences are, for example, depersonalization and de-individuation problems as well as difficulties in turn-taking and disorganization (Kiesler, Siegel & McGuire, 1984). Hiemstra (1982, p. 883) concluded: “As bandwidth narrows from face-to face interaction, the communication is likely to be experienced as less friendly, emotional, and personal, and more serious, businesslike, depersonalized, and task-oriented.” Research results within the cues-filtered-out perspective have been derived from experimental studies with zero-history groups who met for only one point in time. Besides, the research primarily emerged in the domain of synchronous group conferencing and organisational e-mail. Communication in distance education differs in many aspects from these settings: the tutor’s status can be perceived distinctively even though social cues may be missing; she or he has not only a higher status but is also an expert in the respective domain of knowledge. Moreover, experiments within the cuesfiltered-out perspective focused primarily on dependent variables such as decision making, group atmosphere, or choice of media and not on learning processes. Social presence theory and the theory of reduced social context cues assume mainly negative effects of computer-mediated communication. These assumptions are challenged by the social-information-processing perspective (Walther, 1992; Walther, Slovacek, & Tidwell, 1999). This theory predicts different effects of computermediated communication. It assumes, firstly, that communicators are motivated to develop impressions and relations despite possible obstacles imposed by media; secondly, that users adapt their efforts to present social information to the media and probably even develop new codes to express social information, and, thirdly, that the development of relations in computer-mediated communication takes more time than in faceto-face communication. If, however, enough time is provided relations based on computer-mediated communication can achieve the same intensity, depth, and quality as relations developed in a face-to-face communication. Empirical analyses support the social information processing perspective: In unlimited interaction computer-mediated communication was more socio-emotionally oriented than in time-limited interaction (Walther, Anderson, & Park, 1994). Time is a crucial factor in this framework. Interaction partners adapt to new circumstances of communication and with progressing time relations and impressions in computer-mediated communication normalize to the level of face-to-face communication. 3. An Empirical Investigation of Learning and Motivation in a Virtual Seminar How important is the visibility and audibility of the tutor, in other word his or her social presence, for the learners’ achievement and for their satisfaction in a virtual seminar? This question was analyzed in an empirical study in which the availability of pictorial and auditory, social, personal information about a virtual tutor was varied. 3.1. Design of a Virtual Seminar For the investigation a virtual seminar was designed and carried out for one term (9 weeks). In the seminar a central web page offered admittance to different communication and presentation services.

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

159

(Translation: Hörsaal = seminar room; Bibliothek = library; Schwarzes Brett = notice board; Sprechstunde = consulting hours – tutor’s office; Testraum = examination room; Studententreff = cafeteria) Figure 1. Start page in the internet for a virtual seminar.

The page offered admittance to the following facilities: • • • • • •

The virtual seminar room contains learning material in the form of computerbased training. The learning material was realised as a tutorial. Students could also obtain these tutorials offline on CD-ROM. A virtual library offers additional learning material, e.g. texts, short computerbased training material, or software referring to the learning contents. General issues about the course or new information can be found on the notice board. In the cafeteria students may communicate simultaneously with each other. Technically, the cafeteria was organised as a chat room. If students want to communicate with the lecturer they have to visit the tutor’s office. It offers the opportunity to send an e-mail to the tutor and to load down his or her messages. The examination room was a special feature of our study. Students had to fill in tests on the learning material and questionnaires on the seminar.

3.2. Research Design 3.2.1. Variation of the Tutor’s Social Presence In the seminar, the social presence of the virtual tutor was varied in four degrees. 1.

The tutor communicated only via e-mail, i.e. by written text and asynchronously with each tutee. In the tutor’s office students received e-mails of the tutor, on the computer-based learning material a balloon indicated a personal message of the tutor. Most of the tutor’s messages were standardized (e.g. on

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

160

In the first screenshot from the learning material (condition 1) a ballot indicates stimulating and motivating information from the tutor. On the second screenshot (condition 2) the same information is accompanied by a personal view of the tutor. Figure 2. Examples for the tutor’s social presence (condition 1 and 2).

2.

3.

4.

the learning material). There, however, was also spontaneous contact, e.g., when the tutor had to answer students’ e-mails (compare Fig. 2). As in condition 1, the tutor communicated only by visual text. The text, however, was accompanied by a personal view, a picture of the tutor corresponding to the respective communication content (compare Fig. 2). The tutor communicated via spoken messages. The same system as in condition 1 was used but, instead of reading an e-mail, the tutee had to open an audio file and to listen to the tutor’s message. On the learning material, students heard voice recordings at specific instructional events. On the computer-based training spoken text was repeated in its visual form in order to prevent misunderstanding. It was controlled that spoken (asynchronous) spontaneous messages did not differ from written ones with regard to style and form. The tutor communicated by spoken messages as in condition 3; in addition, a personal view of the tutor corresponding to the respective communication content was shown.

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

161

It was assumed that the tutor’s social presence is lowest in condition 1 and highest in condition 4 when students see the tutor’s picture and obtain also paraverbal information by listening to her spoken statements (the tutor in the course was female). Students received messages from the tutor either in her “office” (e.g., when obtaining feedback on test results or answers to questions), on the notice board (e.g., the tutor’s introduction of herself), or on the learning material from the lecture hall. Not all instructional events were presented audibly and eventually embellished by a personal view but rather those in which a high degree of social presence seemed supportive to the learning processes. These instructional events had been derived from Aronson and Brigg’s instructional theory (1983). They included learning situations such as “providing learning guidance”, “providing feedback”, “motivating” etc. When students wished to communicate with the tutor they had to use the text-based e-mail system in the virtual tutor’s office. From the tutor’s point of view students communicated only via textbased e-mail and hence with rather low degree of social presence. The virtual seminar described here was carried out for one term (about 9 weeks) at the University of the Federal Armed Forces, Munich. The topic of the course was an introduction to Psychology with aspects such as “Knowledge structures”, “Learning with text”, or “Learning with pictures”. At the University of the Federal Armed Forces in Munich every student is obliged to attend an introductory course in Educational Sciences or Psychology. Therefore, students from all faculties (computer sciences, business administration, engineering etc.) attended the course. The course was organised so that every second week students received new learning material and had to fill in a test (in this case one speaks of a predetermined distribution of the learning material; Kerres, 1997/1998). Within these two weeks they could decide when, where, and how to work with the learning material. Over the course of nine weeks students received four different learning units and filled in four tests with questions to the learning material. 3.2.2. Students’ Achievements and Evaluations of the Tutor After each of the two-week long learning periods, students filled in a test on the learning contents and were asked to evaluate the tutor. In summary, three measurements were obtained: •

•

•

Students’ achievements: The students’ answers to the questions to the learning material were analyzed and the percentage of correct statements was recorded. For the statistical analyses described later the mean of all answers in one test was computed. Students’ evaluation of the tutor’s ability to motivate them: Students evaluated the tutor’s ability to motivate them by three ratings on a seven-point scale: whether the tutor was friendly, whether she showed enthusiasm, and whether the communication was inspiring for the learning. Students’ evaluation of the tutor’s ability to impart knowledge: After each learning period, students assessed by seven ratings whether the tutor could explain difficult learning contents, whether she fostered the reflection of the learning contents etc.

For the statistical analyses of the evaluations, for each point in time and each student the mean of the three (evaluation of motivation) respectively seven ratings

162

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors Table 1. Dependent variables in the four learning conditions.

Written messages (1)

Written messages + picture (2)

Spoken messages (3)

Spoken messages + picture (4)

Achievement (%)

66.97

66.32

66.49

72.25

Evaluation of tutor’s ability to motivate

4.04

4.77

4.11

4.40

Evaluation of tutor’s ability to impart knowlege

4.38

4.27

4.40

4.22

Number of cases

23

25

25

25

(evaluation of ability to impart knowledge) was computed; then the mean of all three points in time was computed. 3.3. Empirical Results 101 male students of the University of the Federal Armed Forces, Munich, studying different subjects (engineering, educational sciences, computing etc.) but all in the second term of their study participated in the experiment. Due to their different technical equipment the data of three students were not included into the study. Nearly all of the students were between 20 and 24 years old. 25 respectively 26 students were allocated to each of the four treatments. It was recorded how often every student used the services of the internet environment. In summary, students accessed the internet environment 3452 times. 756 times they filled in tests in the examination room, 414 times they visited the virtual library, 913 times the notice board, 1208 times they read the tutor’s messages and 161 times they sent messages to the tutor. The tutor wrote 570 messages (many of these were sent to more than one participant). 3.3.1. Students’ Achievement with Different Degrees of the Tutor’s Social Presence Firstly, it was investigated whether the students’ performance differs with regard to the tutor’s social presence. A multivariate analysis of variance (factors were the form of the communication and the points in time) confirmed this assumption (F = 2.74, df 3, 89, p < 0.05). Students’ performance becomes better as the tutor’s social presence increases. This effect, however, is mainly based on the difference between learning condition 4 and the three other conditions (compare Table 1). Compared to all other conditions, students show better achievements in condition 4. 3.3.2. Students’ Evaluation of the Tutor Secondly, it was investigated whether the students’ assessments of the tutor differ with regard to the form of the communication. Therefore, a multivariate analysis of variance was carried out for the two aspects of the students’ evaluation. Significant differences were found for the evaluation of the tutor’s ability to motivate the students (F = 3.49, df 3, 89, p < 0.05). This effect is due to differences between condition 1 and condition 2. Students evaluate the tutor significantly better in condition 2 than in condition 1. No differences between the four conditions were found for the evaluation of the tutor’s ability to impart knowledge.

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

163

4. Discussion Studies on teacher-student interaction point out that the tasks of a teacher go beyond imparting knowledge. Teachers should also motivate their students, arouse their curiosity and interest, give encouraging feedback, etc. The studies emphasize that the emotions, attitudes, and opinions that teachers communicate in the classroom widely influence students’ learning and their motivation. What, however, happens if in an online seminar affective information cannot be conveyed by the usual nonverbal means such as gesture, mimics, or voice intonation? Does the lack of affective nonverbal information in the teacher-student interaction influence students’ achievement, satisfaction, and motivation? Theories on computer-mediated communication disagree in their answers to these questions. The cues-filtered-out perspective assumes that the lack of affective information will impair learning and motivation in the virtual seminar. For our study it would predict that the students in the condition with the highest level of the tutor’s social presence will show the highest achievement and will score higher on satisfaction with the tutor. The social-information processing perspective would assume that students can adapt to the different circumstances of a communication setting. Students would try to extract affective information from the tutor’s written messages. Hence, the socialinformation-processing perspective would not expect differences between the four learning conditions. The results of our empirical study seem to favor the position of the cues-filteredout perspective. Students’ achievement is highest in the condition with a high degree of the tutor’s social presence. Yet, in their evaluation of the tutor the students distinguished between different instructional tasks. With regard to the evaluation of the tutor’s ability to motivate learners we found only differences between condition 1 (written messages only) and condition 2 (written messages accompanied by a personal view). No differences at all could be found for the evaluation of the tutor’s ability to impart knowledge. The availability of additional information seemed to make a difference for the evaluation of the tutor’s ability to motivate students but not for the evaluation of the tutor’s ability to impart knowledge. Affective information was considered to be only important for those instructional tasks which aim at the motivational processes of learning. Contrary to the expectations of the cues-filtered-out perspectives the addition of audible information to the personal views in condition 4 did not lead to a better evaluation of the tutor. The tutor received the best evaluations in condition 2 (written messages plus personal views). Probably not all types of personal information are equally important and possibly pictorial information is more important than audible information. Here, however, further research needs to be carried out. When interpreting the research results one should take into consideration that a special form of communication was investigated. In the communication among peers it is certainly easier to gather information about the communication partner’s emotions or attitudes, e.g. by directly asking him or her about his/her opinions or moods. This is certainly much more difficult or even not possible in the communication between a tutor and the students. Therefore, students could not actively gather affective information about the tutor. Neither the cues-filtered-out theories nor the social information processing perspective take this aspect into account. Here, also further research needs to be carried out.

164

M. Paechter and K. Schweizer / Learning and Motivation with Virtual Tutors

Literature Aronson, D.T. & Briggs, L.J. (1983). Contributions of Gagné and Briggs to a prescriptive model of instruction. In C.M. Reigeluth (Ed.), Instructional-design theories and models: An overview of their current status (pp. 75–100). Hillsdale, N.J.: Lawrence Erlbaum. Brophy, J.E. & Good, T.L. (1986). Teacher behavior and student achievement. In M.C. Wittrock (Ed.), Handbook of research on teaching (pp. 328–375). New York, NY: Macmillan. 3rd edition. Coates, W.D. & Smidchens, V. (1966). Audience recall as a function of speaker dynamism. Journal of Educational Psychology, 57, 189–195. Culnan, M., & Marcus, M.L. (1987). Information technologies. In F.M. Jablin, L.L. Putman, & K.H. Roberts (Eds.), Handbook of Organizational Communication. An interdisciplinary perspective (pp 420–444). Newbury Park: Sage Publications. Daft, R.L. & Lengel, R.H. (1984). Information richness: A new approach to managerial information processing and organization design. In B.M. Staw & L.L. Cummings (Hrsg), Research in organizational behaviour (pp. 191–233). Greenwich: JAI Press. Dubrovsky, S. & Sethna, B.N. (1991). The equalization phenomenon: Status effects in computer-mediated and face-to-face decision making groups. Human Computer Interaction, 6, 119–146. Hiemstra, G. (1982). Teleconferencing, concern for face, and organizational culture. In: M. Burgoon (Ed.) Communication Yearbook 6 (pp. 874–904). Beverly Hills, CA.: Sage. Kerres, M. (1997/1998). Telelernen – Szenarien, Technik und Didaktik. PAE – Arbeitshilfen für die Erwachsenenbildung, 3, 3–4. Kiesler, S., Siegel, J., & McGuire, T.W. (1984). Social Psychology aspects of computer-mediated communication. American Psychologist, 39, 1123–1134. McGrath, J.E. & Hollingshead, A.B. (1993). Putting the “group” back in group support systems: Some technical issues about dynamic processes in groups with technological enhancements. In L.M. Jessup & J.S. Valacich (Eds.), Group support systems (pp. 78–96). New York: MacMillan. Paechter, M., Schweizer, K., & Weidenmann, B. (2001). When the tutor is socially present or not. In U.D. Reips & M. Bosnjak (Eds.), Dimensions of Internet Science (pp. 305–321). Lengerich: Pabst. Rosenthal, R. & Jacobson, L. (1968). Pygmalion in the classroom. New York. Short, J., Williams, E., & Christie, B. (1976). The social communications of telecommunication. Communication Research, 19, 52–90. Sproull, L. & Kiesler, S. (1986) Reducing social context cues. Electronic mail in organizational communication, Management Science, 32, 1492–1512. Walther, J.B. (1992). Interpersonal effects in computer-mediated communication. Communication Research, 19, 52–90. Walther, J.B., Anderson, J.F., & Park, D. (1994). Interpersonal effects in computer-mediated interaction: A meta-analysis of social and antisocial communication. Communication Research, 19, 50–88. Walther, J.B., Slovacek, C.L., & Tidwell, L.C. (1999). Is a picture worth a thousand words? Photographic image in long-term and short-term computer-mediated communication. Unpublished paper presented at the annual meeting of the International Communication Association, May 1999, San Francisco. von Hentig, H. (1987). Das allmähliche Verschwinden der Wirklichkeit: Ein Pädagoge ermutigt zum Nachdenken über die neuen Medien. München: Hanser. 3rd edition. This research project was funded by the Deutsche Forschungsgemeinschaft (German Science Foundation). We wish to thank the Deutsche Forschungsgemeinschaft for the support.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

165

Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning Jürgen HELLER1, Dietrich ALBERT, Michael KICKMEIER-RUST and Markus KERTZ Department of Psychology, University of Graz, Austria

Abstract. In the past, formal models of cognitive psychology successfully contributed to the development of computerized adaptive tutorial systems. Emotional and motivational aspects, however, were rarely considered, although a variety of studies demonstrated their significant influence on learning and performance. The aim of the current pilot study was to investigate whether the sound and empirically valid knowledge space theory is able to cover learning and performance in two different motivational states, which were hope for success and fear of failure. Moreover, within a factorial design these motivational states were combined with two different learning conditions. Pre-structured learning sessions within an adaptive tutorial system were contrasted with rather free text-based learning. The data collected with 104 high school students in the domain of elementary probability theory indicate that knowledge space theory is able to represent the responses obtained in a post-test for both motivational states, as well as for both learning conditions. These results lay out a promising route to integrating cognitive and emotional/motivational aspects into a comprehensive psychological model for adaptive tutorial systems. Keywords. eLearning, adaptive tutorial systems, knowledge space theory, motivation

1. Introduction Developing computer-based tutorial systems has a tradition that reaches back to the nineteen eighties. Triggered by the rapid evolution of technology, we are currently facing a boom of eLearning systems, which provide a powerful technology with the potential of enhancing human learning. This goal, however, is not reached if a purely technology-centered approach is followed. Within a technology-centered approach the development focuses on cutting edge advances of multimedia technology, e.g. the integration of multimedia in communication technology or the development of virtual, interactive systems. However, pure technology-centered approaches failed to lead to lasting improvements in human learning [1]. In 1922 Thomas Edison predicted that motion pictures will supplant largely, if not entirely, the use of textbooks. In 1945 William Levenson predicted that radio receivers will be as common in the classroom as chalkboards and that radio instructions will be an integrated component of every day’s school life (cited in [1], p. 19). The 1 Corresponding Author: Jürgen Heller, Department of Psychology, University of Graz, Austria; Email: [email protected].

166 J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning

promises of 20th century computer technology were very much the same. However, technology-centered tutorial systems failed to produce better learning than traditional teacher-lead instructions [2]. These insufficiencies are possibly due to the fact that instead of technology adapting to human learning, humans were forced to adapt to new technology-driven systems. Consequently, the prediction that eLearning would revolutionize learning turned out to be false. eLearning still plays a subordinate role [3], and even in the currently booming market we are facing a number of prominent insolvencies of eLearning companies [4]. One reason for this might be that many systems only offer digitized contents, more or less refined with some psychological and pedagogical basics. They seem to be nothing more than rich media versions of their predecessors of the nineteen eighties. To develop more successful approaches, a comprehensive and profound cognitive-psychological basis is required. Successful tutorial systems have to adapt to the human learner [5,6]. The focus must be on using multimedia as an aid to human cognition [2]. eLearning platforms have to evolve from systems with strictly hierarchical content presentation to adaptive, personalized tutoring systems. Cognitive and educational psychology provided considerable contributions to foster this development [7–9]. Knowledge Space Theory (KST) suggested by Doignon & Falmagne [10] provides an appropriate psychological framework that can serve as a basis for implementing the required adaptivity of the learning system. We will briefly introduce KST below. Up to now emotional and motivational aspects, however, played a minor role in the development of these systems, and most often they actually were neglected. In order to integrate these factors into the outlined psychological framework, they have to be compatible with the cognitive model. The subsequently presented pilot study is a first attempt to address this issue. 1.1. A Brief Introduction to Knowledge Space Theory Knowledge Space Theory (KST) is a well-elaborated formal psychological theory founded by J.-P. Doignon and J.-C. Falmagne [10]. It provides an intuitive and simple set-theoretic framework for representing and assessing knowledge. The starting point is the notion of a so-called knowledge domain Q, which is nothing else but a set of problems taken from a certain content area, like, for instance, basic algebra consisting of problems involving additions, subtraction, multiplication, and division of positive integers. To provide an example, assume that the knowledge domain Q = {a, b, c, d, e} consists of the five problems a, b, c, d, and e. Now, let the knowledge state of a person be represented by the set of problems the person is capable of solving. Then the knowledge state of an individual is simply a subset of the knowledge domain Q. However, if we look at the solution behavior that a sufficiently large number of subjects exhibits on these five problems then most certainly not all of the possible subsets (there are 2|Q| = 32 subsets in our example) will actually occur. A person who is capable of solving a problem that requires to multiply two positive integers will also be capable of solving a problem that only involves an addition of two positive integers. This means that from a correct solution to the first problem we can surmise a correct solution to the second problem. This kind of mutual dependency is captured in a so-called surmise relation, or prerequisite relation. The left panel of Fig. 1 illustrates an example of a surmise relation, which is represented by line segments that connect some of the nodes corresponding to problems. For instance, from solving problem c we can surmise correct solutions to problems a and b, and an individual capable of solving problem e will solve all remaining problems, too.

J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning 167

(a)

(b) {a,b,c,d,e}

e

c

{a,b,c,d}

d

{a,b,c}

{a,b,d}

{a,b} a

b {b}

{a}

{}

Figure 1. Panel a shows a surmise relation on the knowledge domain Q = {a, b, c, d, e}. The relation is represented by upwards directed sequences of line segments. Panel b shows the corresponding knowledge space.

The surmise relation restricts the possible knowledge states (i.e. subsets of solved problems). To comply with the surmise relation of Fig. 1, for example, each knowledge state containing problem c should also contain problems a and b. The collection of the knowledge states corresponding to a surmise relation is called a (quasi-ordinal) knowledge space, which, for the given surmise relation is {{}, {a}, {b}, {a, b}, {a, b, c}, {a, b, d}, {a, b, c, d}, {a, b, c, d, e}}. Figure 1b provides a diagram illustrating this knowledge space. The upwards directed line segments (representing set-inclusion) may be interpreted as the possible learning paths leading from the naïve knowledge state {} to the knowledge state of full mastery Q = {a, b, c, d, e}. They may serve as a basis for personalized teaching that matches to the student’s current knowledge state. If a student is in knowledge state {a, b}, for example, then content related to problems c or d should be presented next to allow for learning to take place (i.e. transition into knowledge state {a, b, c} or {a, b, d}, respectively). Knowledge spaces also form the basis for an efficient adaptive knowledge assessment. The rationale of the assessment procedure is exemplified below. Observing a correct response to problem d implies that, given the knowledge space of Fig. 1, the corresponding knowledge state is one of {a, b, d}, {a, b, c, d}, or {a, b, c, d, e}. If then problem c is not solved, we have uniquely identified the knowledge state to be {a, b, d} by presenting only two out of the five problems constituting the knowledge domain. Knowledge Space Theory, as a formal approach to a psychological representation of knowledge, and various of its extensions [11,12] have been implemented to provide adaptive knowledge assessment as well as personalized knowledge acquisition. Implementations in adaptive tutorial systems, for instance, include the research prototype RATH [13–15], AdAsTra [16], and the commercial eLearning platform ALEKS [11].

168 J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning

1.2. Emotional and Motivational Aspects of eLearning Emotional and motivational aspects play an important role in acquiring new knowledge as well as for recalling previously learned knowledge in testing situations [17–19]. [20] emphasize the significant impact that emotional and motivational aspects have on cognitive structures, which, however, is commonly not reflected in the development of computerized eLearning systems. There are some formal models that represent emotional and motivational concepts, and might provide a promising basis for an extension of the existing eLearning systems. In particular, we want to mention the Hull-Spence-Spielberger anxiety-performance theory [21–24], and its competitors, the anxiety-learning theory by Albert [25] on the one side, and the achievement motive and risk taking behavior theory by McClelland and Atkinson [26–29] on the other side. As an example let us consider the model of Atkinson and McClelland, as it refers to well-known and extensively studied concepts. Atkinson distinguishes between an achievement motive and an avoidance motive. The tendency to achieve success TS is computed as the product of the subjective probability of success PS (hope for success), the achievement motive MS, and the anticipated positive effects of success IS (e.g. the expected amount of pleasure) TS = PS * (MS * IS).

(1)

Similarly, the tendency to avoid failure TF is defined as a product of the subjective probability of failure PF, the avoidance motive MF, and the anticipated negative effects of failure IF (e.g. anger, shame) TF = PF * (MF * IF).

(2)

Consequently, the resultant motivational tendency TR to tackle a task or to avoid it is TR = TS + TF = PS * (MS * IS) + PF * (MF * IF).

(3)

In this formalization we may additionally assume that the more difficult a task is, the prouder is a person in case of success (IS = 1 – PS). Likewise, the easier a task is, the more ashamed is a person in case of failure (IF = – PS). Together with the fact that PF + PS = 1 (exactly one alternative will happen) we get TR = TS + TF = MS * [PS * (1 – PS)] – MF * [PS * (1 – PS)]

(4)

= (MS – MF) * [PS * (1 – PS)].x The predictions that can be derived from this formula are illustrated in Fig. 2. According to the two plotted functions, persons with a relatively high achievement motive (MS > MF) will prefer problems with intermediate subjective success probability, while persons with a relatively high avoidance motive (MF > MS) are supposed to prefer problems with more extreme subjective success probabilities. These predictions may be contrasted with the navigation behavior that can be observed in eLearning environments.

J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning 169

Motivational Tendency (T) R

+

TS> TF

(MS > MF)

TR

0.00

0.50

TS

1.00

0.00

0.50

TF

TR TF> TS

-

1.00

(M1 > MS)

Subjec tive probability of success (P) S

Figure 2. Illustration of the resultant motivational tendency in the model of Atkinson and McClelland.

The outlined motivational model focuses primarily on approach and avoidance tendencies. Supplementary studies have shown that, for the present context, the model will have to be extended to incorporate additional factors. The observed effects will not only depend on approach and avoidance tendencies [30–32] but also on situational controllability of behavioural results [33], and internal or external causal attribution [34,35]. It is, however, beyond the scope of the present paper to provide a more detailed account of this issue. 1.3. Combining Cognitive Approaches with Emotional and Motivational Aspects Combining cognitive models (like KST) with formal models of emotional and motivational factors seems to be a promising route to arrive at a more comprehensive psychological model of human learning in computerized environments. So far, we have represented learning as a transition between knowledge states (e.g. from {a, b} to {a, b, d}). The probability of this transition, however, will not only depend on the current knowledge state, but also on the available relevant skills and competences. There are various extensions of KST that integrate these kinds of psychological constructs [12]. Moreover, probability of learning will heavily depend on emotional and motivational states, and these will strongly interact with cognitive states. A simple example may suffice to clarify this dependence. If a student successfully acquires new knowledge then this is likely to increase the motivation, which in turn may increase the probability of learning related content. Representing moderator effects like this calls for further extending KST. Introducing latent emotional/motivational states, however, requires that they are compatible with the structural assumptions in the knowledge space. This is completely in line with the fundamental compatibility of the underlying skill, or competence structures with the knowledge space [11,12]. The question whether we are able to represent the observed solution behaviour under different motivational states within the same knowledge space has to be answered empirically. The subsequently presented pilot study takes a first step into this direction. Since most of the previous investigations on motivational and emotional aspects were conducted in noncomputerized environments, the experiment includes different types of learning condi-

170 J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning

tions, which are computer-based learning and common text-based learning. The experimental setup thus also allows for detecting possible interaction effects.

2. Method 2.1. Design In a 2x2 factorial design we investigated two emotional/motivational states, hope for success (HS) and fear of failure (FF), and two learning conditions, learning within an adaptive tutorial hypertext system and traditional text-based learning. The experiment was divided into two parts, an initial session for assessing motivational states, and a subsequent learning and testing session. Parallel groups in both learning conditions were obtained by forming matched pairs with respect to motivational state (FF, HS). Achievement was measured by a paper-and-pencil test administered after the learning phase. 2.2. Material and Apparatus The Multi-Motive-Grid (MMG), a diagnostic tool proposed by [36], was applied to differentiate between the two different motivational states hope for success and fear of failure. It consists of 14 test items that resemble those of the TAT. The MMG is a reliable projective technique that uses standardized response categories instead of a free response format. The two learning conditions consisted of the presentation of a course on elementary probability either in a print-out (text-based learning), or employing the adaptive tutorial system RATH (available at http://wundt.uni-graz.at/rath). The research prototype RATH [13] is a relational adaptive tutoring hypertext WWW environment based on KST [10,11] and a relational hypertext model [15]. The system adaptively responds to a students knowledge state. Contents, examples, and exercises are presented “just-in-time” according to the current state. Too easy and too difficult content is hidden. Thus, RATH guides a student adaptively through the structure of learning objects along one of the possible learning paths. RATH is a server-side application, i.e. students connected to RATH with school computers via network connections. For traditional text-based learning (TBL) a folder was created that contained printouts of the complete course including examples and exercises. The main difference between both learning conditions is that in the RATH condition students were adaptively guided by the system, while in the TBL condition students were free to explore each part of the course at any time, irrespective of the sequence of the material in the folder. The learning material consisted of a course on elementary probability theory by Held [37]. He identified ten cognitive demands that are required to understand the learning material and, therefore, to solve all related problems. The ten demands comprise capabilities ranging from understanding simple Laplace probabilities to a general understanding of events and their probabilities. These demands were supplemented by an additional demand, the understanding of the concepts of random experiments, results, and events [13]. From the resulting eleven demands six problem classes (A to F) were constructed. Thus, the knowledge domain is given by Q = {A, B, C, D, E, F}. This

J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning 171 {A,B,C,D,E,F} 10 {B,C,D,E,F} 9 {B,D,E,F}

5,6

5,6

{C,D,E,F} 9

8

{D,E,F}

{C,E,F}

7

9

8

{D,F}

{E,F} 8

7 {F} 0,1,2,3,4 {}

Figure 3. Diagram of the knowledge space on the knowledge domain Q = {A, B, C, D, E, F} of problems on elementary probability theory. The numbers refer to the required demands. The diagram is taken from Hockemeyer ([13], p. 42). Please notice that the knowledge state {C, D, E, F} is missing in the original figure of [37].

knowledge domain is assumed to be structured according to the knowledge space illustrated in Fig. 3. It shows the ten knowledge states (subsets of Q), and indicates the demands (numbers 0 to 10 labeling the arrows) that have to be mastered to be able to move from one state to another. Achievement was measured by a paper-and-pencil test administered after the learning phase. The total of 18 problems created by [37] resulted from three instances for each of the six problem classes A to F. The problems were presented in an order individually randomised for each of the participants. The complete course and all the problems are in English language. 2.3. Subjects To participate in the present study the subjects had to meet two requirements. On the one hand, they should be able to understand the used learning material (e.g. its form and content). On the other hand, it was necessary to reduce the amount of preknowledge to a minimum. Thus, we recruited a total of 116 students from six Austrian high school classes (7th and 8th grade). Due to drop-outs the presented results are based on 104 of these students. There were 64 female and 40 male students with an age range from 16 to 19 (M = 16.6, SD = 0.74). 2.4. Procedure The study was conducted as a group experiment in the class-room of each of six high school classes, and consisted of two phases. In an initial session, after a general introduction, the students filled in a form querying personal data, like age, gender, pre-

172 J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning

knowledge in elementary probability theory, mathematics, and English grades. After that the MMG was administered with a time limit of 30 minutes. According to the results of the MMG matched pairs were formed with respect to motivational state, which resulted in 21 pairs for the motivational state hope for success, and 31 pairs for the motivational state fear of failure. The members of each pair then were randomly assigned to one of the two learning conditions. The time period between initial and main sessions was 7 to 16 days (identical for all students of the same high school class). The main session consisted of a learning phase that lasted 75 minutes. In the text-based learning condition the students were presented with a print-out of the course material, while in the computer-based learning condition they were able to work with the adaptive tutorial system RATH. Please remember that in the text-based learning conditions the students were free to explore the material in any order. In the RATH condition the students were forced to process the lessons, examples, and exercises according to the knowledge space illustrated in Fig. 3. After the learning phase and a ten minutes break, the students had take the same paper-and-pencil test in both learning conditions. For completing the test there was a time limit of 40 minutes.

3. Results The main interest of this pilot study was to investigate whether KST can cover different motivational states (HS, FF) of a student in different learning conditions (RATH, TBL). Thus, we analysed the deviations of response patterns in the paper-and-pencil test from the underlying knowledge space proposed by Held [37]. For this we applied a strict criterion for deriving the response patterns over the problem classes from the obtained results. A problem class (A to F) was considered to be solved only if all of the three associated problems were solved correctly. Two indicators of goodness-of-fit of a knowledge space were used as dependent variables. The minimal symmetric distance is defined as the least number of differing elements that result when a response pattern is compared to all knowledge states in a given knowledge space. A minimal symmetric distance of 0 indicates that the response pattern is identical to a knowledge state, a value of 1 indicates that it coincides with a knowledge state only after adding or deleting a single problem, and so on. We consider the distribution of the minimal symmetric distance over subjects as well as its average. The second indicator refers to the compatibility of the order of the problem-specific solution probabilities with the assumed surmise relation. The number of solved problems is used as an additional variable indicating a level of achievement that is independent of the assumed knowledge space. To analyse the deviations between the response patterns and the knowledge space we used the software tool DI, version 2.4.5 from the Knowledge Space Tools package by Held, revised by [38]. DI computes the minimal symmetric distances for each response pattern. Figure 4 shows the resulting average of minimal symmetric distances for all experimental conditions. In the RATH condition smaller values result (FF: .19; HS: .14) than in the textbased learning condition (FF: .32; HS: .33), whereas in the HS and FF conditions the values are very close (.31 versus .35). Figure 4 seems to suggest a slight interaction between the factors. Table 1 provides the distributions of the minimal symmetric distances for all experimental conditions. It shows that over all conditions a majority of response patterns

Average of minimal symmetric distance

J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning 173

Fear of failure Hope for success RATH

TBL

Figure 4. Average of minimal symmetric distances between response patterns and proposed underlying knowledge structure. Table 1. Distributions of minimal symmetric distances for all experimental conditions. The table provides absolute frequencies and percentages (in parentheses) of response patterns for each distance. Distance

RATH

TBL

0

25 (80.65)

23 (74.19)

1

6 (19.35)

6 (19.35)

2

0 ( 0.00)

2 ( 6.45)

3

0 ( 0.00)

0 ( 0.00)

FF

0

15 (71.43)

1

3 (14.29)

5 (23.81)

2

0 ( 0.00)

1 ( 4.76)

3

0 ( 0.00)

0 ( 0.00)

HS

coincide with knowledge states, ranging from 71.43% in the TBL/HS condition to 80.65% in the RATH/FF condition. The RATH condition seems to yield a somewhat higher percentage of patterns in accordance with the knowledge space. For a statistical test we lump together the categories with distances 2 and 3. Then the difference between learning conditions turn out to be non-significant in a chi-square test (X2 (2) = 3.51, p = .173, Fisher’s Exact Test p = .210). Moreover, there are no differences between motivational states FF and HS (X2 (2) = 0.67, p = .967, Fisher’s Exact Test p = 1.000). Applying a multinomial logit model with main effects provides a close to perfect fit between observed and predicted frequencies. Consequently, the resulting chisquare statistic (based on the likelihood ratio) leaves no room for an interaction effect (X2 (2) = 0.3503, p = .839), contrary to what Fig. 4 seems to suggest. The results based on the problem-specific solution frequencies also provide ample evidence that there are no differences concerning the goodness-of-fit of the supposed knowledge space over experimental conditions. Figure 5 shows the obtained relative frequencies plotted in the graph of the respective surmise relation. For all experimental conditions these frequencies decrease the higher the ‘difficulty’ of the problem class, which is to be expected if the response patterns conform with the knowledge states. The results also indicate a slightly better performance in the TBL condition. In order to evaluate the result of the test independent of the assumed knowledge space we consider the number of solved problems, relative to the total of 18 problems that were presented. Due to the time limits, however, the students were often not able

174 J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning RATH / FF

TBL / FF

.05

.23

A

A

.10

B

C .41

.16

B

C .54

.66

D

E .48

.67

D

E .55

F

F

.87

.72

RATH / HS

TBL / HS

.06

.21

A

A

.08

B

C .25

.46

B

C .57

.51

D

E .30

.76

D

E .67

F

F

.73

.87

Figure 5. Surmise relation with problem-specific relative solution frequencies for all experimental conditions. Table 2. Average number of solved problems conditional on tackling and solving the associated exercise in the learning phase. See text for details. RATH

TBL

FF

6.94

7.74

HS

5.43

9.29

to review the complete course. The actually covered content is evident from the associated exercises, which had to be completed in the learning phase. The number of exercises tackled and solved correctly within the time limit in the learning phase was very similar in all conditions, ranging from 63.44% in condition TBL/FF to 70.63% in condition RATH/HS. Thus, it is reasonable to confine consideration to those problems, the associated exercises of which have been tackled and solved. Table 2 provides the resulting average values for this case. An analysis of variance shows that there is no main effect of the motivational state (F(1,100) < 0.001; p = 0.984), while the main effect of the learning condition turns out to be significant (F(1,100) = 6.129; p = 0.015). There is no significant interaction between the factors (F(1,100) = 2.623; p = 0.108).

4. Discussion While models of cognitive psychology significantly contributed to the development of adaptive tutorial systems in the past, emotional and motivational aspects were only rarely considered. A psychological model for a successful computerized adaptive tutorial system, however, has to cover all these aspects. The idea underlying the present paper is to build a starting point for integrating formal models of cognitive and emotional/motivational aspects into a comprehensive psychological model that can serve

J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning 175

this purpose. We outlined Knowledge Space Theory (KST) suggested by [10,11], a well-established cognitive model for computerized adaptive testing and teaching, and discussed the motivational model of Atkinson and McClelland as a candidate that may be used to extend KST in a proper way. Integrating latent emotional/motivational states, however, requires that they are compatible with the structural assumptions in KST. The research question whether we are able to represent the observed solution behaviour under different motivational states within the same knowledge space is addressed by the presented pilot study. It focuses on the motivational states hope for success (HS) and fear of failure (FF). For compatibility with previous investigations on motivational and emotional aspects, the experiment included different types of learning conditions, which are computer-based learning (RATH) and common text-based learning (TBL). The results of the pilot study show that under all conditions the vast majority of response patterns coincide with a knowledge state of the assumed knowledge space. This means that the knowledge space suggested by [37] offers a considerably good fitting representation of the response patterns, and thus of the knowledge in the considered domain. Statistical analysis of the data in Table 1 reveals that the goodness-of-fit does not differ over experimental conditions. There is neither a main effect, nor an interaction effect of the two factors motivational state and learning condition. This conclusion is also corroborated by the fact that the order of the problem-specific solution frequencies is in line with the underlying surmise relation under all conditions (see Fig. 5). The frequencies decrease with increasing “difficulty” of the respective problem class. This finding complies with the notion of the surmise relation. If solving problem a implies solving problem b then the number of students solving b has to be greater than or equal to the number of students solving a. The average number of solved problems (conditional on tackling and solving the associated exercise in the learning phase) exhibits no differences in the two motivational states. There is, however, an effect of the learning condition, but no interaction is present. This result is remarkable, because, although there are differences in the over-all level of performance, the assumed knowledge space was able to represent the obtained responses with the same high precision. We may call this property sample invariance of the knowledge space. It is an essential requirement for extending KST by introducing additional factors, like motivational states. In order to conclude, the results of the presented pilot study strongly encourage the research program to integrate cognitive and emotional/motivational aspects into a comprehensive psychological model for adaptive tutorial systems. As we have only treated two prototypical motivational states, however, future research should address other emotional and motivational aspects. These may also include factors like controllability, or causal attribution. A further step then should aim at implementing such a comprehensive psychological model into a tutoring system.

References [1] L. Cuban, Teachers and machines: The classroom use of technology since 1920. Teachers College Press, New York, 1986. [2] R.E. Mayer, Multimedia learning. University Press, Cambridge, 2001. [3] L.v. Rosenstiel, Wissensmanagement heißt Enteignung der Experten [Knowledge managment means dispossesion of experts]. Süddeutsche Zeitung (2001, December 12). [4] S. Wienecke & D. Kern, E-Learning – Die besten Anbieter [E-Learning – The best suppliers]. Personalwirtschaft, 12 (2001), 36–44. [5] D. Albert, C. Hockemeyer, & T. Mori, Memory, knowledge, and e-learning. In L.G. Nilsson & N. Ohta (eds.), Memory and Society. Routledge and Psychology Press, London, 2004.

176 J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning

[6] D. Albert & T. Mori, Contributions of Cognitive Psychology to the Future of e-Learning. Bulletin of the Graduate School of Education Hiroshima University, Part I (Learning and Curriculum Development), 50 (2001), 25–34. [7] D. Albert, E-learning Future – The Contribution of Psychology. (Keynote). In R. Roth, L. Lowenstein, & D. Trent (eds.), Catching the Future: Women and Men in Global Psychology – Proceedings of the 59th Annual Convention, International Council of Psychologists, July 8–12, 2001, Winchester, England (pp. 30–53). Pabst Science Publishers, Lengerich, 2001. [8] D. Albert, Contributions of the Psychology of Knowledge to Learning Science and Education. Journal of Learning and Curriculum Development, 2 (2003), 111–116. [9] D. Albert & C. Hockemeyer, State of the art in adaptive learning techniques. EASEL Consortium, D03 Requirements Specification, version 1.4 (2001), 24–40. [10] J.-P. Doignon & J.-C. Falmagne, Spaces for the assessment of knowledge. International Journal of Man-Machine Studies, 23 (1985), 175–196. [11] J.-P. Doignon & J.-C. Falmagne, Knowledge Spaces. Springer, Berlin, 1999. [12] D. Albert & J. Lukas (eds.), Knowledge Spaces: Theories, Empirical Research Applications. Lawrence Erlbaum Associates, Mahwah, NJ, 1999. [13] C. Hockemeyer, RATH – A Relational Adaptive Tutoring Hypertext WWW–Environment (Technical Report, No. 1997/3). Institut für Psychologie Karl–Franzens–Universität Graz, Austria, 1997. [14] C. Hockemeyer, T. Held, & D. Albert, RATH – A Relational Adaptive Tutoring Hypertext WWW– Environment Based on Knowledge Space Theory. In C. Alvegaard (Ed.), CALISCE’98: Proceedings of the Fourth International Conference on Computer Aided Learning in Science and Engineering (pp. 417–423). Göteborg, Sweden: Chalmers University of Technology, 1998. [15] D. Albert & C. Hockemeyer, Adaptive and Dynamic Hypertext Tutoring Systems Based on Knowledge Space Theory. In B. du Boulay & R. Mizoguchi (eds.), Artificial Intelligence in Education: Knowledge and Media in Learning Systems (Vol. 39, pp. 553–555). Amsterdam: IOS Press, 1997. [16] C.E. Dowling, C. Hockemeyer, & A.H. Ludwig, Adaptive assessment and training using the Neighbourhood of knowledge states. In C. Frasson, G. Gauthier, & A. Lesgold (eds.), Intelligent Tutoring Systems (Vol. 1086, pp. 578–586). Springer, Berlin, 1996. [17] A.J. Elliot & C.S. Dewck (eds.), Handbook of competence and motivation. Guilford, New York, in press. [18] M. Jerusalem, & R. Pekrun, (eds.), Emotion, Motivation und Leistung [Emotion, motivation, and achievment]. Hogrefe, Göttigen, 1999. [19] K. Schneider & H.-D. Schmalt, Motivation. 2nd ed. Kohlhammer, Stuttgart, 2000. [20] Y.-G Lin, W.J. McKeachie, & M. Naveh-Benjamin, Motivation and student’s cognitive structure. Chinese Journal of Psychology, 41 (1999), 121–130. [21] C.L. Hull, Principles of behavior. Appleton-Century-Crofts, New York, 1943. [22] K.W. Spence & J.T. Spence, The psychology of learning and motivation: Advances in research and theory. Academic Press, New York, 1969. [23] C.D. Spielberger, Theory and research on anxiety. In C.D. Spielberger (ed.), Anxiety and behavior (pp. 3–20). Academic Press, New York, 1966. [24] C.D. Spielberger, Manual for the State Trait Anxiety Inventory (Form Y). Consulting Psychologists Press, Palo Alto, 1983. [25] D. Albert, Anxiety and learning-performance. Archiv für Psychologie, 132 (1980), 139–163. [26] J.W. Atkinson, Motivational determinants of risk-taking behavior. Psychological Review, 64 (1957), 359–372. [27] J.W. Atkinson & D. Birch, The dynamics of action. Wiley, New York, 1970. [28] J.W. Atkinson & D. Birch, Motivation and achievement. Winston, Washington D.C., 1974. [29] D.C McClelland, J.W. Atkinson, R.A. Clark, & E.L. Lowell, The achievement motive. AppletonCentury, New York, 1953. [30] L.H. Chusmir & A. Azevedo, Motivation Needs of Sampled Fortune. 500 CEOs: Relations to Organization Outcomes, Perpetual and Motor Skills, 75 (1992), 595–612. [31] E.G. French, Some characteristics of achievement motivation. Journal of Experimental Psychology, 50 (1955), 232–236. [32] D.C. McClelland & C.E. Franz, Motivational and other sources of work accomplishments in mid-life: A longitudinal study. Journal of Personality, 60 (1992), 679–707. [33] B. Weiner, An attributional theory of motivation and emotion. Springer, New York, 1986. [34] P. O’Connor, J.W. Atkinson, & M. Horner, Motivational implications of ability grouping in school. In J.W. Atkinson & N.T. Feather (eds.), A theory of achievement motivation (pp. 231–248). Wiley, New York, 1966. [35] K.M. Sheldon & A.J. Elliot, Goal striving, need-satisfaction, and longitudinal well-being: The SelfConcordance Model. Journal of Personality and Social Psychology, 76 (1999), 482–497.

J. Heller et al. / Achievement Motivation, Performance Structure, and Adaptive Hypertext Learning 177

[36] H.-D. Schmalt, K. Sokolowski & T.A. Langens, Das Multi-Motiv-Gitter zur Erfassung von Anschluss, Leistung und Macht – MMG. Swets, Frankfurt, 2000. [37] T. Held, Establishment and empirical validation of problem structures based on domain specific skills and textual properties. Unpublished doctoral dissertation, Universität Heidelberg, Germany, 1993. [38] C. Hockemeyer, Tools and Utilities for Knowledge Spaces (Unpublished Technical Report). Institut für Psychologie Karl–Franzens–Universität Graz, Austria, 2001.

178

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

An Interactive Dictionary of Concepts: An Exploratory Platform for Enhancing Communication Between the Concepts Which Form and Inform Us Ania LIAN Critical Pedagogy and Technology Consultants, Australia

Abstract. In this chapter, we describe a non-commercial, web-based communication platform which we have called an Interactive Dictionary of Concepts. The platform was designed in order to facilitate and support a negotiation process between various interest groups and, as a result, between the concepts in which the interests of these groups are embedded. We find support for the development of such a platform in the concern expressed by Calhoun (2002) that increased communication between individuals and groups requires practical experiments based on models of interaction that seek to “improve[s] the quality of opinions, educate[s] the participants and form[s] a collective understanding of issues that advance[s] beyond pre-existing definitions of interests or identities”. Calhoun claims further that without such models, our use of IT might be reduced to “websites giv[ing] the impression of consisting simply of the spontaneous postings of the public” (Calhoun, op. cit., The idea of the public sphere section, para. 6). The Interactive Dictionary of Concepts is a management structure written in the LAMP (Linux-Apache-MySQL-PHP) protocol consisting of a set of forms used to enter information into a database and another set of forms for retrieving it. Retrieval is also supported through the use of search engines such as Google in combination with special identification headers. The Dictionary is still in a developmental stage. It responds to the challenge issued by Calhoun by offering a set of conceptual and software tools designed to overcome the interactional limitations of conventional course design (often implemented in management systems such as WebCT despite the wider potential offered by their innate structural flexibility). Typically, these courses limit communication possibilities by being structured around objectives which pre-determine the relevance of the interactions which they generate. On the other hand, in the Interactive Dictionary of Concepts, the objectives which motivate its design do not exert such a limiting function. On the contrary, the relevance of the interactions which they generate is a function of the impact that the entries in the Dictionary have on subsequent interactions which they inform. In other words, their relevance is in the communication which they help to establish between their own terms and those which inform other contexts. The structure of the Dictionary has no other purpose but to facilitate the flow between different terms or concepts. It does so with tools designed to prevent compartmentalisation of interests, thus enabling and facilitating a dialogue across different groups and beliefs. In turn, this allows individuals and groups to expand and, as a result, to reformulate the terms which form and inform their criteria of judgment. Unlike standard dictionaries, the Interactive Dictionary of Concepts does not come with ready-made entries. Instead, it enables individuals who are investigating various concepts and issues to create their own entries, which are then situated and organised within the constraints of the management structure of the Dictionary. These creative constraints themselves function as tools for stimulating and gener-

A. Lian / An Interactive Dictionary of Concepts

179

ating the critical reflection of authors upon the concepts which they are investigating. This is achieved by offering conditions which challenge, enrich and help to systematise the associations which inform the ideas of authors and their beliefs. Thus the structure of the Dictionary generates conditions which enable communication or dialogue to flow where it would otherwise have stopped when concepts reflect opinions, rather than being the product of a methodology which requires systematised reflection. The Dictionary can be thought of as a framework supporting and generating production (and investigation) of critically-informed entries on all kinds of issues of interest, each sourced in different perspectives and none exhausting them. Entries are organised according to a specific framework which is common to all entries. This shared framework then forms a structure which links the independently produced entries and relevant information together, thus effectively creating from these entries a single, large (potentially infinite) text produced by many, for many, and on issues whose relevance to others can emerge in serendipitous and unpredictable ways through use of the search facilities of the Dictionary. Thus the Dictionary has the potential to store a conceptual history of the ideas of individuals (or groups) from all over the world, systematised and made available in the form of searchable entries. In this perspective, the structure of the Dictionary was not conceptualised as a way of improving the affective and emotional aspects of human-computer interaction as such. Instead, we made it our priority to utilise the exploratory potential of its tools to enhance the quality of the interactions of humans, thus enhancing the process by which individuals and groups feel and become integrated with one another. In turn, this form of integration develops individuals’ and groups’ emotive links with each other by facilitating the process of conceptual and emotional advancement raised by Calhoun. Moreover, the framework of the Dictionary enables and supports varied forms of engagement which help to develop, in individuals, a heightened sense of the value of their own contributions to the community. Thus the potential of the Interactive Dictionary of Concepts is multiple. It functions as a platform which generates dialogue and, therefore, as a system which can assist individuals or groups in their task of investigating issues of concern. Further, the interactions which it generates and accumulates can support different projects, unpredictable in form, which share a concern for a dialogic mode of inquiry. The Interactive Dictionary of Concepts is also a management structure which provides an extremely flexible system for storing and retrieving its concepts and relevant information. Moreover, the search categories of the Dictionary function as tools in support of a critical analysis of texts, and is particularly useful in the process of reading and writing. Overall, the structure of the Interactive Dictionary of Concepts helps us to create an environment which is thoroughly organic, i.e. where different concepts and projects have the possibility of impacting upon, informing or triggering unpredictable dialogic contexts. In this chapter, we will discuss the intellectual framework behind the structure of the Dictionary, its potential educational applications and discuss the analytical and systematising capacity of its tools against practical examples of the entry-like texts created by students of Thai and English, and by the author, while experimenting with the tools and the concept of critical inquiry which informs them. Keywords. Information technology, the international public sphere, dialogic inquiry, critical inquiry, critical pedagogy, critical thinking

1. Introduction: The Significance of the Interactional Dictionary of Concepts Interest in constructing an Interactive Dictionary of Concepts was born out of the Thai News Network (TNN) project conducted at Khon Kaen University, Thailand, with first year undergraduate students of Thai. The overall objective of the course which adopted the model of the TNN was to contribute to the enhancing of students’ critical reading

180

A. Lian / An Interactive Dictionary of Concepts

and writing skills (Buranapatana & Lian 2002).1 The TNN project issued a challenge to students to establish their own news channel with goals and means corresponding to the kinds of ways in which they would see it appropriate to affect others and hence effect change. Thus the guiding intention behind the project was to facilitate and support an educational platform where critical investigation implied a process where students would not only engage in the criticism of texts but, more importantly, would do so in order to enhance their interactions with others. This meant that the objective of their writing was not primarily to serve the disciplinary interests of the particular subject which used the TNN model. Instead, the aim was to engage students in interactions enabling them to integrate their academic work in the context of a diversity of interests of the community at large. The objective was to engage students in exploring issues in relation to those interests and, as a result, in formulating their own identities as belonging to the community and as working in the interest of the community. As the project showed, this form of integration helps to develop students’ emotive links with others, and hence a sense of the value of their contributions and, ultimately, of their own personal value: Now I read more. My parents are surprised that I have changed in this way in that I think more when I talk with them and when we are watching TV together. I can talk with them about political issues. Normally I used to like watching cartoons, not serious matters like political or academic issues. My mother asked me the other day ‘What has happened to you?’ ‘Are you O.K.?’ I told her that I participated in the experimental group, and that we were taught to think critically. I also told her that I have learned a lot as a result. (Student 5)2 Working on the TNN project was fun because I had a chance to express my ideas to others as much as I could. When writing on-line articles, I always thought of the readers and what they might say. (Student 4)

In the course of the project, students developed a web-based information channel while engaging each other, the faculty staff and students and the community at large, in discussions on issues that students identified as relevant and worthy of exploration. Within the three months of the project, they created thirteen articles reporting on different issues ranging from social, political to scientific subjects. They also advertised their own university to the community letting people know about the things that the university did for the students and for the Khon Kaen community. In order to gain the interest of the community for their channel, students advertised their website around the university, among other students, with teachers from different subjects, and in the community in general. They established contact with a wide range of people and obtained email feedback from hundreds of readers. They also established a discussion forum for their readers. In short, the project allowed students to approach the task of writing in a manner where success did not depend on teachers’ judgments alone, but on their ability to participate in, and generate, negotiation among members of the general public. In the process of news production, students reported learning a number of skills. These included acquiring the courage to talk to people from villages or from universities on specific issues of concern, developing a professional attitude to their own work, or learning to look for information. Students themselves learnt to understand the power of informed participation. Students expressed excitement that their work was no longer 1 2

See the TNN project’s site at: http://web.kku.ac.th/~tnnproject/. Students’ comments can be found on: http://www.anialian.com/Maliwan_students.html.

A. Lian / An Interactive Dictionary of Concepts

181

to be ‘an essay for the teacher to grade’, but was directed toward creating a resonance in those who would come to their website: I think presenting our work on the TNN website is an important way to encourage all the students in the experimental group to work harder, and produces benefits for readers inside and outside the university. Having lots of readers giving feedback to us was very useful for our project. (Student 1)

However, despite the enthusiasm that the TNN project generated, a comment from one of the project evaluators raised the problem of the visibility of the critical thinking process. The question is not new. However, the question acquired a new meaning as a result of the TNN project. The evaluations showed that making the critical process visible would result in the creation of a framework for organising communication between individuals and groups. The framework would constrain the authors to a very particular process of text production, thus prioritising the methodology of the critical approach which the framework adopts as opposed to focusing only on the content of texts. In short, other than functioning as a means for systematising texts, the framework would also challenge the authors to submit to its process of organising thought. These objectives do not imply that the framework would replace all forms of writing genres with its own. Of course not. But it would offer a new genre of writing, one which could stand on its own, or could function as a support for assisting the process of writing or analysis of texts. This versatility of the framework would open it up to a multitude of contexts in which it could prove to be of value. Below we provide examples of some of those contexts. (a) In its stand-alone form, the framework could prove to be useful in the following contexts: •

•

As an exercise structure for reflecting upon issues of concern. By focusing on the structure of the process of reflection, the authors are given the opportunity to take a step away from what they think is relevant and focus on the process which, in turn, may help them to detect what they would have otherwise ignored. These exercises in reflection would be stored on the web and, if retrieved, might be of use to others. As a form of academic writing to be used in scholarly journals. There are many advantages that this framework might offer for this type of text. One of them is that it allows for comparing and contrasting the arguments raised on the same issue by different scholars quickly and in an ordered fashion. The retrieval forms, which are structured according to this shared framework, would allow for the entire spectrum of concerns included by the authors to be displayed at once. The framework would generate a dynamic platform for everyone to explore, evaluate and enhance their understandings of specific areas of concern. Unlike search engines such as Google, a textbook, or most on-line resources, the framework of the Dictionary would allow for retrieving of entries which are unpredictable in range, but predictable in their capacity to open up connections and flows which, otherwise, are hidden away from us largely by our own prejudices which prevent us from looking elsewhere: This is the point that people who never come close to the fabrication of science have the greatest difficulty in grasping. They imagine that all scientific

182

A. Lian / An Interactive Dictionary of Concepts

articles are equal and arrayed in lines like soldiers, to be carefully inspected one by one. However, most papers are never read at all. No matter what a paper did to the former literature, if no one else does anything with it, then it is as if it never existed at all. You may have written a paper that settles a fierce controversy once and for all, but if readers ignore it, it cannot be turned into a fact; it simply cannot. (Latour 1987: 40)

•

As a tool for analysing texts. It is possible for individuals to take texts produced without the support of the framework and to analyse those texts according to the framework’s criteria. Thus the structuring function of the framework can be used to create texts about texts. There is no limitation on the kinds of texts that can be analysed. One interesting context is the media. It is very often said that the media provide us with uncritically researched news and information. It would be an interesting exercise to explore the critical quality of those texts with the use of the analytical tools offered by the framework.

(b) On the other hand, the critical tools of the framework could be used as a support for assisting the process of writing texts. This is an important function in projects such as TNN. It is not impossible to imagine that students from different schools, colleges and universities might create a uniting network, a World News Channel, for themselves and the general public to enjoy access to debates, news, educational programs, all developed and made available in order to increase communication between various interest groups. Thus while the producers of those programs, e.g. the students, may experiment with different forms of expression in order to enhance their interactions with others, the tools of the framework can help to develop a critical approach to the production of the students’ texts. Subsequently, students’ projects or news stories can be linked to specific exercises in the thinking which informed those stories and which were constructed with the help of the critical thinking tools of the framework. Consequently, the viewing public would gain access not only to the students’ (or others’) productions, but also to the ‘deeper structure’ of those views, i.e. the history of the concepts which students investigated and to the ways in which they worked with those concepts. A double structure of this kind, one expressing views and another revealing the work that contributed to their formation, is certainly an alternative to the currently opaque formats of news or information presentation. Further, it also demonstrates that no views come from nowhere. Therefore, it is not the views that generate break-downs in a dialogue. Rather it is the belief that we use the same concepts in the same way. Breaking down these concepts and identifying the contexts which make them relevant is a critical aspect of a process concerned with facilitating a dialogue which educates its participants (Calhoun, ibid.). The above reflections on the notion of creating a framework for making the critical thinking process visible convince us that there is value in such a model. In this chapter we present one such model whose development was first sponsored in 2004 by the

A. Lian / An Interactive Dictionary of Concepts

183

Faculty of Humanities and Social Sciences in Khon Kaen University and supported by its former Dean Associate Professor Dr. Sripanya Chaiyai. We have discussed the model with a number of colleagues in Thailand and Australia, experimented with its structure in both countries in different teaching subjects 3 and settled on the model whose conceptual underpinning we will describe next.

2. The Concept of Critical Reflection: Intellectual Background As we showed above, the advantages of making the critical thinking process visible become apparent when the task is approached as a problem of methodology of inquiry (or thinking), rather than as a problem of the perspectives from which a particular form of reasoning is evaluated. The confusion between the two approaches to the concept of critical thinking may well be at the root of the debate as to what counts as critical. Atkinson (1997) gives a number of examples which reflect this confusion: Likewise, Fox (1994) describes interviewing seven university professors who had extensive experience working with nonnative graduate students on their academic writing. Hoping to receive from them precise understandings of the notion of analysis and critical writing – terms that she equated directly with critical thinking and that the professors commonly used – Fox asked them to define these terms. She reports, however, that “this question was surprisingly difficult for them to answer, despite their confidence in using these terms in the language of their assignments” […] and despite the ease with which they were able to identify such characteristics as “good analysis” and “difficulties with analysis” in their students’ writings. (Atkinson, op. cit.: 75)

As a result, Atkinson decides to present the problem of critical thinking as a “tacitly learned behaviour” (Atkinson, op. cit.: 77) which, when taught explicitly, is likely to result in privileging valued behaviours (Atkinson, ibid.). This leads Atkinson to believe that teaching students to approach texts critically is a matter of socialising them to practices particular to specific “content domains” (Atkinson, op. cit.: 88). He justifies his view in Vygotskian learning theory which represents learning as a hierarchical process (Cheyne & Tarulli 1999). In other words, points of view or perspectives are not differences to be embraced, but to be overcome. In a hierarchical model of learning, therefore, the object of change is the “novice”, rather than the terms which validate beliefs. As a result, the relevance of the direction of this change is validated in the positive, rather than critical, evidence which the expert constructs about his/her own judgment. Thus, the hierarchical model of interaction adopts a methodology where the significance of judgments is established in relation to a single text, here the knowledge of the expert, and in isolation from other texts. The expert does this by refracting the logic of potential external determinants (e.g. the novices’ own criteria of judgment) in terms of his/her logic so that external determinants can have an effect only through transformations dictated by this logic (Johnson 1993: 14). As a result, the expert structures the 3 Among our collaborators who helped us with the project, each in a different way, were the faculty staff from the School of Humanities and Social Sciences, Khon Kaen University, Thailand; Ubon Ratchathani University, Thailand; Dhurakijpundit University (DPU), Thailand; the Rajabhat University in Nakhon Ratchasima, Thailand; lecturers in Thai at the University of Sydney, Australia and at the Australian National University, Australia; doctoral students in the fields of critical pedagogy and second language learning at the University of Canberra, Australia.

184

A. Lian / An Interactive Dictionary of Concepts

concept of success, and of the conditions facilitating success, monologically, i.e. in terms which the expert understands, thus limiting the possibilities engendered in the novices’ own histories to the boundaries of his/her constructs. Thus, the hierarchical model of interaction vests experts with the authority to distinguish between legitimate and illegitimate concerns. ‘Critical’ means submitting to this authority. On the other hand, the notion of critical thinking as a problem of methodology of inquiry, rather than of the source of the authority, approaches the concept of knowledge formation as involving negotiation, rather than imposition, of the criteria which validate the relevance of beliefs/constructs. In our framework of critical inquiry we build on and expand Luke’s (2004) understanding that being critical involves “call[ing] up for scrutiny […] the rules of exchange within a social field” (Luke, op. cit.: 26). For Luke, this process of scrutiny entails a form of “epistemological Othering and ‘doubling’ of the world” (Luke, ibid.), “some actual dissociation from one’s available explanatory texts and discourses, a denaturalisation and discomfort and ‘making the familiar strange’” (Luke, op. cit.: 26–27). However, in our view, the process of positioning oneself as ‘Other’ by generating ‘dissociation from one’s available explanatory texts and discourses’ is technically not possible. This is so because our available explanatory texts and discourses are all we have and are. Thus whatever the ‘Other’ that we might construct, it is always the product of the explanatory texts and discourses available to us. In other words, the ‘Other’ can be constructed only in relation to the associations available to us. The disruption of those associations would prevent us from constructing any positions, i.e. be it the ‘Other’ or the ‘Self’. However, if a critical approach is to mean generating different associations, Luke’s model does not give us the criteria to inform the direction of this process and, therefore, the means for evaluating its impact upon the inquiry. “[W]atching oneself watch oneself as an object of power and naming oneself as such” (Luke, op. cit.: 28), while undoubtedly a valuable concern, does not yet specify a methodology for ensuring that the process and its outcomes facilitate increasingly informed participation in or across the social fields. In our view, the model of interaction that Luke offers is at the extreme end of the spectrum of the apprenticeship model discussed by Atkinson. There the expert holds the criteria for distinguishing between that which is relevant and that which is not. On the other hand, in Luke’s model, the focus is on dissociating from the comfort which comes from rejecting possibilities other than those generated by one’s own criteria of judgment. While we wholeheartedly agree with this direction, we would like to propose a methodology of critical inquiry/thinking whose objective is not to relinquish our authority over our judgments, as seems to be the case in Luke, or to impose the authority of the expert, as in the apprenticeship model. Instead, it is to generate conditions enabling individuals (or groups) to assert their own authority over discursive situations through a process which supports their informed participation (Freadman 1994: 21). The advantage of the methodology that we propose is that it does not separate the context of investigation from the context of the participation of individuals in social fields. Instead, we propose a model of interaction where both contexts inform one another and whose main objective is to facilitate this communication. In fact, we suggest that an uncritical approach is characterised by a break in this communication. Thus it is already apparent that in an inquiry of this kind, the focus is not on generating questions as such, but on conditions enabling the negotiation of the criteria which validate the relevance of our beliefs or constructs. The aim of this negotiation process is to enhance or expand the perceptions which inform our understandings by enabling a flow be-

A. Lian / An Interactive Dictionary of Concepts

185

tween conflicting perspectives and the concerns that form them. Further, in order to facilitate such conditions, the model does not seek to disrupt the integrity of our logics or concepts, but to enhance it. Thus we would concur with Latour’s (2002) definition of the critic as “not the one who debunks, but the one who assembles” (Latour, op. cit., Conclusion section, para. 12). In light of the above, we would like to formulate the concept of critical inquiry as being a form of communication between conflicting perspectives (conflicting interpretive systems) where the objective is to explore the contributions that these perspectives offer on the object of concern.4 And, as in communication, these contributions are constructed and interpreted in relation to the strategic value that they offer to the negotiation process itself. In contradistinction to self-referential modes of validation, in a critical inquiry there is no arbitrary structure which validates its terms. There is no expert and there are no concepts limiting the intellectual scope of investigations. Instead, the validity of their terms is the object of strategically – (pragmatically) oriented struggles. The concern with validation being formed in the context of dialogue between conflicting perspectives gives the methodology of critical inquiry its theoretical, strategic (pragmatic), intellectual and ethical features. A critical inquiry is theoretical because it sets itself the objective of challenging and, as a result, of expanding, the conceptual bases by which it validates the relevance of its terms. These terms are the stories which form and inform our concepts. Thus, while we may use the same concepts (war, honour, love, time, life), what we mean by them is informed by stories which reflect our individual histories, not the history of the CONCEPT whose social basis can be realised only through individual use.5 A critical inquiry is also strategically or pragmatically oriented as decisions regarding the explanatory power of its constructs are established in relation to the conflicting perspectives upon which the inquiry draws and which it seeks to affect, not in relation to an arbitrary construct. The intellectual aspect of critical inquiry refers to the dialogic basis of its interrogations which effectively enrich the intellectual capital with which the inquiry can approach its current and subsequent interrogations. The ethical aspect of critical inquiry comes from the recognition that all our attempts to understand are ultimately sourced in our condition as humans and reflect interests which emerge from the history of the interactions which inform our condition. In this sense, a critical inquiry does not relativise everything to the detriment of all values. Instead, it is embedded in the objective of overcoming what prevents us from enhancing the conceptual bases which inform our interactions and, hence, our history conceptualised as that which forms “an integral and living part of our present world of semblance” (Becker 1931, Section III, para. 4). Together, all four aspects of critical inquiry form a methodological framework which acknowledges the sociohistorical basis of our concepts. In this perspective, concepts and perceptions are conceptualised as coming to us through time where time is defined as interactions which form and inform us and which, therefore, bring about change (cf. Hobson 1998: 25). The passing of time is marked by such conceptual (incl. affective) changes. Therefore, critical inquiry is always located in time (Calhoun 1995: 11). Its interpretive frameworks are an object of change embedded in interactions 4 “Our hypotheses, therefore, should not be accorded predictive value in relation to reality, but strategic value in relation to the question raised” (Lyotard 1984: 7). 5 “The paradox of communication is that it presupposes a common medium, but one which works […] only by eliciting and reviving singular, and therefore socially marked, experiences” (Bourdieu 1991: 39).

186

A. Lian / An Interactive Dictionary of Concepts

that seek to relate rival perspectives so as to construct the points from which positions, or possibilities, become more perceptible (Hobson, op. cit.: 24). Latour and Stenger (Latour 1999) refer to this process of relating or negotiation as involving investigators in seeking out contexts (conflicting positions) able to endanger their present beliefs. They see this act of endangering as a critical mechanism that enables an inquiry to qualify its conclusions against a multitude of other possible explanations. The result are richer articulations between concepts and the criteria of judgment: “rendering talkative what was until then mute” (Latour, op. cit., The Stengers-Despret falsification principle section, para. 1). Below we provide an extensive quote from Latour’s (2004) critique of the concept of peace-making and the kinds of prejudices which come into play when negotiators seek to invent peace proposals. In this critique he gives the example of the “Valladolid controversy” which illustrates how our conceptual frameworks, i.e. the stories by which we live and which inform our concepts, can surprise us if we allow them to become visible: […] the “Valladolid controversy,” [refers to] the famous disputatio that Spaniards held to decide whether or not Indians had souls susceptible of being saved. But while that debate was underway, the Indians were engaged in a no less important one, though conducted with very different theories in mind and very different experimental tools. Their task […] was not to decide if Spaniards had souls – that much seemed obvious – but rather if the conquistadors had bodies. The theory under which Amerindians were operating was that all entities [have] a soul […] For the controversialists at Valladolid, the opposite was the case but they remained blissfully unaware that there was an opposite side. Indians obviously had bodies like those of Europeans, but did they have the same spirit? Each side conducted an experiment, based on its own premises and procedures: on the one side to determine whether Indians have souls, and on the other side to determine whether Europeans have bodies. The Amerindians’ experiment was as scientific as the Europeans’. Conquistador prisoners were taken as guinea pigs and immersed in water to see, first, if they drowned and, second, if their flesh would eventually rot. […] If the conquerors drowned and rotted, then the question was settled; they had bodies. But if they did not drown and rot, then the conquerors had to be purely spiritual entities, perhaps similar to shamans. […] The relevance of this anecdote should be apparent: at no point in the Valladolid controversy did the protagonists consider, even in passing, that the confrontation of European Christians and Amerindian animists might be framed differently from the way in which Christian clerics understood it in the sixteenth century. […] Bartolomé de Las Casas, the Dominican priest, held that Europeans and Amerindians were basically the same, and he complained of the un-Christian cruelty of Christians against their “Indian brothers.” But how would he have responded, how might his views have modified, had he witnessed the systematic drowning of his fellow Spaniards in a scientific experiment designed to assay their exact degree of bodily presence? Which “side” would Las Casas, after the experience, be on? (Latour, op. cit., Blessed are the peacemakers section, para. 3)

The important feature of the critical inquiry framework presented above, and which underpins the structure of the Interactive Dictionary of Concepts, is the creation of a platform engendering negotiations capable of revealing the surprises which hide within the concepts by which we live. The presence of features that we ourselves find surprising means that we are often not aware of the contexts, or stories, which inform our concepts and which give them sense. In the next sections we will illustrate the relationship between the model of critical thinking/inquiry described above and the descriptors of the Interactive Dictionary. These descriptors have a double function. First, they systematise information in the

A. Lian / An Interactive Dictionary of Concepts

187

Dictionary, thus offering search facilities which provide the interrogators with the entire context for each entry (or inquiry) which contains the requested information. Second, they function as ‘creative constraints’ which structure the critical investigation of individual authors as they go about creating their own entries in the Dictionary. The constraints achieve this by offering conditions which challenge, enrich and help to systematise the associations which inform the ideas and beliefs of authors. We will discuss the methodology of this process against examples of experimental texts created by students and by the author while testing our model of critical inquiry and the structure of the Dictionary.

3. The Interactive Dictionary and Its Descriptors The Interactive Dictionary is a communication platform informed by a model of interaction whose objective is to enhance the collision of perspectives and, as a result, the flow between different concepts and terms. Hence the objective of the Dictionary is to function as an environment where interests and beliefs crisscross and, as a result, impact upon one another in the most unpredictable ways. Consequently, unlike standard dictionaries, the Interactive Dictionary of Concepts does not come with ready-made entries. Instead, it is a structure, a shell, which enables individuals who are in the process of investigating various concepts and issues to create their own entries. These entries are then situated and organised within the constraints of the structure of the Dictionary. Entries are then stored in a database and can be retrieved at later time. The descriptors of the Dictionary give it a structure which enables the following functions: (a) Individuals (or groups) can create their own entries. They do so while using the descriptors to systematise, and open up for further review, the implications of the dialogic interactions which the descriptors makes possible. (b) Each entry is stored and can be searched according to the information contained in the descriptors or inside the text of those entries. (c) The Dictionary stores and retrieves information about the entries. However, as Appendix 2 illustrates, the retrieval forms contain links to information provided by authors and stored on different websites. This increases the interactivity of the Dictionary. (d) The process of constructing entries is interactive as it challenges authors to engage in dialogue with the Dictionary which is designed to help them to systematise the associations which inform their ideas and beliefs. In other words, given that our interpretations are always internally constructed, our dialogic contexts always involve associations which are created by us. Generating their internal dialogue helps to enrich the articulation between them and to enhance their integrity as well. The structure of the Dictionary attempts to enable this process. (e) The process of searching the entries allows for retrieval of concepts used in contexts where they are least expected. Thus the process of retrieving the entries is interactive as, like the process of creating one’s own entries, it can challenge the associations which different concepts generate in us and, as a result, can generate a dialogue between unexpected connections. Thus this dialogue changes us and transforms the conceptual bases with which we approach subsequent interactions.

188

A. Lian / An Interactive Dictionary of Concepts

(f) Despite its apparently rigid structure for writing and retrieving entries, we have discovered through our experimental work that there is no one way of understanding the descriptors. People will approach them differently. We have learnt that the advantage of the Dictionary is not in that it somehow imposes a single way of approaching questions and, as a result, guarantees the ‘clarity’ of the entries. Instead, the advantage is in the flow which its dialogic interactions generate and support between the diversity of ‘contact points’ or concepts which they help to elicit. (g) The objective of the descriptors is to generate those unexpected connections and flows without there being any other agenda. The Dictionary, therefore, is an excellent tool for supporting a dialogue between the differences which divide us while, at the same time, providing an interactional platform for inspecting the contributions which these difference offer to the contexts of our own beliefs and actions. While we do see the use of the Dictionary in interdisciplinary projects, we would prefer to talk about its potential in stimulating interactions across interest groups, thus avoiding the trap of narrowing the nature of those groups to some predetermined features.

4. Experimental Texts and the Functions of The Dictionary’s Descriptors In order to demonstrate the structuring facilities of the Dictionary, we will use three different types of texts, all informed by the concept of critical inquiry described above, but each produced under different circumstances. The texts are: 1.

2.

3.

Is it enough for doctors to have a high IQ? (Appendix 1). The text is translated from Thai and was produced by the students6 from the original TNN project in 2002. We use this text to illustrate and compare its structure against the descriptors of the Dictionary. The text was produced prior to the conceptualisation of the descriptors and was informed only by the concept of critical inquiry as identified above. Violence in the South of Thailand. 7 The text was produced by students 8 of English from Dhurakijpundit University (DPU), Thailand, in 2004.9 The students who produced this text did not create entries in the Dictionary but used the descriptors to inform their task of constructing a text. The concept of dialogue (Appendix 2).10 A draft version of a Dictionary entry produced by Lian, A., 2004.

The descriptors of the Dictionary organise its structure, and hence the process of assisting with the construction and search of its entries, into the following categories: Topic, Keywords, Author, Affiliation, Languages, Formulating the Question for Analysis, Identifying the Concept for Investigation, Organising Information, Fresh Perspectives and New Questions. An example of the retrieval form for Dictionary entries is 6 Katanchalee Pranom, Raritta Phinichkit, Chin-anong Prachumchit, Wanida Ngamlert, Sutthanee Chaiyasuan & Jittra Kaewloyma. 7 Full text available from: http://www.geocities.com/orissa_panatte/. 8 Sukitpat Puengtamsujarit, Boonaek Thawornsirisakul & Orissa Panatte. 9 The work of the students from the Dhurakijpundit University, Thailand, can be found on the following site: http://geocities.com/dpuwriters/index.html. 10 Also available from: http://criticalpedagogy.com/cptc/html/resources/cptc_resou_maint10.php?res_id=43.

A. Lian / An Interactive Dictionary of Concepts

189

included in Appendix 2. The first set of descriptors includes information about the entry itself (the topic, keywords, author, affiliation, languages) and an example of this data entered under these descriptors can be found in Appendix 2. These descriptors concern the following information:11 •

The topic: The category reflects the general key concept or the topic of the entry. (E.g. dialogue, national anthem, culture, corruption, money, religion.) • Keywords: The category includes as many descriptors as possible which, according to the author beliefs, best describe the entry which they are constructing and the contexts to which their investigation may be relevant. (E.g. religion, science, politics, etc.) • Author: The name of the author of the entry. • Affiliation: The affiliation of the author(s). • Languages: The languages in which the resource is written and into which it is translated. The critical aspect of constructing information begins with the descriptors which follow below. 4.1. Formulating the Question As we said earlier, the objective of the Dictionary is to provide a means for helping to open up our concepts to interactions able to endanger the terms that validate them and, as a result, for expanding the conceptual bases (the interpretive systems) with which we approach our current and subsequent interactions. In order to facilitate this model of endangering through dialogue, the first step that we identified as crucial is that of formulating the question of concern to the author of the entry. Formulating the question is critical because the question opens up the field for different perspectives to occupy different strategic positions in relation to the concept which the question identifies as the object of negotiation. Thus the dialogic interactions which will ensue will not simply attempt to ‘answer’ the question. They will reflect upon the contributions that different perspectives (i.e. the products of different dialogic contexts) offer to that object of negotiation The outcome of this process is not an answer as such, but an expansion of the terms which impact upon the understanding of the object of negotiation and, ultimately, its conceptual change. As a result, the question can then be reformulated and give rise to new negotiations which will now be constructed on a more informed basis. As we signalled earlier, as these negotiations occur, our concepts acquire more history and, consequently, their impact upon our explanatory frameworks becomes more perceptible. Now, we will illustrate the methodology of the process of formulating a question. The objective behind the task of ‘formulating question’ is to focus the inquiry on very specific object of concern. The steps of this process go from a very general aspect of concern, i.e. the topic, to a very detailed question which specifies the concern of the author. The example below comes from one of the questions raised by the students in the TNN project: Topic / issue: E.g. Topic / issue: 11

Refers to a general statement about the object of concern: ‘What do you want to investigate?’ condom vending machines

There is no obligation on the authors to fill all the slots.

190

A. Lian / An Interactive Dictionary of Concepts

Initial question:

Here authors are asked to turn their concern into a question which will reveal more information on the exact source of the concern. E.g. Initial question: Should we have condom vending machines in Thailand? Relevance: E.g. Relevance:

Hypothesis: E.g. Hypothesis:

Examined question:

Examined question:

Here we simply challenge the authors to question themselves about the purpose of asking their question. Why ask about condom vending machines in Thailand? We request from authors to identify their own reasons for asking the question which they are posing. The presence of condom vending machines in Thailand may encourage immorality particularly in young people. We are now at the stage when we can specify the exact parameters of the question. We do so by combing the information which we obtained from the ‘Initial question’ and the ‘Hypothesis’. Does the presence of condom vending machines in Thailand encourage immorality particularly in young people? (In this question we combined the theme from the Initial Question: ‘the presence of condom vending machines in Thailand’ with the concern expressed in the Hypothesis: ‘the presence of condom vending machines as a potential factor encouraging immorality particularly in young people.)

It is important to emphasise that the step of ‘formulating question’ focuses on the process, thus making the content an object of this process. We will now turn to the students’ experimental texts and the ‘Dialogue’ entry in the dictionary to see how these texts define their object of focus, i.e. their question. We will then use these questions to explore whether the authors of the experimental texts investigated what they set out to examine, or whether they lost the focus and went about answering a different question altogether. Text 1: Is it enough for doctors to have a high IQ? (Appendix 1). Topic / issue: Source:

high IQ The title

Initial question: Source:

Should we trust people with a high IQ score? “In general, people tend to trust and sometimes even give special status to those who are skilled.” (para.1)

Relevance:

Why ask about trusting people with a high IQ score?

A. Lian / An Interactive Dictionary of Concepts

Hypothesis: Source:

Examined question: Source:

191

Because high IQ alone may not guarantee one’s professional success. “But those of us who only have a high IQ cannot be guaranteed of success in work duties, because we have to work with lots of people; some will have ideas and feelings which are different from ours, and sometimes there may be conflicts.” (para. 1) Is a high IQ score sufficient for generating trust? “People in control of their emotions may have more success in creating positive relationships with other people. EQ, or Emotional Quotient. Some believe that EQ is a possible indicator of personality and is related to emotional control.” (para. 1)

Our analysis: We had no problem in identifying the above steps in the students’ text. Text 2: Violence in the South of Thailand.12 Topic / issue: Source:

Violence in the South of Thailand The title

Initial question: Source:

Why is there violence in the South of Thailand? “Recently there has been a major upsurge in the terrorist activities in the south. Despite the content attention the issue has received, we still don’t know why this has occurred in our ‘Peaceful’ country.” (para. 1)

Relevance:

Why ask about the violence in the South of Thailand?

Hypothesis:

Because recently there have been terrorist attacks in the South of Thailand and despite the attention, we still do not know why? as above

Source: Examined question:

Source:

How is it that despite the attention given, we still do not know the reasons behind the recent terrorist attacks in the South of Thailand? as above

Our analysis: As in Text 1, the first paragraph summarises quite well the focus of the students’ article. Text 3: The concept of dialogue. (Appendix 2). Topic / issue: Initial question: 12

Dialogue What is a dialogue?

When transcribing the text, we corrected students’ typing errors.

192

A. Lian / An Interactive Dictionary of Concepts

Relevance: Hypothesis:

Examined question:

Why ask about the concept of dialogue? It is important to study the concept of dialogue, because at the heart of many, if not all, problems may be our failure to engage in a dialogue and, therefore to understand not only others, but also ourselves. What makes dialogue a method for including or taking account of others?

Our analysis: The above task of formulating a question was not inferred by us from the content of an article. Instead the text reflects the steps taken by the author to prepare her investigation of the concept of ‘dialogue’. The steps follow in agreement with the principles of the methodology of ‘formulating question’. The exemplified structure brings together the concerns raised in the ‘Initial question’ (‘What is dialogue?’) and the ‘Hypothesis’ (dialogue as a context enabling to engage and understanding others, and hence oneself). The question about the features of the dialogue focuses the investigation on studying the characteristics which allow dialogue to engage and enhance understandings. To summarise, the point of this section was to demonstrate the first step of critical analysis with the help of the tools of the Dictionary. In our experience, the task of formulating the question is not easy. Our experimental work showed that the descriptors function as new concepts and that people take time to adjust to the methodology which they represent. 4.2. Formulating the Concept of the Inquiry In the next step, we turn ‘The examined question’ into a concept which forms the object of negotiation. The task of formulating the concept of the inquiry is relatively straightforward. It involves turning ‘The examined question’ of the inquiry into a concept whose terms (that which gives sense to the concept) can then be investigated from the perspective of a variety of interests. We will describe this process of investigation in the sections below. First we will illustrate the technique of turning a question into a concept: Example: Condom vending machines Examined question: Does the presence of condom vending machines in Thailand encourage immorality particularly in young people? Concept analysed: The presence of condom vending machines as a possible factor encouraging immorality particularly in young people. Now, we will transform the questions of the experimental texts into concepts: Text 1: Is it enough for doctors to have a high IQ? (Appendix 1). Examined question: Concept analysed:

Is a high IQ score sufficient for generating trust? The adequacy of high IQ scores to generate trust.

A. Lian / An Interactive Dictionary of Concepts

193

Text 2: Violence in the South of Thailand. Examined question:

Concept analysed:

How is it that despite the attention given, we still do not know the reasons behind the recent terrorist attacks in the South of Thailand? Making sense of the recent terrorist attacks in the South of Thailand.

Text 3: The concept of dialogue. (Appendix 2). Examined question: Concept analysed:

What makes dialogue a method for including or taking account of others? Features that make dialogue a method for including or taking account of others.

4.3. Organising the Information In order to facilitate the crisscrossing and collisions between the interests which impact upon, and form, our interpretive systems, we have created this stage of analysis. Here, authors are required to identify the contexts in which different interests attribute different qualities to the concept under investigation. It is the interactions between these different qualities that are the focus of the next phase of the inquiry. In order to identify and to systematise the contexts which assist in revealing the different interests, the descriptors of the Dictionary help to distinguish between (a) the contexts of use of the particular concept and (b) the bibliographical information describing its source and providing a link, wherever possible. When outlining the contexts of use, it is important that authors do not express their opinions on those contexts. At this stage, authors are asked simply to outline the examples (provide a list) of different ways in which the concept is approached and interpreted in view of different interests. Outlining examples is different from providing bibliographical annotations and links. As Appendix 2 shows, the same bibliographical source may contain examples of different perspectives in which a given concept is used. It follows that when showing these examples, authors cannot simply offer a link to an article. They are requested to give information relevant only to a specific context of use. This may involve providing quotations which are preceded with short annotations giving a description of the context. The more contexts of use are identified, the greater the range of interactions which they facilitate. This way of organising information in the Dictionary allows for the perspectives which authors considered to be illustrated. Also, the Dictionary will show when authors happened to consult a limited range of interests. The range of interests which authors consult is important, because it helps to increase the dialogic interactions of the investigation at hand. These interests become the criteria against which the impact of the dialogic interactions which they generate is evaluated. Little or no conflict will imply little or no interactions. We will now look at the experimental texts and illustrate with their help the concept of ‘the context of use’. When seeking to identify such contexts, it is critical to connect the concept of the inquiry (the ‘Concept analysed’) with those contexts the ‘Contexts of Use’). We do this in our examples of the experimental texts. As we main-

194

A. Lian / An Interactive Dictionary of Concepts

tain this link, we are able to see whether the contexts exemplified in those texts were linked to those concepts, and, therefore, whether they provided perspectives on those concepts. We will limit our discussion to the students’ texts only. As for information regarding the ‘Dialogue’ entry in the dictionary, we refer readers to Appendix 2. It needs to be mentioned that the example in Appendix 2 is not perfect. The greater differentiation of perspectives that the ‘Contexts of Use’ descriptor can show, the more informative the entry would be. Text 1: Is it enough for doctors to have a high IQ? (Appendix 1). Concept analysed: The adequacy of high IQ scores to generate trust. Context:

Reference: Source:

Context: Reference: Source:

Context: Reference: Source:

The adequacy of high IQ scores to generate trust and the problem of stress generated by differences between people. own experience “But those of us who only have a high IQ cannot be guaranteed of success in work duties, because we have to work with lots of people; some will have ideas and feelings which are different from ours, and sometimes there may be conflicts.” (para. 1) The adequacy of high IQ scores to generate trust and the concept of EQ work in psychology (no bibliographical note and no links provided) “People in control of their emotions may have more success in creating positive relationships with other people. EQ, or Emotional Quotient.” (para. 1) And “Psychologists generated a different definition of EQ, but in summary, EQ is the ability to control emotions and to respond to one’s own needs in an appropriate way, to know how to overcome tension and various obstacles so as to be able understand one’s own emotions and those of other people.” (para. 2) The adequacy of high IQ scores to generate trust and AQ, MQ. work in psychology (no bibliographical note and no links provided) “Apart from EQ, psychologists are certain that, in order to improve continuously and create a pleasant society, there is a need for a high AQ (Adversity Quotient) which represent endurance and survival skills, and a high MQ (Morality Quotient) which means having a value system and good behavior.” (para. 2)

A. Lian / An Interactive Dictionary of Concepts

Context: Reference: Source:

Context: Reference: Source:

Context:

Reference: Source:

195

The adequacy of high IQ scores to generate trust and crime perpetrated by doctors. media “From time to time, in Thailand, we get news about horrific acts perpetrated by doctors.” […] “Examples of such shocking acts are: the Nuanchawi case [a doctor’s wife killed some years ago], the Sayamon case [a doctor’s wife killed some years ago] who was murdered by her husband who was a doctor because he wanted to have a love affair with a new woman, the case of Seum Sakhornras [a doctor who killed his lover 2–3 years ago] who killed and cut up the body of his lover because of jealousy, and the most recent case, that of doctor Sorachart Sirichot [at the time this article was written] who killed and burnt his girlfriend, Doctor Phatthanan Chayawong.” (para. 5–6) The adequacy of high IQ scores to generate trust and the ten precepts for being a doctor http://www.yasiam.com/txt/med450929.html “1. Have sympathy for the patient; treat them equally 2. Do not be greedy for gain from any patient, 3. Do not be boastful about your knowledge so as to deceive the patient” […] (para. 9) The adequacy of high IQ scores to generate trust, EQ and ways of instilling appropriate ethical behaviour can be instilled at the early stage of life Dr. Terdsak Dejkhong, head of the Department of Psychiatry, Charoenkrung Pracharak Hospital “[…] especially EQ which can be instilled in the unborn child during pregnancy. A mother should be eventempered, and avoid stress or pressure.” (para. 11)

Our analysis: By now it has become clear that there is a mismatch in the text between the concept that the article poses (‘The adequacy of high IQ scores to generate trust’) and the contexts upon which the text draws. In the title, the students focus on the medical profession (‘Is it enough for doctors to have a high IQ’). Then in the introductory part of the article they broaden the focus to the relationship between high IQ scores and trust which creates the expectation that the article is raising the question of the adequacy of high IQ scores to generate trust. Then they bring in the EQ factor and limit the problem of the relationship between high IQ scores and trust to the medical profession only. This focus is unexpected and leaves the reader with the feeling that there is huge room for improvement in the way in which we educate our doctors while, at the same time, other professions (or contexts of life) which deal with people, their hopes and grief are let off the hook. Text 2: Violence in the South of Thailand. Concept analysed: Making sense of the recent terrorist attacks in the South of Thailand.

196

A. Lian / An Interactive Dictionary of Concepts

Context:

Reference: Source:

Context: Reference: Source:

Context:

Reference: Source:

Context: Reference: Source:

Context:

Reference: Source:

Making sense of the recent terrorist attacks in the South of Thailand and the attitude of the Prime Minister media (assumed) “Since the beginning the Prime minister seems to have been unsure why these problems are occurring.” (para. 2) Making sense of the recent terrorist attacks in the South of Thailand and public opinion. media (assumed) “Many people have concluded that surely it was about religion because the terrorists confessed that they do follow religious teachings.” (para. 2) Making sense of the recent terrorist attacks in the South of Thailand and the view of a local Muslim student. personal interview “Mr Pinitch, a Muslim student at DPU believes that it is not an issue of a guerrilla band who want to make problems in order to have their own rule.” (para. 2) Making sense of the recent terrorist attacks in the South of Thailand and the view of another student. personal interview (assumed) “Ms. Daranee, a Muslim girl at Ramkhamhang University, said many Muslim people in the south were proud to be Thai and were happy to live in Thailand but some groups of people wanted to destroy peace in the south because they wanted to seek power in their area.” (para. 2) Making sense of the recent terrorist attacks in the South of Thailand and the view of a person living in the South. personal interview (assumed) “Mr. Suhaimin Vani, a Muslim man who lives in Pattani said government officials used unlimited [unrestricted] power to intimidate both Muslims and Buddhists and that these government officials deserve to die.” (para. 3)

Our analysis: The examples above indicate that students sought to balance the confusion of the media with opinions obtained from interviewing the people who lived in the South or who are Muslim. While these interviews in themselves are invaluable, the article as a whole shows a mismatch between the expectations which its title generates, and which are supported by the concept that we identified as the focus of the text’s investigation, and the actual question which the article answers, but does not pose. In our view, the fact that the article focused mainly on the interviews indicates that it did not seek to ‘make sense of the recent terrorist attacks in the South of Thailand’.

A. Lian / An Interactive Dictionary of Concepts

197

Instead, it sought to show opinions, largely from the Muslim strata of Thai society, on the conflict. The objective of ‘making sense of the recent terrorist attacks in the South of Thailand’ is not the same as ‘showing the public opinion on the attacks’. The former cannot be reduced to the latter, and public opinion, or interviewing people, may not provide sufficient diversity of contexts or perspectives to generate interactions able to impact upon the question that the inquiry set out to investigate. To sum up, giving the impression that the article will make an attempt at creating some sense of the events in the South is a claim (“we still do not know why this has occurred in our ‘Peaceful’ country”) which creates false expectations in the reader: we simply do not know when we read the article where it is going to take us. Thus the path that it takes and its endpoints are a surprise. 4.4. Fresh Perspectives and New Questions The ‘Fresh perspectives’ phase deals with the process of interaction between the different perspectives or dimensions of the examined concept elicited in the preceding stage of analysis. In this phase, the aim is to generate a dialogue between (epistemologically) varied interests and the concerns that shape them. To this end, it is necessary to construct contexts where these different interests can collide and where the impact of these collisions upon those perspectives and, therefore, upon our explanatory frameworks can be explored. Thus the task for this stage is to generate such contexts and to offer tools for evaluating the relevance of these dialogic collisions in relation to the concern expressed in the question that the text (or a dictionary entry) posed initially. The descriptors of the Dictionary achieve this by challenging authors with questions which are designed to help to systematise the varied interests in play by evaluating the contributions which they make to the object of concern. The questions are: • • •

• • • •

What dimensions or aspects of the examined concept appear to be shared between the different ‘contexts of use’ that you identified? Exemplify. What dimensions or aspects of the examined concept appear to be an object of conflict in the different ‘contexts of use’ that you identified? Exemplify. Can you suggest a way of grouping the differences and similarities by identifying the functions, i.e. the actual meaning of the goals or interests, that they serve? You may discover that differences and similarities stem from different meanings being attributed to what first appears to be the same goals. What kind of picture emerges regarding the question that you posed in your resource? How did this analysis help you to expand your understanding of the question that you posed in your entry? Did this investigation help you to suggest new questions that should be raised for further exploration? Did this investigation help you to suggest a new question that should be raised for further exploration?

Earlier in this chapter, we defined conceptual expansion as the objective of critical inquiry. Thus the final test for a critical inquiry is its capacity to broaden the perspectives which generated the initial concern and to reformulate, or transform, this initial concern into a new question, now constructed on a more informed (i.e. expanded) basis. It is these progressive reformulations of the initial concern/question which illustrate the

198

A. Lian / An Interactive Dictionary of Concepts

impact that the dialogic interactions have on the concepts of the inquiry. Moreover, new questions as new concerns are an indication of the potential of an inquiry to engage, impact upon and be affected by, more interactions. The relevance of an inquiry lies precisely in its capacity to be affected by earlier interactions (i.e. concepts which are the product of such interactions) and to affect subsequent interactions by generating new contexts for dialogue in the form of new questions, or concerns. Keeping in mind these objectives, we will now analyse the experimental texts from the perspective of their capacity to offer a more informed basis for new questions to be asked. Text 1: Is it enough for doctors to have a high IQ? (Appendix 1). The examined question: Is a high IQ score sufficient for generating trust? Our analysis: As we stated earlier, in our view, there was an inconsistency between the questions which we believed that the text posed in its introductory phase and the rest of the text. As a result, we cannot say that the interactions which the text generated informed the question. However, if we were to assume that the text sought to inform the question posed in its title, we would argue that the students’ article generated the foundations for helping us to redefine the question which it asked and, as a result, transform its original conditions on the basis of the information and the interactions which it generated. Thus we would have to conclude that, in view of the concerns raised by the authors, the problem is not so much whether ‘it is enough for doctors to have a high IQ’, but rather whether our medical institutions are prepared to meet the challenges generated by the need to ensure ethical conduct of medical staff. In other words, our view is that the students’ essay opened up the discussion regarding the concept of ethical conduct as it relates to the kinds of behaviours which are expected from the medical profession. Text 2: Violence in the South of Thailand. Examined question: How is it that despite the attention given, we still do not know the reasons behind the recent terrorist attacks in the South of Thailand? Our analysis: As in the case of Text 1, we believe that there is a dysjunction between the expectations which the text generates in its introduction and the path that it subsequently follows. Thus, we cannot say that the interactions in Text 2 were directed toward informing the question implied in its introduction. However, having said that, if we assume that the role of the text was to illustrate opinions about the conflict experienced by people who live in the South, or who are Muslim, we accept that the interviews demonstrated that there is a dissonance between the ‘official’ view, which presents the South as a source of religious conflict, and the views presented by the interviewees who emphasised the willingness on the part of the people from the South to be seen as Thai and for the need to replace violence with political solutions to the problem. Thus the interviews compiled by students helped to demonstrate that the problem in the South cannot be reduced to geographical or religious divisions. In this perspective, new investigations would need to take this conclusion on board and construct an inquiry which would help to further identify the true nature of the dissonance. We cannot resist quoting Latour again on the concept of peace:

A. Lian / An Interactive Dictionary of Concepts

199

On the other hand, peace proposals make sense only if the real extent of the conflicts they are supposed to settle is understood. (Latour, ibid.)

Text 3: The concept of dialogue. (Appendix 2). The examined question: What makes dialogue a method for including or taking account of others? Our analysis: Appendix 2 shows the exact steps of the analyses which can then be followed further online.13 As the Appendix shows, the contexts which the inquiry identified as using the concept of dialogue are relatively broad, ranging from dialogue as a means for reconciliation between adversarial parties, dialogue as an educational environment, dialogue as a means for resisting ideology and fundamentalism, through to dialogue as a method of scientific inquiry. But the contexts also revealed features of the concept of dialogue which are more and less consistent with one another, and which related exactly to the concept of inclusion and accounting for others. Thus the contexts identified by the inquiry helped to throw more light on the issue which generated the inquiry in the first place. As we show in the Appendix, new questions that the Dictionary entry entitled “Dialogue’ posed the problem of inclusion and accounting for others as signifying expansion, change and transformation, i.e. qualities which the discussion then suggests should be followed up by the proponents of Vygotsky’s model of dialogue.

5. Conclusion Our presentation of the Interactive Dictionary of Concepts began with the ambition to respond to the challenge issued by Calhoun and to construct a model of interaction capable of increasing communication between varied interest groups and, as a result, to contribute to a better understanding of each other. Against the background of our critique of the apprenticeship model and while building upon Luke’s concept of critical approach, we proposed a framework for a critical model of inquiry seen as communication and, therefore, as necessarily embedded in features which characterise communication. We identified these features as referring to the theoretical (concept-based), strategic (pragmatic), intellectual (incl. emotive) and ethical aspects of communication. We defined the interactions which form and inform critical inquiry as a history-generating process, embedded in the interactions (concepts) which precede it and as constructing the conceptual bases for subsequent interactions. We identified the interactive aspect of the Dictionary in its capacity to support and facilitate interactions which explore the ‘past’ in order to bring about change in their ‘future’. It is this interactive aspect of the Dictionary which, in our view, fulfils Calhoun’s requirement for a model of interaction capable of generating educational change in people as members of a society with differences sourced in their respective histories, histories which need to be embraced as the building blocks of the links which we create between us. We argued that this process takes the form of progressive steps involving the reformulation or transformation of those links. We illustrated the analytical capacities of the descriptors of the Dictionary against examples of experimental texts created by students and by the author. Our analysis showed that despite some structural prob13

The Dictionary entry ‘Dialogue’ can also be found on: http://anialian.com/Dialogue.htm.

200

A. Lian / An Interactive Dictionary of Concepts

lems which the analytical tools of the Dictionary helped to bring out, the value of the reflections which these texts illustrated lay somewhere else. Indeed, this value resided in the possibilities that the dialogic interactions in those texts made available for asking new questions and for approaching problems on a more informed basis. It is in this spirit of taking small steps but in the right direction that we designed the Interactive Dictionary of Concepts and introduced it to the readers.

Appendices Appendix 1. Is It Enough for Doctors to Have a High IQ? Is it enough for doctors to have a high IQ? by Katanchalee Pranom, Raritta Phinichkit, Chin-anong Prachumchit, Wanida Ngamlert, Sutthanee Chaiyasuan & Jittra Kaewloyma. (translated from Thai) In general, people tend to trust and sometimes even give special status to those who are skilled. In today society, people constantly compete against each other. There is a common belief in society that smart people will be successful. By smart people we mean people with a high IQ (Intelligence Quotient), which is a measure of intellectual ability. If you have a high IQ, you will be an intelligent person, you will think insightfully. But those of us who only have a high IQ cannot be guaranteed of success in work duties, because we have to work with lots of people; some will have ideas and feelings which are different from ours, and sometimes there may be conflicts. People in control of their emotions may have more success in creating positive relationships with other people. EQ, or Emotional Quotient. Some believe that EQ is a possible indicator of personality and is related to emotional control. Psychologists generated a different definition of EQ, but in summary, EQ is the ability to control emotions and to respond to one’s own needs in an appropriate way, to know how to overcome tension and various obstacles so as to be able understand one’s own emotions and those of other people. People who have high EQ will know generosity, have endurance and purpose in their various activities, have empathy with others, will compete with themselves, will not compare themselves with others, and will try to develop their own skills to the maximum. Apart from EQ, psychologists are certain that, in order to improve continuously and create a pleasant society, there is a need for a high AQ (Adversity Quotient) which represent endurance and survival skills, and a high MQ (Morality Quotient) which means having a value system and good behavior. Those who lack EQ will lack self-control and the ability to think. Thus they are unable to display endurance or give proper consideration to value systems. They also have low AQ and MQ which are components of EQ. This ability to control emotions is important for all who need to socialize, especially in professions which deal or treat people all day, especially, those which deal with sick people or people who hope to receive care. Such a profession is the medical profession. It is accepted by all that people who are to be doctors must have a high level of intelligence, and from annual university entrance examination scores the medical

A. Lian / An Interactive Dictionary of Concepts

201

faculties of all universities are filled with people with high to very high scores. However, no one can be sure that people high IQs also have high EQs. From time to time, in Thailand, we get news about horrific acts perpetrated by doctors. Even though the number of such acts in medical profession is very small compared with other professions, the cases generate shock because of the expectation that people who choose the medical profession should have higher moral values and qualifications than other people. This is particularly shocking in cases where doctors are criminals, and, in particular, in cases where doctors are unable to control their emotions. Examples of such shocking acts are: the Nuanchawi case [a doctor’s wife killed some years ago], the Sayamon case [a doctor’s wife killed some years ago] who was murdered by her husband who was a doctor because he wanted to have a love affair with a new woman, the case of Seum Sakhornras [a doctor who killed his lover 2–3 years ago] who killed and cut up the body of his lover because of jealousy, and the most recent case, that of doctor Sorachart Sirichot [at the time this article was written] who killed and burnt his girlfriend, Doctor Phatthanan Chayawong. All these cases are about love, desire, lust and emotions, and there are still many cases which are about violence, bad and immoral conduct of doctors. For example, the case of a doctor who harmed sick people, the case of a doctor who raped and behaved indecently with a young girl who applied for a job. This behavior would not have arisen at all, if people controlled their emotions in a good way. At least they should have stopped and reflected on the appropriateness of their behaviour. At the conclusion of the murder case of Doctor Phatthanan, Doctor Sorachart, the criminal, said that, even though his girlfriend had died, he still loved her and wanted to look after her family. That shows that he committed the murder, because of sudden emotion, but that he regretted it. What is the value of regret after the act? Had he been able to exercise control over his emotions from the start of the affair, it would not have ended up in this way. In view of the shocking crimes caused by doctors from time to time, what will guarantee that a patient can trust his/her life to a doctor? All medical students have to study medical ethics as well as medical subjects. All doctors know very well that there are ten precepts for being a doctor. They are: 1. 2. 3. 4. 5.

6. 7. 8. 9. 10.

Have sympathy for the patient; treat them equally. Do not be greedy for gain from any patient. Do not be boastful about your knowledge so as to deceive the patient. Do not be envious of other doctors who have better knowledge. Do not seek the four prejudices, which are: the prejudice of love or desire, the prejudice caused by hatred, the prejudice caused by delusion or stupidity, and the prejudice caused by fear. Do not feel afraid of things that are the ‘Eight World Dharma’. Be ashamed of vice. Do not be lazy or careless. Be delicate in your dealings with others; justify your reasons for action. Do not put yourself at risk of ruin.

Source of information: http://www.yasiam.com/txt/med450929.html. But various horrible cases about some doctors who lack control of their emotions become known from time to time. Do doctors not pay attention to medical eth-

202

A. Lian / An Interactive Dictionary of Concepts

ics and think of it as only a subject of study that once they get a passing grade everything is finished? It is difficult to live by these ten precepts, because doctors are just ordinary human-beings who cannot always behave in a principled way. However, these ten precepts can be used to develop EQ, AQ, and MQ. For instance, let us consider the fifth item: ‘Do not seek the four prejudices’ is obviously adapted from Dharma (and Eight World Dharma) whose concepts some doctors do not even know about. This precept should be known not only by doctors, but also people in general. There are many factors which result in each person having different moral and emotional skills. Dr. Terdsak Dejkhong, head of the Department of Psychiatry, Charoenkrung Pracharak Hospital, said that having appropriate ethical behaviour can be instilled at the early stage of life, especially EQ which can be instilled in the unborn child during pregnancy. A mother should be even-tempered, and avoid stress or pressure. When it is time for a child to learn, the parents or guardians should teach them to learn about his/her own emotions including ways of expressing their feelings in appropriate ways. Apart from this, our environment is another factor which could affect our emotional skills as, everyday, we are involved with people who have different characters, minds, and habits, so that our EQs are constantly challenged. For doctors who work very hard and have to face unpleasant circumstance from the time that they are medical students to the time that they leave the profession, it is quite possible that this could negatively affect their EQs. After each shocking case, there is a view among Thai people that the process of selecting students for medical training should be reviewed. Apart from formal examinations which can predict the levels of IQ, it is suggested that different kinds of measurements should also be used. At the present time, there is just an interview and a personality test to select applicants in each institution. But the question remains: is that enough? Dr. Wallop Thainua, Permanent Secretary for Public Health, proposes a policy for students whose grades are good enough at the end of school to study in a faculty of medicine in order to become doctors or teachers in this field. In addition to good grades, admission will depend on good EQ scores. A second source of medical students will be those who have bachelor degrees in the sciences and who have work experience. In the workplace, these students have learnt to adapt in order to be able to work with their students or colleagues. This can be seen as evidence that they have had the opportunity to manage their EQ. There are many EQ tests which can be used [as a test], some of which can indicate whether a person shows signs of mental illness. Dr. Terdsak Dejkhong said that no EQ test has been found so far which works unequivocally and, therefore, there is no EQ test as yet which is acceptable to all. To sum up, should there be an EQ test before a person is granted a licence to practise medicine? Should we repeat this test subsequently? The answer to this question is something everybody would like to know so that they might feel comfortable about consulting a doctor. It may be that we all have unrealistically high expectations of doctors and believe that they should have higher levels of moral, ethics and EQ than ordinary people. We do not realize that doctors are just human beings who have flesh and blood, and feelings. They all can love, be greedy, be angry and be deluded. It is not only doctors that should think about their EQs, but also everyone as we all need to develop our own EQs so that terrible things will not keep happening. Before it is too late.

A. Lian / An Interactive Dictionary of Concepts

Appendix 2. An Example Retrieval Form from the Dictionary Title: The concept of dialogue, Ania Lian, 2004.

203

204

A. Lian / An Interactive Dictionary of Concepts

A. Lian / An Interactive Dictionary of Concepts

205

206

A. Lian / An Interactive Dictionary of Concepts

References [1] Atkinson, D. (1997). A critical approach to critical thinking in TESOL. TESOL Quarterly, 31, 1, 71–94. [2] Becker, C. (1931). Everyman his own historian. Annual address of the president of the American Historical Association, delivered at Minneapolis. December 29, 1931. American Historical Review, 37, 2, 221–236. Retrieved November 22, 2004 from: http://www.historians.org/info/AHA_History/ clbecker.htm. [3] Bourdieu, P. (1991). Language and symbolic power. Harvard University Press, Cambridge, Massachusetts. [4] Buranapatana, M. & Lian, A.B. (2002). Thai News Network: Critical thinking in a Thai reading programs. A paper presented within the Interdisciplinary Research Seminars, University of Canberra, Australia. Retrieved December 2, 2003 from: http://www.anialian.com/TNN_project.html. [5] Calhoun, C. (2002). Information Technology and the international public sphere. Paper presented to the International Sociological Association, Brisbane, Australia. Retrieved December, 2, 2003 from: http://www.ssrc.org/programs/calhoun/publications/infotechandpublicsphere.pdf. [6] Calhoun, C. (1995). Critical social theory. Blackwell. [7] Cheyne, J.A. & Tarulli, D. (1999). Dialogue, difference, and the “Third Voice” in the Zone of Proximal Development. Theory and Psychology, 9, 5–28. Retrieved November, 22, 2004 from: http://www. arts.uwaterloo.ca/%7Eacheyne/ZPD.html. [8] Freadman, A. (1994). Models of genre for language teaching. The 1994 Sonia Marks Memorial Lecture. University of Sydney. [9] Hobson, M. (1998). Jacques Derrida: Opening lines. Routledge. [10] Johnson, R. (1993). Editor’s introduction. In: P. Bourdieu. The field of cultural production: essays on art and literature, 1–25. Cambridge, Polity Press. [11] Latour, B. (2004). Whose cosmos, which cosmopolitics? Comments on the peace terms of Ulrich Beck. Common knowledge, 10, 3, 450–462. Retrieved May, 8, 2005 from: http://www.ensmp.fr/PagePerso/ CSI/Bruno_Latour.html/articles/article/92-BECK-CK.html. [12] Latour, B. (2002). Why has critique run out of steam? From matters of fact to matters of concern. Retrieved November 22, 2004 from: http://www.ensmp.fr/PagePerso/CSI/Bruno_Latour.html/articles/ article/089.html. [13] Latour, B. (1999). How to talk about the body. The normative dimension of science studies. A paper written for a symposium organized by Akrich and Berg, Paris, Theorizing the Body. Revised January 2000, November 2002. Retrieved November 22, 2004 from: http://www.ensmp.fr/~latour/articles/ article/077.html. [14] Latour, B. (1987). Science in action: how to follow scientists and engineers through society. Harvard University Press: Cambridge, MA. [15] Luke, A. (2004). Two take son critical. In: B. Norton & K. Toohey (eds.) Critical pedagogy and language learning, 21–29. Cambridge: Cambridge University Press.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

207

Human-Computer Interaction: Sharing of Intergenerational Wisdom and Cross-Cultural Knowledge Elspeth McKAY Ph.D. School of Business Information Technology, RMIT University, Australia [email protected]

Abstract. For some, the quest for new knowledge is an inherent pastime. Previously, much of what humans learnt was handed down by previous generations. The tools by which this knowledge transfer took place limited to a human-tohuman context; where speech and drawings provided the transmission of ideas. However, the implementation of electro-communication tools through the internet is bringing about a global anthropological revolution in human knowledge transfer. Web-based technologies have promoted the concept of learning partnerships and lifelong learning [1], with interest in globalised learning communities gaining momentum [2]. Although there is an expectation that lifelong learning will extend, in a natural sense, into the broader community, thereby promoting learning societies, the current tools are not effectively reaching key sectors of the community such as the special education arena [3]. Furthermore, little is known about the interactive effect of cognitive development and Web-mediated multimedia instruction on performance outcomes in the education and training sectors [4]. This Chapter will discuss a project underway linking Japan and Australia in an intergenerational cross-cultural study. The project is designed to investigate attitudinal changes amongst young children, undergraduate students, and senior citizens, in both synchronous and asynchronous learning environments when engaged in Webmediated collaborative knowledge sharing activities. Keywords. Collaborative learning, cross-cultural communication, e-learning, e-museum

1. Background There can be no doubt that the renewed interest in human-computer interaction has opened the way for a plethora of technological innovations designed to entice people to use the internet. Consequently, game playing on the internet has become a popular pastime particularly enjoyed by the young members of our community. However, in the rush to implement the most popular game, it appears that the system developers have paid little attention to the repercussions of the game-playing environments on offer. Many of the games involve chase scenarios or aggressive fighting that employ a curious combination of repetitious eye-hand coordination tasks that keep the user locked to the keyboard and screen for hours. The time spent playing these games is raising alarm bells in some quarters. Thankfully the social consequences of computer overuse are now being documented. Hopefully we can nip some of the negative effects of this overuse in the bud.

208

E. McKay / Human-Computer Interaction

It is not surprising to see that interest in the psychological effects of spending long hours in front of a computer screen is emerging from within social science. Problems associated with addictive disorders have been identified, with the internet cited as the cause. We should be taking a long hard look at the effects of how damaging it must be for young children who spend long hours playing with computerized play stations [4]. It may only be a matter of time before there is a voice rising from researchers calling for health warnings to be printed on the packaging of computerized multimedia toys. However, it may prove difficult to claw back the beckoning temptation in these visually pleasing entertainment platforms. At best we need strategies to be put in place for improving the ergonomic problems of incorrect physical conditions (it is so easy to slouch over a computer keyboard). Nevertheless the more difficult problems associated with addictive behaviours may take a little longer to solve. Wholistic courseware design will improve the position. Immediate issues that need attention include: the delivery mechanisms (ergonomics), and affective and emotional consequences of playing with interactive multimedia.

2. Human Computer Interaction The community interested in the human-dimension of human-computer interaction has an important role to play in setting much needed safety standards. The concept of human-computer interaction is not new. The Association for Computing Machinery (ACM) was established in 1947 (see http://www.acm.org/). Not surprisingly, much of the developmental work for computer systems thus far, has concentrated on the mechanistic side of the phenomenon. However an increased acceptance for internet game playing has brought forward renewed vigour for human-computer interaction. Describing human-computer interaction is complex; there are many seemingly competing and interactive variables. People who build computerized systems have struggled for many years to define these interactive relationships. Figure 1 attempts to distil the plethora of components that formulate the human-computer interaction context. 2.1. Technology Dimension Looking at the right-hand side of Fig. 1 we see the familiar variables associated with computer system’s design; like the commonly used graphics to denote files, hardware and peripherals. It is through the use of these symbols that the language of system’s design is often best conveyed. The added value of Web-mediated learning support systems has shown there is an immense attraction towards implementing information and communications technologies (ICT) to enhance our learning potential. Some believe this rather new approach to increasing our knowledge development is now achievable through rich internet applications (RIA). However the instructional/learning qualities that are highly interactive, immersive and known as constructionistic have been ignored for several years for technical reasons [5]. Fortunately the newer systems’ developmental tools like Java, Flash, Dreamweaver, VRML and X3D that have entered the Web courseware design field, reestablish the essential criteria for bringing about more effective learning environments. The current ICTs provide a range of powerful tools that include: easier access, updating capability, scheduling of tasks, and flexible environments for both learning facilitators (teachers, corporate trainers) and students. There are three ICT elements that

E. McKay / Human-Computer Interaction

209

Figure 1. Human Computer Interaction (adapted from (Preece 1994):16).

represent an RIA: rich client technology, server technology, and development tools. Rich client technology; the Flash player is a good example of an effective tool to provide all the benefits of the Web by keeping costs to a minimum (automatic compression and loading of components on demand), etc. Server technology provides the markup languages to connect to the rich client technologies; for example Web database language tools. Development tools provide an environment that provides the ability to create the various pieces of an application – from user interfaces to server-side logic. The technologists say that RIAs offer the benefits of distributed, server-based internet applications with the rich interface and interaction capabilities for desktop applications [5]. Moreover, they promote the notions that RIAs should possess the functionality to interact with and manipulate data, rather than simply visualize or present it. Enhanced leverage for a richer human-computer interface (HCI) capability providing realtime status indicators whenever background processing requires the user to wait is bringing the techno-development closer to the human-dimension of human-computer interaction. 2.2. Human Dimension Language and communication form important relationships within the humandimension of the Preece (1994) model (Fig. 1). Therefore it is no surprise to see that linguists are amongst the early computerized game-playing innovators. It will be largely through their efforts that the interactive effect of affective and emotional aspects of human-computer interaction will be investigated. This is where the cognitive psychologists and application developers can harmonize their system’s development

210

E. McKay / Human-Computer Interaction

work. One example of this was conceptualized by Dr. Yuri Nishihori, Professor at the Institute of Language and Culture Studies, Graduate School of International Media and Communication, Hokkaido University, Sapporo, Japan. Dr. Nishihori’s Japanese students translate English sentences sent from other countries. This work relies on an attractive interactive language interface to motivate the students. To complement this earlier work, a global eMuseum system (GEMS) proof-of-concept study has been underway since January 2003, at RMIT University, Melbourne, Australia collaboration with Hokkaido University, Sapporo, Japan.

3. Project Aim The primary aim of the GEMS project is to investigate the socio-cultural effects of global Web-mediated collaborative experiential learning environments that involve intergenerational and cross-cultural knowledge sharing. The literature reveals there are no research studies that combine the full range of these internet user community members. Researchers do however deal with the age-dimension by taking separate views of the population. These categories are the most common: the young (early childhood development), youth (adolescents) and the elderly. While there are many fine research projects for the younger end of the age scale, there are limited works on the upper end of the dimension. It has been noted that access to educational technology should be through a ubiquitous design respecting the values and requirements of the broadest end-user community possible [6]. It was felt that the time was ripe to tackle the issue of intergenerational knowledge sharing on the internet. Generally speaking, the boundaries of the age-dimension appear to be less apparent amongst internet users. With careful planning there are many exciting ways in which we can bring people together using the internet that involve us directly with the affective and emotional aspects of our electronic communications. For example, there are many sites available where a young child can sing nursery rhymes with their parents and others.

4. System Overview The GEMS prototype drawings and specifications have been finalized, with building of a system prototype to be completed by late 2005. A caricature of an actual museum building represents the initial HCI. This metaphor is steeped in historical meaning. Museums represent a place where the affective and emotional aspects of our lives are documented. Not only do they enable us to experience both the physical presence of treasured collections of historical artefacts; they enrich our lives through experiential awareness for things long gone. Those that visit museums can attest that when they visit, they feel a sense of reaching out across socio-cultural, intergenerational boundaries. A visitor’s entry point into the eMuseum of the GEMS project is through the front door that is used for viewing the Treasure Case of jewels. These Jewels represent games and traditional stories, fables, and legends. While inside the Treasure Case visitors can play with one of the Jewels; proceeding into a Playing Cave System if they wish to share their experience with others. It is

E. McKay / Human-Computer Interaction

211

Figure 2. GEMS Component Diagram [7]:180.

precisely in this type of experiential/knowledge sharing environment that the internet is bringing about tremendous changes to human-dimension of human-computer interaction. 4.1. Gallery Treasure Case This main storage area stores and maintains the Jewels collected from various countries, differing cultural persuasions and age groups. Drawing on the multimedia technologies, with this type of knowledge exchange associated with these games and the telling and retelling of traditional stories is unique. This free exchange promotes cross-cultural awareness in the community and allow for the important transfer of key human knowledge from generation to generation.

5. Activity Specifics 5.1. Playing Cave System This is a virtual set of rooms (caves) in which children can experience hands-on activities with the Jewels they enjoy playing with. They can either practice games on their own in a Practice Cave, or engage in collaborative activities with another child from a different country or region in the Playing Cave. The option to play interactively or practice in isolation may be beneficial in special education arenas, where skill and confidence levels may vary greatly. The Playing Cave is a key human-computer interaction component that enables children and visitors to enjoy the games and stories, while illustrating the regional differences of Jewels over the generations. It is proposed that through this type of Web-mediated knowledge exchange, the value of interactive col-

212

E. McKay / Human-Computer Interaction

laboration with other children throughout the world will be recognized by the participants. Upon entering an empty Playing Cave the GEMS visitor can look for other global playmates for interactive play. 5.2. Jewel Receiving Window If visitors would like to make donations to the eMuseum, they can enter through the Jewel Receiving Windows. Senior citizens can post any Crystal (game or traditional story) in their own language, or English if they like. Their name and the date of posting added to each Crystal, so they can see their donation in GEMS at any time. Children are also invited to post contemporary Crystals they play, in their own language, or using English. Each new Crystal is screened by GEMS Headquarters for its suitability in the eMuseum and categorised for cutting and polishing in the Jewel Cutting Factory. This donation process ensures that games and stories are passed from one generation to the next and from one culture to another, in an innovative information system designed to replicate the old word of mouth dissemination of stories. 5.3. Jewel Cutting Factory The Factory is staffed by English as second language (ESL) students in various countries, as part of their eLearning activities. Their role is to take the raw Crystals donated by visitors and cut them into Jewels for display in the eMuseum. These Jewel cutting activities are carried out in the collaborative bulletin board for students to exchange ideas among classes from various regions and cultures.

6. Significance and Innovation It is anticipated that spin-offs from this project will provide an understanding of new ways to unite members of the community in ways that are enjoyable, playful and encourage lifelong learning. There is a real chance to discover ways to engage a diverse cross-cultural user network, including elderly citizens and younger family members to benefit from the collaborative story telling activities, thereby strengthening family relationships. 6.1. Significance Australia faces significant socio-economic challenges this decade. This nation is one of the most multi-cultural countries in the world [8], weaving cultural diversity and associated tensions into the social fabric. Furthermore, Australia is an aging nation giving rise to communication problems associated with intergeneration discord [9]. Unfortunately, research is still ignoring the importance of socio-cultural interaction and Webmediated knowledge exchange [10]. There is an expectation afforded by the Web that facilitates the engagement of people in cognitive processes through collaborative team work [11]. Research therefore needs to deal with the complexity of the interactivity between humans and technology [12], and learning intelligence environments [13,14]. Work has commenced to investigate the ontological complexity in Web-based collaborative networks [15].

E. McKay / Human-Computer Interaction

213

The GEMS project addresses the gap between theory and practice [16], which exists within the Australian population, in terms of access to knowledge building techniques available through Web-mediated educational technology for: enhancing multicultural sharing, promoting knowledge transfer between generations, and facilitating quality outcomes in special education. These facets of socio-cultural interaction that work towards improving the affective and emotional well being of people will now be reviewed. (a) The issue of cross cultural sharing and the ever widening digital divide Ethnic and racial tensions are aggravated by social inequities. The media unwittingly feeds this dilemmama. Look at how often we are directed to the internet for further information. While exploring the internet may be easier for computer literates, the converse reveals a complete avoidance for this type of knowledge exchange. Misunderstandings that occur between cultural communities may be exacerbated by the digital divide through lack of access, for whatever reason, causing a gap in cultural differences, and henceforth leading to serious communication breakdowns. A discourse analysis may reveal how these breakdowns may be overcome in a Web-collaborative learning environment. According to Henry Tsang OAM, the success of cultural diversity depends on the spirit of sharing [8]. One common vehicle for cross cultural sharing is through the abundant courses available for learning a second language (including ESL). According to the second language acquisition (SLA) theories, oral interaction is necessary for enhancing inter-language development [17,18]. Since (electronic) chat communication is considered similar to oral interaction, the potential benefits of network-based communication for SLA and learning are often advocated by many researchers. Because text chat communication is considered to be an excellent tool to enhance abilities [19], the GEMS project explores and evaluates the effectiveness of text-chat as a means to bridge the digital divide, across cultural boundaries. Nevertheless, the difficulties in understanding the trigger negotiations of meaning amongst ethnic groupings must be overcome. While there is evidence in the literature of this type of research [20], it is generally concerned with the grammatical items, and although an inter-cultural communication gap is referred to, it is not explained. It is therefore valuable to focus on this inter-cultural communication gap to improve the affective side of human-computer interaction. (b) The issue of cross generational communications breakdown Difficulties that arise through intergenerational perspectives [21] can be witnessed in many parts of the community: at home with parents and siblings, in work-place reporting networks with age differences of employees, and at school between staff and students. This project is investigating the quality of the Web-mediated environment as a conduit, linking the vast knowledge/memories of elderly citizens with the inquisitive minds of the young. (c) The issue of accessing educational technology for special education Special education is defined here in a broad sense, as pedagogical practice especially designed as assistive technology for people with learning disabilities [22]. While a spe-

214

E. McKay / Human-Computer Interaction

cial education literature on computerized learning programmes has emerged in the last decade, mapping the collaborative interaction in terms of knowledge transfer has not been achieved anywhere in the world, much less than for special educational needs. This research is monitoring how children engage in cognitive and scientific knowledge development, drawing on human-computer interaction in a manner akin to the revolutions in conceptual change [23]. Consequently, for the first time people are watching living knowledge taking shape.

7. Benefits The multi-cultural aspects of this project innovates an exemplary commitment towards strengthening a cohesive community. The findings will provide the basis for benchmarking global human-computer interaction design from a socio-economic perspective, thereby informing future implementations of Web-mediated collaborative experiential learning programmes. Furthermore, the ripple effects of the technology and knowledge transfer will extend through international bodies such as UNESCO, allowing for crosscultural sharing on a global learning platform.

8. Conclusion Unless there is a concerted effort to interlace the human-dimension into each humancomputer interaction, the gap between theory and practice will continue to widen. Lost will be the understanding of the benefits that semiotic/discovery activities may elucidate for cross-cultural sharing. Lost too will be an important opportunity to discover how assistive technologies may benefit special education pedagogy. Systems’ designers are duty bound to promote equitable access to information technology to ensure each person has the choice to partake in the richness of 21st century communications technologies. This Chapter has described one way in which educational technologies and the internet can be synergized to capture Web-mediated collaborative interactivity. The GEMS research project extends the principles of a Generative Virtual Classroom [24], and Schank’s Sickle Cell Counselor [25], by combining social interaction through linguistics in a communicative collaborative model [26]. The project team manage the complexity of dealing with diverse subjects by adopting three distinct experiential environments. The first, involves senior citizens recalling traditional stories and games. The second defines undergraduate student interaction in a multi-cultural setting. While the third, relates to young children’s propensity for playful sharing of experiential learning materials. Through the interaction of these three environments the GEMS project achieves its initial purpose to provide a nexus for sharing cross-cultural knowledge, through intergenerational activities. In time it may well set the benchmark to reveal the effectiveness of the translation of the traditional games into a common language (English) for global use as an English language study tool for Foreign Language (EFL) in classroom activities and in the home environment. In this way, willing people such as university students can volunteer to support senior citizens whose memories are a magnificent resource of traditional stories and games [27], but whose expressive skills and computer literacy may not be good enough for this kind of activity.

E. McKay / Human-Computer Interaction

215

While there are Web-sites that purport to recognise and address racism in the learning environment [28], with others that offer games for children to play [29]; there are none that facilitate for interaction between the source program and multiple concurrent users. Moreover, in storing shared Web-mediated interaction, it is anticipated that the data gathered for the GEMS project will enable analysis of hands-on collections of stories and games in a user’s traditional context. The tracking of the collaborative nature of the involvement with the eMuseum will enhance the knowledge of interaction between generations, creating a capacity for emergent innovative global legends.

References [1] Geertshuis, S.A., Bristol, A., Holmes, M.E.A., Clancy, D.M. and Sambrook, S. 2000. Learning and business: Supporting lifelong learning and the knowledge worker through the design of quality learning systems, Kinshuk, C. Jesshope, and T. Okamoto, Editors, Paper presented at the International Workshop on Advanced Learning Technologies (IWALT 2000): Advanced Learning Technology: Design and Development Issues. Palmerston North, New Zealand: IEEE Computer Society. 186–187. [2] Boshier, R. 2000. Running to win: The contest between lifelong learning and lifelong education in Canada. New Zealand Journal of Adult Learning, 23(2): 6–28. [3] Langone, J., Clees, T.J., Reiber, L. and Matzko, M. 2003. The Future of Computer-Based Interactive Technology for Teaching Individuals with Moderate to Severe: Issues Relating to Research and Practice Disabilities, viewed 08/06/05, http://jset.Unlv.Edu/18.1t/langonet/first.Html. [4] McKay, E. 2005. Cognitive skill capabilities in web-based educational systems, in Interactive Multimedia in Educational Systems, R.C.Sharma and M.Sanjaya, Editors, Idea Group, Inc. UK, 213–248. [5] McKay, E., ed. 2005 forthcoming Special Edition Journal.(editorial). International journal for continuing engineering education and life-long learning: Special edition – the effectiveness of rich internet application for education and training, viewed on 08/06/05 at http://www.Inderscience.Com/browse/ index.Php?Journalcode=ijceell, ed. E. McKay and P. Kommers: UK. [6] Stephanidis, C. 2001. Towards universal access in the information society: Achievements and challenges, Paper presented at the Workshop on Universal Accessibility of Ubiquitous Computing: Providing for the Elderly. ICS-FORTH (HCI Laboratory) Internal Report. Viewed 08/06/05 http://www.ics. forth.gr/hci/files/PAPER_ON_UNIVERSAL_DESIGN_stephanidis_2001.pdf: Crete, 4. [7] McKay, E. and Nishihori, Y. 2004. Towards closing the digital divide: A multicultural, intergenerational ICT case study, Paper presented at the International Conference on Computers in Education – Acquiring and Constructing Knowledge Through Human-Computer Interaction: Creating new visions for the future of learning. E. McKay Editor, Melbourne, Australia: Common Ground Publishing. 179–189. [8] Tsang, H. 1995. Designing for diversity: The multicultural city, Paper presented at the 1995 Global Cultural Diversity Conference. Sydney: Department of Immigration and Multicultural and Indigenous Affairs: viewed 08/06/05 http://www.immi.gov.au/multicultural/_inc/publications/confer/10/ speech44a.htm. [9] Sternberg, J. 1997. Generating x: Lifestyle Panics and the New Generation Gap. Media International Australia, 85(November): 79–89. [10] Schank, R.C. 1990. Tell Me a Story: A New Look at Real and Artificial Memory. New York: Charles Scribner. [11] Kearsley, G. and Shneiderman, B. 1999. Engagement Theory: A Framework for Technology-Based Teaching and Learning, Naval Sea Systems Command: N00024-97-4173 viewed 08/06/05, http://home. sprynet.com/~gkearsley/engage.htm. [12] Sims, R. 2000. An interactive conundrum: Constructs of interactivity and learning theory. Australian Journal of Educational Technology, 16(1): 45–57. [13] Schank, R.C. and Morison, S.G. 1995. Tell me a story: Narrative and intelligence. Rethinking Theory: Northwestern University Press. [14] Garner, B.J. 2002. Role of solutions architects in learning intelligence, E. McKay, Editor, Paper presented at the Invited paper in eLearning Conference on Design and Development: International best practice to enhance corporate performance, held Oct 21–25. Melbourne, Australia: RMIT Informit Library. 18–25.

216

E. McKay / Human-Computer Interaction

[15] McKay, E., Garner, B.J. and Okamoto, T. 2002. Understanding the ontological requirements for collaborative web-based experiential learning, Kinshuk, et al., Editors, Paper presented at the International Conference on Computers in Education 2002. Auckland, NZ: IEEE Computer Society. 356–357. [16] Baine, D. 1982. Instructional Design for Special Education. New Jersey: Educational Technology Publications. [17] Pica, T. 1994. Research on negotiation: What does it reveal about second-language learning conditions, processes, and outcomes? Language Learning (44): 493–527. [18] Swain, M. 1993. The output hypothesis: Just speaking and writing aren’t enough. The Canadian Modern Language Review, 50(1): 158–164. [19] Nishihori, Y., Okabe, S., Yamamoto, Y. and Kurosaki, D. 2000. Chat’n’debate: Computer supported collaboration in language learning. Intelligent Systems and Applications (ISA 2000), 2: 677–683. [20] Kitsuregawa, M., Toyoda, M. and Pramudiono, I. 2002. Web community mining and web log mining: Commodity cluster based execution. Conferences in Research and Practice in Information Technology, 13th Australasian Database Conference (ADC2002), 5. [21] Thomas, T. 1998. Intergerational perspectives: Older persons through the eyes of the younger generations, Paper presented at the 33rd Annual Conference of the Australian Association of Gerontology. Melbourne, Australia. [22] Atkinson, D. and Walmsley, J. 2003. Time to make up your mind: Why choosing is difficult. British Journal of Learning Disabilities, 31(1, March): 3–17. [23] Thargard, P. 1992. Conceptual Revolutions. NJ: Princeton University Press. 285. [24] Schaverien, L. 2000. Towards research based designing for understanding fundamental concepts: The case of the web delivered generative virtual classroom for teacher education. Australian Journal of Educational Technology, 16(1): 1–12. [25] Bell, B., Bareiss, R. and Beckwith, R. 1994. Sickle cell counselor: A prototype goal-based scenario. The Journal of the Learning Sciences. [26] Cecez-Kecmanovic, D. and Webb, C. 2000. Towards a communicative model of collaborative webmediated learning. Australian Journal of Educational Technology, 16(1): 73–85. [27] Nishihori, Y., Okabe, S., Yamamoto, Y. and Kurosaki, D. 2001. Communicative and collaborative language learning with knowledge media (IntelligentPad), C.-H. Lee, et al., Editors, Paper presented at the Proceedings of the 9th International Conference on Computers in Education: Enhancement of Quality Learning Through Information and Communication Technology (ICT). Seoul, Korea: Incheon National Unviersity of Education. 1579–1584. [28] Anonymous. Conference of Education Systems Chief Executive Officer 2000. Racism. No way! Viewed 17/02/04 http://www.Racismnoway.Com.Au/. [29] Anonymous. Children’s Playrooms, viewed 08/06/05 http://homepages.ihug.com.au/~umm_pub/ GamesRm.html.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

217

Designing the Stimulation Aspect of Hedonic Quality – an Exploratory Study Michael BURMESTER and Annely DUFNER Stuttgart Media University (HdM), Germany

Abstract. It is widely discussed that products should provide more qualities than just providing the right functionality (being useful) and giving access to it (being easy to use). Hassenzahl [1] postulated that users perceive pragmatic and hedonic qualities when using an interactive product. The hedonic quality has a stimulation and an identity aspect. The central question of this study is which type of product features support the stimulation aspect of hedonic quality and whether this quality can be systematically increased in the design process. In a moderated workshop, experienced design experts were introduced to the theoretical model of perceived pragmatic and hedonic qualities and were asked to invent stimulating product features for a given online shop. The hedonic qualities of these feature ideas and of online shops showing more or less of these features were evaluated by shop users in an online survey. First conclusions for systematic design of the stimulation aspect of hedonic quality will be derived. Keywords. User interface design, emotional design, hedonic quality, pragmatic quality, design process

1. Introduction It is widely accepted in the human-computer interaction (HCI) community and in industry as well that products should provide the right functionality and be easy to use so that the users can accomplish their tasks and reach their goals. The required quality is usability as it is defined in ISO 9241 part 11 [2] as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use”. But it is increasingly also discussed, how products should be designed so that they serve more than being useful and easy to use. Emotional factors, experience factors and hedonic values are gaining more and more awareness [e.g. 3–5]. Some theoretical approaches were published [e.g. 1,6–9]. Some case studies [e.g. 10,11] as well as a few dedicated process approaches for the design of attractive products in a systematic way [e.g. 12] can be found. On the one hand, it could be argued that as long as the theoretical background is not completely clarified it is not the right time to talk about methods and process approaches. On the other hand, the requirement to design and develop attractive products is already relevant for industry [13]. Furthermore, there are products on the market already which have the power to fascinate people, such as mobile phones allowing extended and spontaneous communication, personalisation possibilities, gaming features, extraordinary design etc. The domain which is interesting for this paper is business-to-consumer. Here you can

218

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

find online shops providing advanced three-dimensional product views, personalised recommendation features, ‘cool designs’ etc. These applications were designed by product development and design teams. Obviously, the knowledge on how to design attractive products is already available. Very often, this is more tacit or implicit knowledge based on the personal education and project experience of design experts and not systematically documented knowledge [14]. In order to develop usable products, it is recommended to apply user-centred design, which internationally accepted to achieve usability in user interface design [15]. According to the usability definition quoted above, the design goal is to design products which are suitable for specific users and their tasks in a specific physical, organisational, social, and technical environment. The users must be able to achieve their goals with an appropriate amount of effort. If this is possible, the users should be satisfied, which means that they will not be harmed and will develop a positive attitude towards the product (as defined by ISO 9241 part 11 [2]. Design of products providing other qualities beyond usability has not been under consideration in this classic view of design in usability engineering [16]. In fact, a main goal is to prevent the user from negative emotions, frustration, anger, and disappointment [17]. The new approaches such as “emotional design” [9] or “funology” [5,18] focus on positive and motivating experiences.

2. Theoretical Model Hassenzahl and colleagues [e.g. 1,19,20] carried out several studies and derived a model of perceived qualities in human-computer interaction. The central conclusion of that model is that attractive products should not only cover usability aspects but should cover hedonic quality as well. Hedonic quality does not focus on task completion, but on activity-oriented aspects of human-computer interaction, i.e. users using a product in order to enjoy the product and to learn more about it. The model consists of the following concepts [1]: 1.

2.

Perceived Pragmatic Quality (PQ) People want to manipulate their environment. They need certain functions and access to them in order to achieve their goals in an effective and efficient way. This quality is very close to what is meant by usability. Perceived Hedonic Quality (HQ) The perceived hedonic quality consists of three sub-dimensions, which are stimulation, identity, and evocation. 2.1. Perceived Hedonic Quality – Stimulation (HQ-S): People want to learn. They want to grow and improve. They have a need for stimulation. Curiosity is the underlying motive of this quality. Interesting functions, new interaction techniques and innovative design can foster the stimulation aspect of hedonic quality. People like to try out and use a product beyond their tasks. 2.2. Perceived Hedonic Quality – Identity (HQ-I): A product must communicate something positive about the user to other people. This hedonic quality is called identity. For example, using an intelligent recommender system in online shops has the power to communicate that its user is a smart person using advanced technology.

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

Designer

Product

User

intended qualities

features

perceived qualities

pragmatic quality

presentation e.g. colour, forms, typography

pragmatic quality

e.g. controllability

interaction e.g. menus, sliders

hedonic quality e.g. exciting (stimulation) exclusive (identity) cosy (evocation)

content e.g. stories, explanations, descriptions functions e.g. search engine

e.g. controllability hedonic quality

appraisal

219

consequences behaviour

attractiveness e.g. beautiful

e.g. usage requency

emotion e.g. enjoyment

e.g. exciting (stimulation) exclusive (identity) cosy (evocation)

Figure 1. Theoretical model for the design of attractive products [based on 1,21].

3.

4.

2.3. Perceived Hedonic Quality – Evocation (HQ-E): The product is able to provoke positive memories. The product acts as a symbol, e.g. for a nice journey (e.g. souvenir). Attractiveness (ATT): Based on the perceived pragmatic and hedonic quality, the user forms an overall evaluation of the product. The product can be “good” or “bad” or “beautiful” or “ugly”. This is called “attractiveness”. Consequences: The attractiveness evaluation has consequences. First, the behaviour of the user is influenced. If the attractiveness of a product is low, then the users tend to use it not very often. Second, emotional reactions will occur. If the users judge the attractiveness very high, then they might be happy to use the product and be proud of owning it.

The theoretical model is shown in Fig. 1. A design team sets the design goals. On that basis, the designer develops the intended pragmatic and hedonic qualities of the product. The combination of the intended qualities forms the intended product character [1,3]. In order to achieve the intended product character on the basis of the combination of the intended product qualities, the designer has the possibility to manipulate the product features. Product features which can be manipulated are content and functions a product offers as well as presentation (e.g. colours used, layout, typography, etc.) and interactive elements (e.g. menus, buttons, links, sliders, etc.). The user experiences the product features when being exposed to it or when using it. Based on that, the user perceives the pragmatic and hedonic quality of the product. The perceived quality of the product can be far away from the intended quality. For example, designers intend to design a very easy-to-use product, but in usability tests it turns out that the product is not easy to use. Of course, the same applies to hedonic qualities as well.

3. Exploring Hedonic Features for Online Shops 3.1. Basic Idea Attractive products can already be found on the market e.g. websites providing fancy interaction possibilities as well as funny and interesting content [see 22]. Design teams

220

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

have invented and designed them. It can be suspected that expert designers should have some knowledge to create attractive products. This is very often based on implicit or tacit knowledge and skills of designers. Understanding their types of ideas for hedonic product features and their tacit knowledge can lead to a better understanding on how to design hedonic products. Based on this, systematic design methods can be developed. Assuming the general ability of expert designers to create attractive products, this ability can be fostered by focussing design experts on a certain intended quality of a product to be designed in the framework of a real design task. Concerning the model on product attractiveness introduced above, it was decided just to focus on the stimulation aspect of hedonic quality. The rationale behind this is to reduce complexity of the design task. Having asked design experts to generate design ideas, the question is whether users will appreciate them. Therefore, users were asked to assess these ideas. If these ideas support the stimulation aspect of hedonic quality, users should perceive products showing a certain amount of these ideas as more stimulating than products showing fewer of these ideas. 3.2. Goals This study is planned to be an exploratory study in order to better understand the design process of designing stimulating hedonic products. The following research questions should be answered: • • • • •

What type of product features will be invented if expert designers focus on the stimulation aspect of hedonic quality as the main design goal? Which of these features are rated as stimulating by users? Do products with a certain amount of stimulating features show more perceived hedonic quality than products having fewer features? What are methods or approaches supporting the creation of hedonic product features? What design principles can be derived from that exploratory study?

3.3. Research Method 3.3.1. General Approach The idea is to develop an understanding of the design process of hedonic product features. As described above, the hedonic quality consists of three sub-dimensions, which are stimulation, identity, and evocation. Each of these sub-dimensions is complex enough. In order to focus on the design work, we selected the stimulation aspect of hedonic quality as the main design goal. Online shops were selected as the domain of application because users of online shops approach these shops with different usage modes [22] resulting from the different phases of the buying process [23]. Users of online shops often begin with a diffuse need, e.g. finding a present for a friend. They want to be stimulated by the shop to find an interesting product. After they have decided that they want to buy a specific type of product, they have to evaluate the options (e.g. different brands). When the options are specified and the decision for a specific product is taken, then they have to transact. Users entering a shop can be in different phases of the buying process.

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

221

Experienced design experts were invited to a workshop. The designers had between 3 and 20 years of professional experience in the field of product and user interface design as well as website design. In the workshop, they had the task to invent features for a specific online shop in order to improve the stimulation aspect of the hedonic quality. The invented product features were evaluated in a user survey. Users of online shops had to rate whether the invented product features are interesting for them. Furthermore, they had to evaluate two online shops showing more or fewer features by filling out the online questionnaire AttrakDiff [24]. The AttrakDiff questionnaire measures perceived pragmatic quality, perceived hedonic quality stimulation, and perceived hedonic quality identity as well as the overall attractiveness. The questionnaire is designed as a semantic differential. The items and their affiliation to the dimensions are shown in Table 1. Table 1. Items of the AttrakDiff questionnaire (translated version of AttrakDiff [24], see also [25]). In this table, the items are ordered according to the four dimensions of the questionnaire: pragmatic quality (PQ), perceived hedonic quality stimulation (HQ-S), perceived hedonic quality identity (HQ-I), and attractiveness (ATT). The presentation order in the questionnaire is different from this. scale

original anchors

translated anchors

pragmatic quality PQ_1

technisch

–

menschlich

technical

–

human

PQ_2

kompliziert

–

einfach

complicated

–

simple

PQ_3

unpraktisch

–

praktisch

impractical

–

practical

PQ_4

umständlich

–

direkt

cumbersome

–

direct

PQ_5

unberechenbar

–

voraussagbar

unpredictable

–

predictable

PQ_6

verwirrend

–

übersichtlich

confusing

–

clear

PQ_7

widerspenstig

–

handhabbar

unruly

–

manageable

hedonic quality – stimulation HQS_1

konventionell

–

originell

typical

–

original

HQS_2

phantasielos

–

kreativ

standard

–

creative

HQS_3

vorsichtig

–

mutig

cautious

–

courageous

HQS_4

konservativ

–

innovativ

conservative

–

innovative

HQS_5

lahm

–

fesselnd

lame

–

exciting

HQS_6

harmlos

–

herausfordernd

easy

–

challenging

HQS_7

herkömmlich

–

neuartig

commonplace

–

new

hedonic quality – identification HQI_1

isolierend

–

verbindend

isolating

–

integrating

HQI_2

laienhaft

–

fachmännisch

amateurish

–

professional

HQI_3

stillos

–

stilvoll

gaudy

–

classy

HQI_4

minderwertig

–

wertvoll

cheap

–

valuable

222

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

Table 1. (Continued.) scale

original anchors

translated anchors

HQI_5

ausgrenzend

–

einbeziehend

non-inclusive

–

inclusive

HQI_6

trennt mich von Leuten

– bringt mich den Leuten näher

takes me distant from people

–

brings me closer to people

HQI_7

nicht vorzeigbar

–

vorzeigbar

unpresentable

–

presentable

evaluational constructs attractiveness ATT_1

angenehm

–

unangenehm

pleasant

–

unpleasant

ATT_2

hässlich

–

schön

ugly

–

attractive

ATT_3

sympathisch

–

unsympathisch

likeable

–

disagreeable

ATT_4

zurückweisend

–

einladend

rejecting

–

inviting

ATT_5

gut

–

schlecht

good

–

bad

ATT_6

abstoßend

–

anziehend

repelling

–

appealing

ATT_7

motivierend

–

entmutigend

motivating

–

discouraging

Note: Order and polarity of items was randomized. Table 2. Design experts. Profession

Background

Years of design experience

Usability Engineer

Computer scientist, usability engineer, experience in all methods of UCD, design experience with web-sites, consumer products, enterprise software

20 years of usability research and consultancy

Information architect

IT group manager, design of information architecture for large commercial web-sites

8 years of design experience

Industrial designer

Industrial designer, interface design for industrial products, web-sites, consumer electronics, corporate identity and corporate design

10 years of design experience

Communication designer

Communication design, experimental design projects, web-site design corporate identity and corporate design, advertisement design

3 years of design experience

3.3.2. The Workshop The central idea of the workshop was that experienced designers should be creative and work together in a moderated workshop situation. Table 2 shows the characteristics of the participants of the workshop. They have been recruited from a usability and user interface design consultancy, a design agency as well as a software and design company. The workshop was moderated by two experienced moderators with a background in usability engineering. Before running the workshop with the design experts, a pilot trial with students was carried out in order to fine-tune the workshop process. The workshop phases are described in Table 3. During the workshop, the participants were asked to analyse the online shop www.extra-tour.com as shown in Fig. 2a in

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

223

Table 3. Phases of the workshop. Phase

Instruction / Theme

Intended Result

1.

Introduction, presentation of participants, research objectives, explanations and agenda of the workshop

Being informed, warming up, set up a good working atmosphere

2.

Presentation of the theoretic model concerning pragmatic and hedonic quality as well as attractiveness. Examples of hedonic aspects in product design were shown. Furthermore, the focus on hedonic quality stimulation and the application domain online shops were introduced.

Common understanding that there are more qualities than usability and how they are perceived by the user. Focus on stimulation aspect of hedonic quality and on-line shops.

3.

The participants were asked to describe a personal fascinating experience with an online shop. Each participant was asked to explain and describe the experience. The moderators were asking questions until the experience was clearly described.

Introduction of the design challenge of hedonic qualities in on-line shops.

4.

Introduction of the relevant design possibilities by introducing the product features which can be designed: presentation, interaction, content, functionality (see Fig. 1).

Participants should be aware of all design options.

5.

Two participants worked together in order to evaluate an online shop (www.extra-tour.com). The participants were instructed to evaluate which stimulating features are implemented in the design of the on-line shop.

Focus on stimulation aspect of hedonic features based on presentation, interaction, content, and functionality.

6.

All results from the evaluation phase were collected on cards and pinned on a pin board. The participants could also present pragmatic aspects.

Overview of good hedonic and missing hedonic features.

7.

Creative Phase: The participants were instructed to write down ideas, how the hedonic quality stimulation of the on-line shop could be improved. Afterwards all participants presented their ideas. During the presentation phase all participants were invited to add more ideas. The moderators asked questions concerning understanding and rationale of the ideas.

Generation of as many as possible ideas for the improvement of the stimulation aspect of hedonic quality

8.

A second online shop (www.globetrotter.de) was presented which was selected by the moderators as more hedonic as the first online shop. One moderator presented that shop and the participants were asked to comment whether the suspected hedonic features are designed and how they can be improved.

Detailed evaluation and criteria for improvement of hedonic features.

9.

The participants are invited to write down criteria and methods which have the power to design hedonic products. Afterwards all participants presented their ideas.

Collection of design criteria and methods for the design of hedonic products.

10.

Conclusion and goodbye

Feedback from the participants concerning the workshop and positive end of the workshop.

order to generate ideas to improve the hedonic quality. The second shop www.globetrotter.de was assessed to be a more hedonic in beforehand by the moderators. This shop was introduced in the workshop in order to discuss the suspected hedonic features in this shop.

224

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

Table 4. Characteristics of survey participants. Age

20 to 40 years, average 27 years

Gender

9 female, 11 male

Education

Secondary school (1), high school (5), university degree (14)

Profession

media designer (2), educationalist (2), student (9), manager (1), economist (3), software developer (1), librarian (1), researcher (1)

3.3.3. User Survey The user survey was conducted online. The survey participants were typical and experienced online shop users. Beginners in online shop usage were excluded from this study because their judgement is may be influenced by the very new experience with online shopping. 20 users participated in the survey. Table 4 shows the main characteristics of the survey participants. The main tasks of the survey participants were to use and experience the online shops and then to fill out the AttrakDiff questionnaire. Afterwards, they were asked to assess the ideas for hedonic features generated in the expert workshop. Each idea was described in two or three sentences accompanied by a written example. The question was “How interesting is the described product feature?” The rating scale was “not at all” (1), “little” (2), “medium” (3), “very” (4) and “extraordinary” (5) (translated from German). Each participant had to follow the following procedure: 5. 6. 7. 8.

They received an email and had to read the instructions and follow the links to the shops and the AttrakDiff questionnaire. Using one of the online shops for 20 minutes and afterwards filling out the AttrakDiff questionnaire. Using the other online shop for 20 minutes and afterwards filling out the AttrakDiff questionnaire. Rating the hedonic features invented by the expert designers in the workshop in an online form.

3.4. Results 3.4.1. Feature Ideas During the workshop, the design experts produced 26 ideas for stimulating product features. The ideas from the workshop are listed in Table 5. The potential for hedonic quality stimulation is described in the column “Potential for stimulation” of Table 5. This has been added on the basis of the verbal comments of the workshop participants and based on the workshop analysis of the moderators. Furthermore, the average rating of the 20 users is shown in column “Rating”. The ideas are ordered from “extraordinary” interesting till “not at all” interesting. In column “Feature”, each feature idea is categorized by the four groups of features which can be designed (see also Fig. 1). This categorization was done by the moderators of the workshop. Comparing the feature ideas with the product features “presentation”, “interaction”, “content “and “functions”, it is obvious that most of the ideas are content and functionality features. There are 15 feature ideas belonging to the group of content features (C), 7 feature ideas belong-

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

225

a) www.extra-tour.com

b) www.globetrotter.de Figure 2. Online Shops used in the workshop and in the survey.

ing to content and functionality (C&F and F), and 4 ideas are presentation features (P). No idea on interaction methods (I) was generated in the workshop. In order to focus the main result from Table 5, the feature ideas which are rated as “extraordinary” or “very” were taken from list. Some of these feature ideas show some

226

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

Table 5. Ideas for features to improve the hedonic quality stimulation of an online shop for outdoor products and services. No.

Feature ideas

Potential for stimulation

Rating

Feature*

Rating “extraordinary” 1.

Tool for systematic comparison of products

Learning more about the products and finding the optimum product.

4,6

C&F

2.

Data sheet with product specifications

Exploring the quality of a product. Learning new things about products. Getting inspiration on product use.

4,6

C

3.

Providing the possibility to have different views on the product, e.g. to check details like a seam of an outdoor sleeping bag.

Exploring the product. Learning new things about the product and broadening the future usage possibilities.

4,0

C&F

4.

Product reviews of other users

Getting more information about products. Learning which products are the best.

3,9

C

5.

Tool for trip planning (e.g. checklists of equipment, trails, interesting sites)

Getting new ideas about trips and how to equip one self with optimum products.

3,8

C&F

6.

Field reports of the provider, e.g. which cloth can be used for extreme coldness

Finding new solutions and may open new possibilities for outdoor activities.

3,8

C

7.

Field reports of other users, e.g. how good was the tent x in Alaska.

Learning from other users.

3,7

C

8.

Illustration of product use (e.g. use of climbing equipment or dress an avatar with clothes of interest)

Learning about products and their fit to the user’s outdoor goals. Enhancing the personal outdoor activities.

3,6

C

9.

Special offers on the home page in order to attract the users

Inspiration for good bargains. Get to know new products.

3,4

C

10.

Recommendations of the provider (e.g. “for vacations in Finland a mosquito net is important”)

Learning more about products and their advantages. Enhancing the personal outdoor activities.

3,4

C

11.

Offering more products fitting to the context of use of the chosen product

Learning about interesting products for the context of use.

3,3

C

12.

Tutorials on product use, e.g. how to use a GPS orientation system in the wilderness.

Learning more about outdoor activities and products enhancing outdoor possibilities.

3,2

C&F

13.

Tips and tricks for different outdoor activities, e.g. rockfall protection in the mountains.

Enhancing the personal knowledge about living in nature and outdoor activities.

3,1

C

14.

Personalisation of the shop, e.g. personal welcome for the specific user

— has more to do with hedonic quality identity —

2,9

F

15.

Introduction of a holiday atmosphere, e.g. by showing pictures of people in the nature, on the mountain etc.

Introduction of good feelings can stimulate interest for outdoor activities

2,8

P

Rating “very”

Rating “medium”

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

227

Table 5. (Continued.) No.

Feature ideas

Potential for stimulation

Rating

Feature*

16.

Notebook function for storing interesting products and personal notes

Possibility to collect products and to take notes in order to a more comprehensive preparation of the vacations planed.

2,7

C

17.

Presentation of the online shop should be according to the products offered, e.g. the shop should look a little wild like outdoor live is.

Enhancing the outdoor experience during shopping.

2,7

P

18.

Recommendations concerning buying decisions of other users (e.g. “user x has bought the product y, because …”)

Learning about the reasons of other users to buy a product in order to take smarter buying decisions.

2,7

C

Rating “little” 19.

Interactive sales assistant for answering questions as well as providing support and information concerning products and shop use

Learning from a virtual coach.

2,3

C&F

20.

Scurrile products, like a small oven for the trouser pocket, should be presented.

Inspiration for new possibilities of product use.

2,2

C

21.

Unobtrusive animations fitting to the online shop theme, e.g. for outdoor shops small insects crawling around

Enhancing the outdoor experience during shopping.

1,9

P

22.

Lists of favourite products of other users.

Learning about new products which are relevant for users who are like me.

1,8

C

23.

The online shop should present sounds fitting to outdoor activities, e.g. crackling of a fireplace.

Enhancing the outdoor experience during shopping.

1,8

P

24.

Thematic related accessories like screen saver with fancy photos of outdoor live.

Having outdoor inspiration by photos.

1,6

C&F

Rating “Not at all” 25.

Offering thematic related games

Giving the opportunity to have some more fun with outdoor themes.

1,4

C&F

26.

Products should be presented embedded in a story, e.g. a photo story of a Nepal trip with all products which can be used on such a tour

Enhancing the outdoor experience during shopping. Inspiration for new outdoor activities.

1,4

C

* Groups of Features which can be designed in interactive products: P = presentation, I = interaction, C = content, F = functionality.

similarities and can be integrated to one feature. The following interesting clusters of feature ideas were identified: 1. 2.

Data sheets and tool for systematic comparison of products (integrated from feature 1 and 2, see Table 5). Different product views (feature 3, see Table 5).

228

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

3. 4. 5. 6.

Product reviews of other users (feature 4, see Table 5). Tool for trip planning (feature 5, see Table 5). Field reports by other users and the shop provider (integrated from feature 6 and 7, see Table 5). Illustration of product use (feature 8, see Table 5).

3.4.2. Design Methods After having generated the ideas for stimulating features, the experts were asked to collect methods and principles for the design of hedonic features: 1.

2.

3.

4.

The main and central method recommended by the design experts was the in deep analysis of relevant target groups. It was extremely important to the experts to understand the experiences of target users with other products as well as their needs and wishes. Here, all methods for context of use analysis like interviewing, contextual inquiry and also online surveys were presented as helpful. In addition to the context and user analysis, the evaluation of products of competitors was rated as very important. These evaluations can be done by experts (e.g. marketing, designer, etc.) as well as users. The more perspectives influence the evaluation of the products the better. A continuous collection of design ideas concerning pragmatic and hedonic features in the design team can be very helpful. Sources of this collection can be user feedback and ideas by the product design team. A catalogue of ideas can be a set up. From that catalogue the most interesting feature ideas can be selected and integrated into a product design process when needed. A design feature should not just be an ornament or just a fun feature. It should have a long-term advantage for hedonic quality.

3.4.3. Online Shop Evaluation Together with the rating of the feature ideas, the participants of the online survey had to answer the AttrakDiff questionnaire concerning the two online shops www.extratour.com and www.globetrotter.de (see Fig. 2a and 2b). The shop www.globetrotter.de provides the following hedonic product features which were rated as extraordinary or very interesting: product data sheets and tool for systematic comparison of products (integrated from feature 1 and 2, see Table 5), for some products different product views (feature 3, see Table 5), tool for trip planning (feature 5, see Table 5) and field reports (see feature 6 and 7, Table 5). The shop www.extra-tour.com provides just a few data sheets and possibilities for comparing products. Figure 3 shows the average rating of the AttrakDiff items (see Table 1) for the four relevant perceived qualities: pragmatic quality (PQ), hedonic quality identity (HQ-I), hedonic quality stimulation (HQ-S), and the overall judgment attractiveness (ATT). Most interesting in this case is the difference of the averages of HQ-S. This difference is statistically significant (.05-level) and the www.globetrotter.de shop shows a higher HQ-S value than www.extra-tour.com. Furthermore, it is obvious that the pragmatic quality of www.extra-tour.com is significantly (.05-level) better than the pragmatic quality of www.globetrotter.de.

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

229

Figure 3. Average user ratings for the items of the AttrakDiff questionnaire for the perceived qualities pragmatic quality (PQ), hedonic quality identity (HQ-I), hedonic quality stimulation (HQ-S) and the overall judgment attractiveness (ATT).

3.5. Discussion Most feature ideas created by the experts in the workshop fall into the category content and functionality (15 feature ideas on content, 7 feature ideas on content and functionality, 4 feature ideas on presentation and no feature idea on interaction methods). This can be interpreted in the following four ways: 1.

2.

3.

4.

The design experts have the opinion that content and functionality features have the highest influence on the stimulation aspect of hedonic quality applied in online shops. This is supported by the fact that the users rated the presentation features as “medium” and “little” interesting. This sample of design experts had special focus on content and functionality, which might be different for other design experts. In order to investigate this, more workshops with other design experts are required. Feature ideas were collected and presented in a verbal format during the workshop. This might facilitate the invention of content and function features. Other workshop or study formats where design experts e.g. have to design a product can clarify this. During the method collection, the experts said that it is important to design hedonic features ensuring a long term interest in the product. Presentation and interaction features are seen as features with a short run effect on hedonic quality – stimulation and content and functionality features are seen as features interesting the users in the long run.

The presentation features were rated quite low in the online survey. One explanation is that this type of features is not of central interest for the users and has just small impact on the stimulation aspects of the hedonic quality. Another reason for that might be that the participants had no experience with the presentation features. In the online survey, all feature ideas were just presented on the basis of a textual description combined with a written example. This might have caused the impression that these features are not that interesting. Studies with prototyped features can clarify that. When having a closer look to the features rated high, it is obvious that they have stimulation aspects of the hedonic quality, but pragmatic quality as well. There are two

230

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

possible explanations. First, the users mixed up hedonic and pragmatic qualities when rating the feature ideas. Here, the instruction to rate the feature ideas whether they are more or less “interesting” might be not clear enough. Other ways of rating the stimulation aspect of hedonic quality can be more precise. One possibility might be to rate all feature ideas based on the HQ-S items of the AttrakDiff questionnaire (see Table 1, items HQ-S 1 to HQ-S 7). Or second, users perceive more than one quality in the same feature. For example, on the one hand, a planning tool can induce a perceived pragmatic quality, as it is more effective and efficient to plan a trip with a planning tool. On the other hand, a planning tool could also cause perceived hedonic quality. Users are stimulated by the suggestions of a planning tool. They might expect a richer experience of a trip, when the tool recommends interesting sites and equipment which increases the enjoyment, e.g. a nice underwater cover for the camera when planning a journey to the barrier reef in Australia. The online shop www.globetrotter.de has some of the high rated hedonic features (product data sheets and tool for systematic comparison of products, for some products different product views, a tool for trip planning and field reports). This shop has a significantly higher rating of the hedonic quality stimulation (HQ-S, see Fig. 3). However, the question why this shop does not show a higher value of hedonic quality-stimulation must be posed. One explanation is that the hedonic features have not been found in the rather complex shop and therefore the users did not experience the hedonic features. It seems to be quite obvious that features which can not be detected are not able to induce hedonic quality. The design method collection of the design experts in the workshop focuses very much on user analysis and analysis of competitor products. This is in conformity with the theoretical model (see Fig. 1). Taking into account that pragmatic and hedonic qualities are perceived qualities, it becomes clear that the experience of the user must be investigated. This is necessary, because users compare a product on the basis of their experience with other products of the same product category or comparable product categories. In this case, users are only stimulated, when a product shows unknown features or features with a potential of new usage possibilities or new content. Of course, these methods are not new in the field of systematic design of interactive products. On the one hand, this could mean that the field of designing hedonic products is so new to design professionals that, at the moment, no new methods are in practical use at least in the workshop sample. However, new design support methods like culture probes [26,27] supporting complete new product ideas are available and applied in projects, e.g. in projects on technology for families [28]. On the other hand, traditional usability methods cannot only be used to increase usability, but also the joy of use. This is discussed by Pagulayan, Steury, Fulton and Romero [29] by applying classic user testing based on thinking aloud and behaviour observation in order to increase the fun of using digital games. Whereas in user testing, the focus is on finding usability problems [29,30], Pagulayan et al. [29] try to identify usage situations preventing fun in order to change the design for inducing more fun. The results of this exploratory study can have some consequences for the definition of a systematic design method for the design of products which show a high degree of perceived hedonic quality-stimulation. Derived from theory and inspired from the study results, the following designrelevant aspects can be listed.

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

231

1.

Users can experience different perceived qualities in one feature Users can perceive pragmatic and hedonic qualities in one feature. As shown above, a feature can help the users to achieve their goals in an effective and efficient way, but it is also possible that the same feature is stimulating because new usage possibilities are opened for example.

2.

Implications of perceived qualities Hedonic quality is perceived by the users. This leads to the following consequences: First, the feature must be detected by the users; otherwise they will have no experience with that feature. Second, if the users have detected the feature, they must understand which stimulating possibilities are opened by the feature. Third, the experts in the workshop stressed very much that the attributes and the experiences of the target user groups must be explored. One reason for that is that users compare their experience of a product with their stored experiences with other products. Therefore, as a designer it is necessary to know more about the products users normally use. Only by knowing that, it is possible to invent and design new interesting features. The design experts recommended evaluating products of competitors in order to explore what the “state-of-the-art” designs are. With this knowledge, it is possible to go further in the design of the product to be developed.

3.

Knowing the usage situations The experience of a product feature depends on the user mode and the usage situation. Especially, in on-line shops users can have different usage modes according to their decision phase in the buying process. When designing product features, it is important to take the different usage modes and the different situations of use into account.

4.

Long-term effects Short-term effects might be interesting to attract the attention of users and to induce “fun” [1]. However, it is also important that the stimulation aspect of the hedonic quality lasts for a longer use of a product. Especially for the application domain online shops, it is important to stimulate the users at each visit. Interesting content and functions seem to be appropriate for long-term effects. Here, feature ideas like field reports, planning tools, different product views and tools for product comparison can help.

5.

Creativity and inspiration Especially the hedonic quality stimulation requires new feature ideas, which are different from what the users already know. For example for online shops, interesting content fitting the needs and interests of the users seem to be very important. It is essential that the basics for inspiration are the user characteristics, user behaviour, and user requirements. A combination of creativity techniques and user research is required.

4. Further Research This exploratory study had the goal to explore types of feature ideas, design principles, and methods in order to design products showing a high degree of the stimulation as-

232

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

pect of the perceived hedonic quality. The approach of combining theoretic foundations and tacit knowledge of design experts with user centred evaluation of the outcome of design work seems to be promising to investigate the design of hedonic quality more closely. Based on this combination, successful strategies to design attractive products can be derived. The strategies described herein have to be developed further on the basis of comparable studies. For further research the following aspects seem to be important: • • •

If tacit knowledge of design professionals for designing hedonic products is in the focus, practical design should be observed and discussed. Just verbal generation of ideas might narrow the design space. Empirical evaluation of hedonic design features should be done on the basis of prototypes and not just on the basis of textual descriptions. The assessment of the hedonic quality by users should be done on the basis of the proved hedonic items of the AttrakDiff questionnaire.

References [1] M. Hassenzahl, The thing and I: Understanding the Relationship between User and Product. In: M.A. Blythe, C.J. Overbeeke, A.F. Monk & P.C. Wright (Hrsg.), Funology. From Usability to Enjoyment. Human-Computer Interaction Series. Volume 3 (Dordrecht: Kluwer Academic Publishers), (2003), 3–42. [2] ISO 9241–11, Ergonomic requirements for office work with visual display terminals (VDTs) – Part 11: Guidance on usability. International Organization for Standardization, 1998. [3] P.W. Jordan, Designing Pleasurable Products. London: Taylor & Francis, 2000. [4] M. Burmester, M. Hassenzahl, F. Koller, Usability ist nicht alles – Wege zu attraktiven interaktiven Produkten. I-Com, 1 (2002), 32–40. [5] M.A. Blythe, K. Overbeeke, A.F. Monk, P.C. Wright, (Eds.), Funology. From Usability to Enjoyment. Human-Computer Interaction Series. Volume 3. Dordrecht: Kluwer Academic Publishers, 2003. [6] M. Igbaria, S.J. Schiffman, T.J. Wieckowski, The respective roles of perceived usefulness and perceived fun in the acceptance of microcomputer technology. Behaviour & Information Technology, 13(6) (1994), 349–361. [7] C.J. Overbeeke, J.P. Djajadiningrat, C.C.M. Hummels, S.A.G. Wensveen, Beauty in Usability: Forget about Ease of Use! In: W.S. Green & P.W. Jordan (Eds.), Pleasure with products: Beyond usability (Taylor & Francis), pp. 9–18, 2002. [8] P.W. Wright, J. McCarthy, L. Meekinson, Making Sens of Experience. In: M.A. Blythe, C.J. Overbeeke, A.F. Monk & P.C. Wright (Hrsg.), Funology. From Usability to Enjoyment. Human-Computer Interaction Series. Volume 3. Dordrecht: Kluwer Academic Publishers, pp. 43–53, 2003. [9] D.A. Norman, Emotional Design. Why We Love (or Hate) Everyday Things. New York: Basic Books, 2004. [10] D. Chao, Doom as an interface for process management. In: Proceedings of the SIGCHI conference on Human factors in computing systems. New York: ACM, pp. 152–157, 2001. [11] J. Karat, C.-M. Karat, That’s Entertainment. In: Blythe, M.A.; Overbeeke, K.; Monk, A.F.; Wright, P.C. (Hrsg.). Funology. From Usability to Enjoyment. Human-Computer Interaction Series. Volume 3 (Dordrecht: Kluwer Academic Publishers), pp. 125–136, 2003. [12] G.H. Hofmeester, J.A.M Kemp, A.C.M. Blankendaal, Sensuality in product design: a structured approach. Conference on Human Factors and Computing Systems. Conference proceedings on Human factors in computing systems. New York: ACM Press, pp. 428–435, 1996. [13] R.J. Logan, Behavioral and emotional usability: Thomson Consumer Electronics. In: M. Wiklund (Ed.), Usability in Practice. Cambridge, MA: Academic Press, 1994. [14] B. Lawson, How Designers Think, 3rd Edition. Oxford: Architectural Press, 2002. [15] ISO 13407, Human-centred design processes for interactive systems. International Organization for Standardization, 1999. [16] J.M. Carroll, Beyond Fun. Interaction, Vol. XI.5 (2004), 38–40. [17] J. Ramsay, A Factor Analysis of User Cognition and Emotion. In: CHI97 Proceedings, 22–27 March, 1997, (Atlanta: ACM), pp. 546–547, 1997.

M. Burmester and A. Dufner / Designing the Stimulation Aspect of Hedonic Quality

233

[18] M.A. Blythe, M. Hassenzahl, P.C. Wright, More Funology. Interaction, Vol. XI.5 (2004), 37. [19] M. Hassenzahl, M. Burmester, F. Koller, AttrakDiff: Ein Fragebogen zur Messung wahrgenommener hedonischer und pragmatischer Qualität. In: J. Ziegler & G. Szwillus (Hrsg.), Mensch & Computer 2003. Interaktion in Bewegung (Stuttgart: Teubner), pp. 187–196, 2003. [20] M. Hassenzahl, A. Platz, M. Burmester, K. Lehner, Hedonic and Ergonomic Quality Aspects Determine a Software’s Appeal. Proceedings of the CHI 2000 Conference on Human Factors in Computing. New York: ACM, pp. 201–208, 2000. [21] M. Burmester, M. Hassenzahl, F. Koller, Wie attraktiv ist mein Produkt? AttrakDiff-2, ein Instrument zur Messung der Attraktivität interaktiver Produkte. In: Ziegler, J. & Beinhauer, W. (Hrsg.). Interaktion mit komplexen Informationsräumen. München: Oldenbourg, in print. [22] M. Hassenzahl, R. Kekez, M. Burmester, The importance of a software’s pragmatic quality depends on usage modes. In H. Luczak, A.E. Cakir & G. Cakir (Eds.), Proceedings of the 6th international conference on Work With Display Units – WWDU 2002 (Berlin: ERGONOMIC Institut für Arbeits- und Sozialforschung), pp. 275–276, 2002. [23] A. Chak, Submit Now – Designing Persuasive Web Site. Boston: New Riders, 2002. [24] AttrakDiff, On-line Questionnaire (URL). Retrieved Sept. 15, 2004 from http://www.attrakdiff.de, 2004. [25] M. Hassenzahl, The interplay of beauty, goodness and usability in interactive products. Human Computer Interaction, 19 (2004), 319–349. [26] B. Gaver, T. Dunne, E. Pacenti, Cultural Probes. interactions, Volume 6, Issue 1 (1999), 21–29. [27] W.W. Gaver, A. Boucher, S. Pennington, B. Walker, More funology: inspiration: Cultural probes and the value of uncertainty, interactions, Volume 11 Issue 5 (2004), 53–56. [28] H. Hutchinson, W. Mackay, B. Westerlund, B.B. Bederson, A. Druin, C. Plaisant, M. BeaudouinLafon, S. Conversy, H. Evans, H. Hansen, N. Roussel, B. Eiderbäck, S. Lindquist, Y. Sundblad, Technology Probes: Inspiring Design for and with Families. In: Proceedings of CHI 2003, April 5–10, 2003, Ft. Lauderdale, Florida, USA. New York: ACM, 2003. [29] R.J. Pagulayan, K.R. Steury, B. Fulton, R.L. Romero, Designing for fun: user testing case studies. In: M.A. Blythe, K. Overbeeke, A.F. Monk & P.C. Wright (Eds.), Funology – From Usability to Enjoyment (pp. 137–150). Dodrect: Kluwer, 2003. [30] J.C. Dumas, User-Based Evaluations. In J.A. Jacko & A. Sears (Eds.), The Handbook of Human Computer Interaction (1093–1116). Mahwah, New Jersey: Lawrence Erlbaum, 2002.

This page intentionally left blank

Emotions and Emotional Agents

This page intentionally left blank

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

237

On the Role of Self Esteem, Empathy and Narrative in the Development of Intelligent Learning Environments Paul BRNA School of Informatics, Northumbria UniversityNewcastle upon Tyne NE1 8ST

1. Introduction In the last few years, developers of interactive learning environments have sought to engage the intended users in the development of systems (Scaife et al., 1997; Cooper & Brna, 2000; Waraich, 2003). The methods have been various but the core of both participant design and informant design methodologies is to acknowledge the importance of drawing on the experience of those involved and to empower these users to some extent. In parallel to this development, designers wanting their applications to be attractive and, perhaps more important, to engage the learner, have sought to utilise notions associated with narrative (Plowman et al., 1999; Waraich, 1998). While the different meanings/interpretations of the term ‘narrative’ need careful treatment, there is no doubting the pervasiveness of the belief that ‘narrative is good’ in some quarters of the learning environments community. However, is it necessary to utilise notions about narrative in the design of learning environments? Is there a relationship between the methods used to produce learning environments and the qualities of the learning environments? We will return to these questions. Again, a parallel development has been the increasing awareness of the importance of the user’s feelings when working with software. HCI researchers have defined usability in terms of effectiveness, efficiency and user satisfaction. In the 1980’s the emphasis was on effectiveness and efficiency but there has been a trend towards placing more importance on user satisfaction, and more recently, even notions of fun and pleasure have begun to be considered more seriously (Price et al., 2003). If the user’s feelings are to be taken into account then how does this fit into the design process? Finally, for this paper, there is the issue of the designer’s duty of care to the learner. In the artificial intelligence in education community, John Self argued that caring for learners involved responding to their needs — often in terms of some form of personalisation/adaptation (Self, 1999). Caring for learners goes beyond simply making sure that the system provides sensible responses that meet the specific needs for effective and efficient learning — to truly care for learners there is a need to create and maintain a mental model of others which draws on our own experiences and feelings (Cooper, 2003). Cooper argues that a lack of empathy in teachers may lead to an unbalanced focus on management and curricular issues and this can have a detrimental effect on the learner’s motivation (Cooper, 2003). This may well be the case but how should designers of interactive learning environments respond? For example, do designers themselves need to have empathy for learners? Do they need to express this empathy for learners,

238

P. Brna / On the Role of Self Esteem, Empathy and Narrative

and how do they do this? If designers of educational software fail to factor empathy into the design process adequately, do they also concentrate too much on design issues connected with management and the curriculum?

2. Empathy Empathy can be defined in a number of distinct ways — all of which have some bearing on the problem of utilising empathy in the design of learning environments. Thinking of empathy in terms of levels of explanation, Preston and der Vaal have sought to establish a unifying view of the various disparate perspectives on empathy (Preston & deWaal, 2001). They are particularly interested in Mayr’s distinction between proximate and ultimate causes, allowing for different, consistent interpretations of cause and effect. Mayr states that “proximate causes govern the responses of the individual (and his organs) to immediate factors of the environment while ultimate causes are responsible for the evolution of the particular DNA code of information with which every individual of every species is endowed” (Mayr, 1961, p. 1503, cited in (Preston & deWaal, 2001)). Preston and de Waal’s process model makes “empathy a superordinate category that includes all sub-classes of phenomena that share the same mechanism. This includes emotional contagion, sympathy, cognitive empathy, helping behaviour, etc. This process model also links empathy to all facilitation behaviours that rely on perception-action” (Preston & deWaal, 2001). Emotional contagion is the notion that, for example, seeing a person smile literally evokes a muscular response emulating a smile — or seeing a child in a state of fear evokes fear in the child’s mother. These forces are very basic, and have been exploited — knowingly or unknowingly — in recent work on believable agents. For example, Lester’s “Herman the Bug” shows high levels of interest — and perhaps is exploiting a form of emotional contagion which might be part of the “Persona Effect” (Lester et al., 1997). On the other hand, the educational community is often more interested in some form of cognitive empathy, a conscious, rational assessment of another’s situation as found in Roger’s work: “... the state of empathy or being empathic is to perceive the internal frame of reference of another with accuracy and with the emotional components and means which pertain thereto as if one were the person, but without ever losing the ‘as if’ condition” (Rogers, 1975)

Haynes and Avery regard empathy as embedded in some interaction. “... the ability to recognize and understand another persons perceptions and feelings, and to accurately convey that understanding through an accepting response” (Haynes & Avery, 1979)

Already we can see that ‘designing in’ empathy into learning environments works both at the conscious and unconscious levels. If we bring into consideration that all learning environments ‘stand in’ in some way for the teacher, then how does a good teacher express empathy? The empathic teacher treats children as individuals by seeking to discover a pupil’s existing skills and seeks to help them develop. The empathic teacher knows the child as a person, knows their con-

P. Brna / On the Role of Self Esteem, Empathy and Narrative

239

fidence levels, as well as their knowledge. The empathic teacher also nurtures each child’s sense of self, supports their academic progress, and seeks to develop each child’s awareness of themselves. Cooper et al. provide a more elaborated description in (Cooper et al., 2000). So we take the position that • •

A strong sense of empathy is a valuable, probably essential, characteristic for designers of learning systems. Good teachers demonstrate a set of empathic characteristics that provide a starting point for the development of better quality interactions between the learning environment and the learner.

3. Self Esteem We recently finished a two year project that — to an extent — extended the work of the EU funded NIMIS (Networked Interactive Media In Schools) project (Brna & Cooper, 2000a; Cooper et al., 2000). The Nuffield Foundation “ICT and the Whole Child” project sought to follow up the evaluation work in the NIMS project with a view to examining more carefully the long term effects of helping children to learn in an ICT intensive environment. During the NIMIS project (Cooper & Brna, 2000; Brna & Cooper, 2000b) we found evidence that carefully designed, high quality, child-friendly ICT facilities integrated within normal classrooms generated high levels of collaboration and engagement in learning as well as the side effect of considerable increases in children’s ICT skills. We had some evidence in the NIMIS project that there were also content-based improvements (Cooper & Brna, 2002b) connected with the software specially designed for developing literacy through story writing (Brna & Cooper, 2002). Over a period of nine months, the children in the NIMIS classroom improved their reading over the school year by 14 months on average, which was five months more than the expected average increase. The children who made most gains were mainly those in the middle and lower areas of the attainment range. We observed growing confidence in both literacy and ICT skills. The experiences underlying these observation could be expected to add to children’s confidence and esteem but we made no specific attempt to confirm this during the lifetime of the NIMIS project. The “ICT and the Whole Child” project sought to examine the growth (or otherwise) of confidence and self esteem over a longer period of time and using appropriate measures. As with the NIMIS project, we sought to examine whether a carefully designed classroom integrated within normal classroom activities and used across the curriculum over two years in Key Stage 1 (Years One and Two) provided support for improved achievement across the curriculum and increased motivation and self-esteem. Standardised tests were used for assessing reading age (Bookbinder, 2002) and self-esteem (Maines & Robinson, 1988). Standardised assessment results were also collected. The results were interesting. Though self-esteem tends to drop as children move through school — the self-esteem tests taken by the children in the project class show self esteem holding up quite well relative to the comparison class and certainly confidence, oracy and skills are evident in their use of the computers (Fig. 1). The project and comparison classes were somewhat different in composition. The project class has 25% children with special needs (the other class had only one SEN

240

P. Brna / On the Role of Self Esteem, Empathy and Narrative

Figure 1. Self Esteem Progress.

Figure 2. Progress in Standardised Assessments 2003.

child). It also lost some of the high achievers during the year. In this, and other respects, the project class children were at a greater disadvantage — and yet, their self esteem held up much better. To partly explain this, the project class had a large touch screen and extra computers arranged in a manner that promoted child-child interaction. In terms of growth of ICT competence, there was really no comparison with the project class which far outstripped the comparison class. More importantly, SAT tests demonstrated that the project class did better in Maths, Science and Reading (Fig. 2). Writing, unsurprisingly, was not so successful — the original NIMIS project had emphasised support for creative writing but the follow on project was not able to take advantage of T’rrific Tales, the software written to support collaborative story writing (Brna & Cooper, 2002).

P. Brna / On the Role of Self Esteem, Empathy and Narrative

241

Evidence was obtained from interviews with teachers and children indicating that the children generally enjoyed learning in the project classroom (Brna & Cooper, 2003). This was also supported through a systematic analysis of the various hours of videos taken of the children at work. Project teacher 1: “It has a motivating, uplifting element to it – they’re not going to sit there and switch off. You can give a child a piece of work and five minutes later you can come back and he’s done nothing – he’s switched off and gone off to his own little planet — but the same activity but on a computer and he will do it — he will do it — I don’t know why it is — I wonder why? — it’s just more interesting – do they think it’s more of a game? — which tells me that maybe we should make our teaching and learning more interactive and more play orientated.” Steven: “I like it in when you get to go on computers in our classroom — and class four that we used to be in had four as well – and they’ve both had big screens so I’m quite glad we went into this classroom because just before we went into that classroom they got a big screen (gesticulating – excited).” Project teacher 1: “they can be sitting there with the key board and you can be sitting there and we’re doing a shared piece of writing and you just sit there, the children are there, and you’re amongst them and you’re with them, you’re there — and you’re all looking at the screen and you’re sitting there with them and it just makes it more different, you’re not facing them — you’re with them and you’re doing a shared piece of writing which is what ‘shared’ should be isn’t it...”

It would be reasonable to seek several kinds of explanation for these effects but one strand of the story is that there is the intention to improve self esteem was present in at least some, if not all, of the class teachers. “I think it’s sometimes — often — the quiet ones who haven’t got a good self-esteem and I’m very aware of every single child in my class. Each one is a very special person ... to their parent and I’ve got the job of looking after those for quite a lot of the year when you think about all the time they’re in school, not just to take care of them socially and through their development but to educate them as well ... you’re a parent and educator all at the same time — we have a big influence, I think we have a big influence on children ... so I am aware of involving everybody — especially on carpet sessions and working with every-body to let them feel really special, they’re special and although we’re in the same class we’re working individually but we’re also helping each other... and to try and build up their confidence and self-esteem” (Cooper & Brna, 2002a)

We find that some teachers have a clear aim to improve the self esteem of these young children — though I suspect that similar aims are less likely to be held by teachers of, say, children at secondary schools. We also find that technology may have a role to play in providing for an environment in which self esteem can be maintained more successfully than otherwise. What we do not know from this later work is whether adaptive learning environments actually help or hinder the development of self esteem. This is an important research question in the context of developing adaptive systems.

4. Narrative Earlier it was suggested that many regard narrative as a ‘good thing’. The discussion of the importance of narrative deserves more attention than can be given here, but we need to draw attention to the main points.

242

P. Brna / On the Role of Self Esteem, Empathy and Narrative

•

•

•

Remembering important facts or processes depends to an extent on the ways in which these experience are coded at the neurophysiological level. There is increasing evidence that this encoding is conditioned by the depth of feelings associated with the experiences (Damasio, 1999). The careful use of narrative in structuring some learning experience can strengthen the ability to recall facts/procedures. Constructing an understanding of the world involves comprehension of the basic data and the generation and revision of interpretations. Children be-gin to learn to do this early on when they both listen and make up stories with their parents, relatives and friends, and later, in primary school. The foundation of reasoning arguably lies in the logic of these early story telling experiences. Constructing a sense of oneself both as an individual and as a social being requires the same kind of processes. The generation of stories of a learner’s relationships with others and their experiences with the world serves to strengthen (or weaken) the learner’s self confidence and the repertoire of actions that help the learner to adapt to situations that threaten their self identity and social well being.

During the “Networked Interactive Media In Schools” (NIMIS) project we had the opportunity to develop a learning environment designed to help children aged 5–6 years old improve their story writing skills. The environment supported the development of the skills in learning to write stories through a number of methods — some of which are based on narrative ideas. The learning environment, called T’rrific Tales, featured a prototype agent, Louisa, which was designed to provide support for changing the learner’s focus, and encouraging the learner to extend their ideas and to develop them — though the final implementation was not as sophisticated as we wished, there was some evidence that children responded well (Cooper & Brna, 2001). Louisa was designed around a model of an agent expressing empathy for children according to Cooper et al’s work (Cooper et al., 2000). The model included six important aspects (Brna et al., 2001). While Louisa was designed to manage the ‘available’ aspect she did not fully manage the fifth aspect of ‘parting’. Attend:

Engage:

Value:

Encourage: Parting:

The empathic teacher (and agent) stops doing what they want to do, thinking what they want and expressing their own emotions and turns to consider the actions, thoughts and feelings of the learner. Here, we indicate the importance of attending to affect, cognition and sensorimotor issues. The teacher begins to align their actions, thoughts and feelings with those that the child is experiencing. The teacher makes it clear that this is going on. The teacher, by actions, words and the expression of appropriate emotions, makes it clear to the learner that they and their work is considered to be valuable to the teacher. The teacher then seeks to encourage the child to go further. This encouragement has emotional, physical and cognitive aspects. The teacher now turns to another matter, the work she was previously doing, or to the needs of another child. The link with the child is not broken, and ‘closure’ is not sought. This is achieved by a combination of

P. Brna / On the Role of Self Esteem, Empathy and Narrative

Available:

243

gesture & facial expression, comment and feeling, and indicates the availability of the teacher for later interaction. In a sense this fifth part is implicitly managed by Louisa because she is always available for consultation. More directly she might smile and suggest that she can be called if needed. The teacher is working at their own work — usually with other children in the class — but is ready to be available whenever needed. The teacher quietly supports the child. The agent is not as limited as a human teacher in that an agent has no requirement to manage a class of children and can in practice be interrupted at any time. This allows a good climate to be developed in the classroom freeing the human teacher to give high quality care.

This continued availability is central to the ethical and caring standpoint we have adopted. At the outset, the teacher is available (before attending) and at the end after returning to their work. This implication of further support is crucial to the teacher’s success, implying care, concern, continuity, support, security — there will always be someone there to value the work even if other children or the teacher are busy. She may not even be used but she is there — her being there allows progress because she imbues confidence and is a symbol of value even if no inter-action takes place. The quality of the support is crucial. Surrounding all these (six) aspects there is the key idea that the teacher (and software agent) models physical actions, thinking and emotional behaviour that the teacher wishes that the learner exhibit. The teacher (and agent) have a responsibility to do so since, when working with the learner’s feelings, it is especially important to remember that there is an ethical system at work. While Louisa was designed to provide a high quality interaction with learners, the environment also provided ‘static’ support. For example, the word bank included sections for story starters, story stirrers and story endings. This provided learners with a reasonably large number of ideas to take and customise to their own wishes. This structuring also reinforced the notion that stories have a beginning, a challenge and an ending — a model of narrative that was a little more sophisticated than the structural view of stories recited frequently in classrooms for this age group — i.e. that a story has a “beginning, middle and an ending”. There was evidence that working in the environment we developed for children did benefit them in terms, for example, of reading age. We also assessed their stories — both written with T’rrific Tales and without, and compared with stories written by children from a comparison classroom. The evidence therefore was that the T’rrific Tales software designed explicitly with both empathic and narrative features was broadly successful. By focussing on these aspects we believe that the system provided a safe and enjoyable environment in which children benefited from the narrative aspects through having vivid experiences and benefited from the empathic treatment by feeling their work was valued. In this sense, self esteem can be seen as an effect of being treated well/sensitively. It would seem that regard for the self esteem of learners was implicit in the design process. Without going into the degree of success achieved, the underlying question is whether these explicit aspects are necessary for the success of the product — or can similarly beneficial results be achieved by focussing on a different set of ‘necessary features’.

244

P. Brna / On the Role of Self Esteem, Empathy and Narrative

5. Conclusion We set out with the objective of discussing the production of learning environments which enhance the learner’s self esteem, ensure that the learner’s best interests are respected through paying attention to the narrative structure of the learner’s experience, and the ways in which communication can be enhanced through empathy with the learner. We discussed the necessity of using narrative in learning environments through the example of the NIMIS project and the work done with a carefully designed primary school classroom. The T’rrific Tales software was designed to both scaffold the development of narrative skills and to actively guide learners. The indication from the evaluations performed is that this worked well. Waraich has also demonstrated the interesting result that involving learners in the development of the software and teaching them a narrative theory sophisticated enough can benefit the resulting system (Waraich, 2003). Waraich has tried this adaptation of Scaife and Rogers’ informant design with a range of learners with good results. The indications are that there is value in involving the learner in the development of the narrative that drives the educational software. From briefly considering the role of empathy in the design of T’rrific Tales, the indications are that empathy practised consciously throughout the design process is worth doing. The careful deployment of both software and hardware and a sensitivity to work patterns can lead to high values for self esteem across a range of subjects. Intelligent learning environments adapt to the learner. From considering the results of Cooper and her colleagues, the indications are that the designer must consider how to empathise with the learner in particular ways. We do not know how intelligent learning environments help or hinder the development of self esteem. In principle this should be the case, and it is certainly likely that designers want this to happen — even if they conceive of their design in other terms. The evidence of the benefits of empathic narrative-based design for explicitly supporting the development of the learner’s self esteem is not yet available even if the indications are favourable. The next step is to set up the research agenda necessary to establish an adequate level of evidence.

Acknowledgements This work was funded in part by the Nuffield Foundation (Grant EDU/00256/G) and the EU (Experimental School Environments Grant No 29301). My thanks are due to Bridget Cooper for her extensive contributions to this work. References [1] Bookbinder, G. (2002). Salford Sentence Reading Test (Revised) Specimen Set. Hodder & Stoughton, London, 3rd edition. revised by d. vincentand m. crumpler. edition. [2] Brna, P. and Cooper, B. (2000a). Lessons from the nimis classroom: an overview of progress towards the ‘classroom of tomorrow’ in an english county primary school. Technical Report 00/11, Computer Based Learning Unit, University of Leeds. [3] Brna, P. and Cooper, B. (2000b). Lessons from the NIMIS classroom: An overview of progress towards the “classroom of tomorrow” in an english county primary school. Technical Report 00/11, Computer Based Learning Unit, Leeds University.

P. Brna / On the Role of Self Esteem, Empathy and Narrative

245

[4] Brna, P. and Cooper, B. (2002). Supporting young children learning to write stories together in a ‘classroom of the future’. International Journal of Continuing Engineering Education and Lifelong Learning, 12(5/6):485–503. [5] Brna, P. and Cooper, B., (2003). Humans and computers working together in schools: Improving achievement, motivation and self-esteem with ICT in key stage one, Final Report to the Nuffield Foundation, available from the authors. [6] Brna, P., Cooper, B. and Razmerita, L. (2001). Marching to the wrong distant drum: Pedagogic agents, emotion and student modelling. In Proceedings of the 2nd International Workshop on Attitude, Personality and Emotions in User-Adapted Interaction. Sonthoven, Bavaria. [7] Cooper, B. and Brna, P. (2000). Classroom conundrums: The use of a participant design methodology. Educational Technology & Society, 3(4):85–100. [8] Cooper, B. and Brna, P. (2001). Fostering cartoon-style creativity with sensitive agent support in tomorrow’s classroom. Educational Technology and Society, 4(2):32–40. [9] Cooper, B. and Brna, P. (2002a). Hidden curriculum, hidden feelings; emotions, relationships and learning with ICT and the whole child. In BERA, Exeter, September 2002. [10] Cooper, B. L. and Brna, P. (2002b). Supporting high quality interaction and motivation in the classroom using ICT: the social and emotional learning and engagement in the NIMIS project. Education, Communication & Information, 2(1/2):113–138. [11] Cooper, B.L. (2003). Care — making the affective leap: More than a concerned interest in a learner’s cognitive abilities. International Journal of Artificial Intelligence in Education, 13(1):3–9. [12] Cooper, B., Brna, P. and Martins, A. (2000). Effective affective in intelligent systems — building on evidence of empathy in teaching and learning. In Paiva, A., (ed.), Affect in Interactions: Towards a New Generation of Computer Interfaces, pages 21–34. Springer, Berlin. [13] Damasio, A. (1999). The Feeling of What Happens: Body and Emotion in the Making of Consciousness. Harcourt Brace, New York. [14] Haynes, L.A. and Avery, A.W. (1979). Training adolescents in self-disclosure and empathy. Journal of Community Psychology, 26(6):526–530. [15] Lester, J., Converse, S.A., Stone, B., Kahler, S. and Barlow, S.T. (1997). Animated pedagogical agents and problem-solving effectiveness: A large-scale empirical evaluation. In Proceedings of the Eigth World Conference on Artificial Intelligence in Education, pages 347–354. Kobe, Japan. [16] Maines, B. and Robinson, G. (1988). B/G-STEEM: a self-esteem scale with locus of control items. Lucky Duck Publishing Ltd, Bristol. [17] Mayr, E. (1961). Cause and effect in biology. Science, 134:1501–1506. [18] Plowman, L., Luckin, R., Laurillard, D., Stratfold, M. and Taylor, J. (1999). Designing multimedia for learning: Narrative guidance and narrative construction. In Proceedings of Computer-Human Interaction, CHI’99, pages 310–317. Pittsburgh, PA. [19] Preston, S.D. and de Waal, F.B.M. (2001). Empathy: Its ultimate and proximate bases. Behaviour and Brain Science, 25:1–72. [20] Price, S., Rogers, Y., Scaife, M., Stanton, D. and H., Neale. (2003). Using ‘tan-gibles’ to promote novel forms of playful learning. Interacting with Computers, 15(2):169–185. [21] Rogers, C.R. (1975). Empathic: An unappreciated way of being. The Coun-selling Psychologist, 5(2):2–10. [22] Scaife, M., Rogers, Y., Aldrich, F. and Davies, M. (1997). Designing for or designing with? Informant design for interactive learning environments. In CHI’97: Proceedings of Human Factors in Computing Systems, pages 343–350. ACM, New York. [23] Self, J. (1999). The defining characteristics of intelligent tutoring systems research: ITSs care, precisely. International Journal of Artificial Intelligence in Education, 10(3–4):350–364. [24] Waraich, A. (1998). Telling stories — the role of narrative in intelligent tutoring systems. In Proceedings of ED-MEDIA 98. AACE, Charlottesville, VA. [25] Waraich, A. (2003). Designing Motivating Narratives for Interactive Learning Environments. Unpublished Ph.D. thesis, Computer Based Learning Unit, Leeds University.

246

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

Empathic Characters in Computer-Based Personal and Social Education João DIAS a, Ana PAIVA a, Marco VALA a, Ruth AYLETT b, Sarah WOODS c, Carsten ZOLL d and Lynne HALL e a Instituto Superior Técnico and INESC-ID, Av. Prof. Cavaco Silva, IST, Taguspark, Porto Salvo, Portugal b University of Herriot Watt, Scotland, UK c University of Adaptive Systems Research Group, University of Hertfordshire, UK d Institute of Theoretical Psychology, University of Bamberg, Germany e School of Computing and Technology, University of Sunderland, Sunderland, UK

Abstract. The advent of techniques for affective computing [1], both in capturing user’s emotional states and allowing for systems to display emotionally charged expressions, has allowed for a new development of interactive media. Following these ideas, in this paper we will discuss the role of empathy in the construction of synthetic characters to interact with learners in intelligent learning environments. We tried to answer the question: How can we build synthetic characters that are able to evoke and establish empathic relations with learners in a interactive learning virtual environment? To do that, we will describe a system, FearNot!, that uses synthetic characters and role playing, developed as a set of bullying situations, which emerge from the actions and interactions between synthetic characters in a 3D virtual world. The system was designed to evoke affective responses by the users, in this case, children. FearNot! has been evaluated with 345 children in June 2004. The results achieved show that empathic interactions were achieved with synthetic characters. Keywords. Synthetic characters and emotions in learning environments

1. Introduction During the last few years, Intelligent Learning Environments brought new aspects such as flexibility and intelligence into current technology enhanced learning systems. However, most of the applications developed so far address learning areas such as maths, physics, languages, among others. The use of these systems in still problematic and complex areas such as science education, allows not only for proactive interventions, but also poses several challenges to the researchers that need to handle questions such as com2 plex problem solving. This need to build good applications in these areas has somehow hidden the necessity in other different areas of intervention, such as personal and social education (PSE) (or more recently as Personal, Social and Health Education – PSHE in the UK). However, the need for affective education has been emphasized as essential and should thus be a dimension of the educational process [3]. In general, PSE covers topics such as education against bullying and racism, on drugs, including smoking and alcohol, and sex education. These topics are addressed in many

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

247

forms in the curricula. One possibility is through the inclusion of specific targeted activities that, in an horizontal way, are developed across different disciplines. A common thread in these topics is that knowledge in and of itself is not sufficient to meet the pedagogical objectives, since attitudes and emotions are at least as important to producing desired rather than undesired behaviour. For this reason, techniques such as smallgroup discussion, role-play and dramatic performance by Theatre-in-Education TiE) groups may be used. So, it is natural that in these areas, which require changes in attitudes, issues such as emotional responses, awareness of the others, empathy, etc, must play a role. And dealing with these issues in a computer based learning environment is not easy. As we all know, technology in general, and computers in particular, are often associated with cold and perhaps non-emotional interactions. Aspects such as emotions, personality or empathy were, until recently, considered only possible in human-human interactions. However, the advent of techniques for affective computing [1] has allowed for a new development of interactive media with users. The use of affective interaction techniques, combined with technological achievements as the ones in the area of synthetic characters, allows for new types of systems, that in an effective way can address personal and social education issues. Combining these two aspects, in this paper we will try to answer the question: Can learners establish empathic interactions in a learning environment? So, can learners respond emotionally to the interactions established with synthetic characters using a learning environment? If we can answer this positively, then we are on the right track to build intelligent learning environments for PSE. To address this problem we will provide an example of an interactive application called FearNot! designed for addressing bullying problems in schools. By using synthetic characters and role playing, the application develops as a set of bullying situations, which emerge from the actions and interactions between synthetic characters in a 3D virtual world. The design of the characters and the situations explored, followed a framework that gives emphasis to the emotional reactions users will have to the synthetic characters. By designing situations, in particular exploring the narrative elements there embedded, the behaviours and establishing a degree of physical proximity between users and characters, we expected to attain emotional responses to the characters created. This paper is organised as follows. First we will discuss the issue of empathy and its relation to synthetic characters. Then we will describe the system (FearNot!) and describe how the characters were build. Finally we will present the results attained in the evaluation done last June, drawing some conclusions and future work.

2. Empathy for Learning Environments The most popular media, tv, films and literature, are masters in evoking empathic responses from the viewers. Indeed, we all establish affective relations with fictional characters, suspend our disbelief and look at them as “alive”. Differently, in computer based environments, characters are mostly seen as action-characters, detached from any kind of emotional behaviour. But, if we want to build environments populated with characters to apply in PSE, we need characters that evoke emotions and change attitudes. So, how can we build characters in computer-based learning environments that allow for the creation of empathic interactions with learners?

248

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

Usually, empathy is considered to be a process between two human beings, more precisely “any process where the attended perception of the object’s state generates a state in the subject that is more applicable to the object’s state or situation than to the subject’s own prior state or situation” [2]. So we have two people involved: one, known as the “subject” or more commonly referred to as the “observer”, and the other, who is observed and is referred as the “object” or “target”. In an empathic situation, the state of the subject changes due to the perception of an emotional state of the object. This is a very vague definition and although we think this is the most precise definition up to date it is still true that “empathy is, and always has been, a broad, somewhat slippery concept” [5]. Furthermore, most contemporary researchers consider that there are two types of empathy, namely cognitive and affective empathy. Cognitive empathy is considered as the basic aspect of empathy and is the awareness, or understanding of another’s state or condition [6]. In case of cognitive empathy the result of an empathic process is a change of the observer cognitive system. For example, when an observer perceives a target crying, he or she could conclude that the target is sad. But the observer doesn’t get sad himor herself. On the other hand, affective empathy is that the feelings or condition of a person (the target) generate strong vicarious emotions on the observer. In case of affective empathy the result of the empathic process is a change in the emotional system of the observer, i.e. the observer is feeling something. This would be the case if the observer is perceiving a target crying and is getting sad him- or herself due to that perception. This later type is the one most commonly considered as empathy. Given these aspects of empathy, we will try to solve the problem of how to build learning environments that evoke empathic reactions in children. We expect that these emotional reactions to situations will lead to attitude change, and awareness of situations, in accordance with PSE objectives.

3. FearNot!: An Application in the Area of Bullying Bullying is a difficult and devastating problem in our schools. Victimization problems are widespread and cross-cultural. Consequences may involve conduct disorder, hyperactivity, physical health problems, sickness, depression, anxiety and low self-esteem. The most common initiatives are theatrical performances or classroom discussions. However, these are necessarily collective, and in any group it is very likely that some individuals will be victims of bullying by some other in the group and will be inhibited in their participation. So our approach was to address this problem through a computer based environment, that children, individually, deal with situations of bullying and try to cope with them, helping our synthetic characters on what to do. FearNot! (Fun with Empathic Agents to Reach Novel Outcomes in Teaching) specifically aimed at anti-bullying education for the 8–12 age group. The structure of FearNot! is inspired by the Forum Theatre approach developed by the dramatist Augusto Boal and is as follows: there are a set of dramatic episodes divided by periods in which advice can be given to a character. Bullying is naturally episodic: it is distinguished from other aggressive behaviour by is very repetitiveness and by the unequal power relationship between bully and victim, the structure of the application follow this division into episodes. Each episode may involve verbal abuse, physical bullying, such as hitting, pushing and taking money, or, in the case of girls, the manipulation of social

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

249

Figure 1. A Snapshot of an episode in FearNot!

relationships against the victim (“You can’t sit with us”, “We’re meeting up tonight but you can’t come”), known as relational bullying. A child using FearNot! logs in and starts by watching an introductory scripted segment where the characters and school are introduced. This small introduction is followed by an agent-driven episode in which one of the characters is bullied (a ‘bullying scenario’ – see Fig. 1). At the end of each episode, the victimized character goes to a resources room (the school library) where the child user is asked to give them advice about how to cope with the bullying problem (see Fig. 2). For example, the child may advise the character to tell someone, like the teacher, or for example, to hit back. Then, the advise of the child is used in the application for the continuation and selection of the next episode. After a number of episodes the drama concludes with an educational message and a questionnaire assessing the extent to which the child user can ‘put themselves in the shoes’ of the characters they have seen and assess their motives and goals. The educational objectives of FearNot! depend on the child user being willing to engage with the problems faced by the victimised character. This requires the child to act as an ‘invisible friend’ – invisible because they are not themselves present in the dramatic episodes, and friend because they can advise and support the character but not act with god-like power to solve their problems for them. The eventual aim in FearNot! is to use the empathic agent architecture that has been developed for its characters (see [4]) to generate dramatic episodes from interaction, in a similar way to improvised drama.

250

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

Figure 2. An interaction window between the victim character and the child (for the physical bullying and relational bullying).

4. Building Synthetic Characters for Empathic Interactions with Children Animators and film makers have been creating unforgettable characters for years, characters that lead viewers to cry, become angry and react emotionally to what happens. However, creating embodied lifelike computer generated characters that have the power to make the user feel emotional reactions is still an unexplored research challenge. There are several factors to take into account and each one of them is, per se, may be regarded as a research topic. For example, what type of agent is more likely to evoke emotional responses from the user? A good starting point for the identification of possibilities to design empathy enhancing synthetic characters is Bischof-Köhler’s distinction of situation- and expressionmediated empathy [8]. The distinction relates to the source of which the observer receives the information about the state of the target. This source can be either the situation the target is dealing with or the emotional expression of the target. Situationmediated empathy is promoted by the story the synthetic agent is part of. For example, if the target agent is beaten up or insulted by another agent, the user can infer that the target is either sad, angry or scared. On the other hand, if the target is crying, it expresses its sadness (expression-mediated empathy). Different possibilities for expressing emotions exist and are perceived by the user. They can be, for example: facial expressions, posture, gesture, psychophysiological cues (e.g. flush) or paraverbal cues (like voice pitch). Following this research on empathy, we will address the problem of building synthetic characters for learning environments by considering the following elements: (1) the situation, that is, who to convey to the user some situation that allow for empathic responses; (2) the behaviour of the characters, that is, how can we develop archi-

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

251

tectures for the agents that lead to believable behaviour, capable of giving the illusion of life; and (3) the expression of the characters. All these elements are part of what we consider to be the needs of a system for evoking empathy. We will describe each one for the case of the characters in FearNot! (for more details, see [10]). (a) The Situation: As has already been outlined, situational aspects play a central role when it comes to promoting empathic interactions between humans and agents. In the case of FearNot! an essential aspect of the design and implementation of believable and interesting bullying scenarios concerns the profiles and roles designated for each character depicted within the scenarios. A number of research studies have been carried out to assess bullying profiles and a classification of distinct characteristics are evident for ‘pure’ bullies, ‘pure’ victim, bully/victim, bully/assistants, defenders and bystanders. Together with teachers and students we have promoted a set of activities for data gathering where a set of typical situations were described. We have used Kartouche [11] to create scenarios through storyboards, in order to assess the children’s preferences and empathy towards the characters with different roles. With these data, we have defined a high level description of typical scenarios, and each simulated episode within the bullying scenario act defines a set of encounters that enacts (rather than dictate) bullying situations. Furthermore, and in order to obtain certain degree of similarity between the children and the characters, we have considered specific situations for both genders (more direct bullying for boys and relational bullying for girls). (b) Behaviour: Characters must act in a believable way and in accordance with their roles. Their actions must be generated in a way that their behaviour allow children to recognize common situations found in schools. Given that we didn’t want to script the actions of the characters, the architecture developed for our FearNot! synthetic characters allows for a dynamic generation of actions in a believable and autonomous way. Such architecture, is used for the emergent version of FearNot! and was done in a way that the characters are able to generate actions, consistent with their role in the episode and their emotional state at the moment. (c) Expression and physical appearance: In order to design the characters we made some preliminary tests with children and designed two types of characters: realistic versus cartoon like. Although at first, one would be tempted to adopt realistic characters because they are more close to reality, it was clear from the studies and results that learners of this age preferred the cartoon characters (see [7]). Inspired by the very popular characters from a Portuguese children’s web portal (Cidade da Malta in http://www.cidadedamalta.pt/) originally in 2D, we have converted the characters into 3D and adopted them adequately. The facial expressions, although simple, convey the emotional state quite clearly as they are a bit exaggerated. Furthermore, the characters and the situations for the age groups we are targeting range a quite distinct set of children’s appearances so that children can easily identify with one or another character. For each country (UK, Portugal and Germany), we designed different characters given that children in the UK have uniforms and in Portugal and Germany do not.

252

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

5. Evaluation and Results One first version of FearNot! was trailed at the “Virtually Friends” event at the University of Hertfordshire, UK, in June 2004. Although not including an emergent narrative yet, the version used had the episodes and the characters designed with the principles here described. 345 children participated in the event. 172 were male (49.9%) and 173 were female (50.1%). The age of the sample ranged from 8 to 11 with a mean age of 9.95 (SD: 0.50). The sample comprised of children from a wide range of primary schools in the South of England. For each day of the event, 2 classes from different schools participated. The FearNot! evaluation session began with a general introduction to bullying and using the system. Children then completed questionnaires on bullying role and empathy profile. After completing the questionnaires, they interacted with FearNot!, firstly with a direct, and secondly, with a relational bullying scenario. The direct bullying scenario had 3 characters; John (the victim); Paul (the bystander) and Luke (the bully). The relational bulling scenario has 4 characters; Frances (the victim); Martinha (the bystander) and Sarah and Janet (the bully and the bully assistant). After each scenario, lasting approximately 15 minutes, the children completed an Agent Evaluation Questionnaire. With this evaluation, we wanted to find out if children responded emotionally to the situations and characters portrayed in the episodes shown. The results were as follows: Did the children like the characters we wanted them to like? Children stated that they liked John the victim the most, followed by Paul and Martinha, who both offered support to the victim in the direct and relational scenarios respectively. Janet and Sarah the female bullies had very little impact. Only 6 children stated that they liked these characters the most. 10% of the sample liked Luke the male bully the most. Children identified with the non-bullying characters the most in terms of who they would choose to be, i.e. 31% of children would choose to be Martinha and 25% would choose to be Paul the defender. 16% of children stated that they wouldn’t choose to be any of the characters. Children stated that characters they would most like to be friends with were John and Martinha followed by Paul and Frances. Few children wanted to be friends with a bully character. In line with our intentions, children could clearly be seen to prefer the non-bully characters. Did the children feel sad for the victims? Using FearNot! did evoke empathic feelings in most of the users, with 79% of children stated that they felt sorry for one/some of the characters in the dramas. Children predominantly felt sorry for John (80%) and Frances (67%) the victims in the dramas. Significant gender differences were revealed for feeling sorry for the characters (X = 18.28, (327) df = 1, p = 0.000) where females (89%) felt significantly more sorry for characters overall compared to males (70%). Children felt less sorry for characters if they chose to be friends with Luke (43.5%) or None (59%). The most empathy was expressed if children chose to be friends with Frances (the female victim).

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

253

Figure 3. Results from the evaluation: Most and Least liked character.

Did children feel anger towards the bullies? Interacting with FearNot! resulted in 72% of children stating that they felt anger towards one/some of the characters in the dramas. Children predominantly felt angry towards the male bully, Luke (70%), 58% felt angry towards Janet the bully assistant, 63% felt angry towards Sarah the bully. Significant gender differences were found (X = 11.06, (315), df = 1, p = 0.001) where significantly more females (81%) felt angry towards characters compared to males (64%). There was a significant relationship between prime character and feeling angry towards characters (p = 0.005). If children stated that they wanted to be Luke they were much less likely to state feeling sorry for the characters compared to others (42%). A significant relationship emerged between feeling angry for characters and which character children would choose to be friends with (p = 000). If children chose to be friends with Luke (26%), Janet/Sarah (64or none (58%) of the characters they were significantly less likely to feel angry with the characters. A significant relationship was found between children’s liked most character and feeling anger towards the characters in the drama (p = 0.007). Children who stated that they liked Luke (47%) the most expressed that they felt the least anger towards characters. This was followed by those that stated that they didn’t have a liked most character (65%).

6. Final Comments In this paper we have provided a discussion on some aspects that need consideration for the creation of synthetic characters that are able to establish an empathic relation with learners. To do that, we have relied on research on empathy and from them derived some guidelines for designing the characters in accordance. By describing FearNot!, an application designed for addressing bullying problems in schools, we have illustrated the issues. The results of the evaluation of the system carried out with 345 children show that children can indeed feel empathy towards the characters, and that there are some gender differences that emerge.

254

J. Dias et al. / Empathic Characters in Computer-Based Personal and Social Education

Although the results obtained are restricted to this particular application in the area of bullying, we do believe that the results can be extended to other areas of Personal and Social Education.

Acknowledgements The work here reported is part of the VICTEC project [9]. We would like to thank all the partners in the project for their contributions in some of the issues here reported.

References [1] R. Picard: Affective Computing, MIT Press, 1997. [2] D. Preston and F.B.M. de Waal: Empathy: Its ultimate and proximate bases, in Behavioral and Brain Sciences, 25(1) 2002. [3] B. Campos and I. Menezes: Affective Education in Portugal, in Affective Education, eds. P. Lang, Y. Katz and I. Menezes, Cassell, 1998. [4] A. Paiva and J. Dias and D. Sobral and R. Aylett and S. Woods and L. Hall and C. Zoll: Caring for Agents and Agents that Care: Building Empathic Relations with Synthetic Agents, in Proceeding os AAMAS 2004, ACM Press, 2004. [5] Eisenberg, N. and Strayer J.: Critical issues in the study of empathy, in Empathy and its development, Eds. Eisenberg, N. and Strayer, J., Cambridge: Cambridge Press, 1987. [6] E. Staub: “Part I Commentary: Historical and theorectical Perspectives” in Empathy and its development, Eds. Eisenberg, N. and Strayer, J., Cambridge: Cambridge Press, 1987. [7] L. Hall and S. Woods and K. Dautenhahn and D. Sobral and A. Paiva and D. Wolke and L. Newall: Designing Empathic Agents: Adults vs. Kids, in Intelligent Tutoring Systems, Springer, 2004. [8] Bischof-Köhler: D. Spiegelbild und Empathie – Die Anfange der sozialen Kognition, Bern: Hans Huber, 1989. [9] www.victec.org. [10] A. Paiva and João Dias and Daniel Sobral and Ruth Aylett and Sarah Woods and Lynne Hall and Carsten Zoll: Learning by Feeling: Evoking Empathy with Synthetic Characters, in Applied Artificial Intelligence Journal, vol. 19, n. 3–4, 2005. [11] www.immersiveeducation.com/uk/Kar2ouche_JISC.asp.

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

255

Using Machine-Learning Techniques to Recognize Emotions for On-Line Learning Systems Mohammed A. RAZEK, Soumaya CHAFFAR, Claude FRASSON and Magalie OCHS Département d’informatique et de recherche opérationnelle, Université de Montréal C.P. 6128, Succ. Centre-ville, Montréal, Québec, Canada H3C 3J7 {ochsmaga, abdelram, chaffars, frasson}@iro.umontreal.ca

Abstract. The effectiveness of intelligent tutoring systems, for instance on-line learning systems, can be improved when the learner’s emotions are taken into account. A necessary condition for this is that the system will be able to recognize the learner’s current emotional state. Traditional methods for doing this are based on measuring physical parameters, most typically the facial expression or muscle tension, however, they are neither comfortable for the user nor useful in a distributed environment such as the Internet. Furthermore, filling out a long questionnaire is a time-consuming task. In contrast, we present an extremely simple method that can be used instead, the Emotion Recognition Agent (ERA), which is devoted to exploit the natural relation between emotions and colors. We have performed experiments demonstrating both the simplicity and the accuracy of our ERA method which employs machine learning techniques for determining a user’s emotion given colors sequence. Keywords. Machine learning, ID3, emotions, colors, Web-based tutoring systems

1. Introduction Emotions are presented everywhere in all time in human life, both in our interpersonal interactions and in front of our computer. In the context of learning, for instance, learners worry, hope, get bored, get embarrassed, envy, get anxious, feel proud, and become frustrated. Their emotions will moreover highly influence their performance [Isen, 2000; Goleman, 1997]. If a tutor knows a learner’s emotion during a learning session, he/she will adapt his/her teaching methods in the way that improves the learner’s learning capacities [Hargreaves, 2000; Picard, 1997]. This shows the importance of the recognition of a learner’s emotion in an ITS. Currently, most methods for determining the learner’s emotion, in particular in the context of an ITS, employ sophisticated systems based on physical sensors. However, these methods can recognize emotion from the facial expression; most of them need sophisticated technologies such as wearable sensors which consist of connecting electrodes to the body of the subject aiming to measure his physiological signals [Cacioppo, 2000]. For example, Faivre et al. [Faivre, Nkambou & Frasson, 2002] has created an affective-tutor, which is affectively adapted to the learner’s emotion.

256

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

Such methods are, however, impractical in the context of online systems. Another way of assessing the user’s emotions is a questionnaire in which he reports his emotions. Yet, it is not always easy to determine one’s own emotions. In this paper we propose a method that is even easier and faster but at the same time reliable. It is also a considerable way to be used on the web. Our method is based on the relation between colors and emotions. In fact, there exists a natural way of representing of emotions by colors [Black, 2002]. For instance, when we are happy we commonly say that we see life through rose-colored spectacles. Yellow has been found to be associated with both sadness [Peretti, 1974] and with cheerfulness [Wexner, 1954]. Red is related to anger and violence [Sch, 43]. However, cool colors are considered to have calming effects such as: sea green, violet, blue, light blue and cyan [Levy, 1984]. Moreover, blue is related to tendress [Schachtel, 1943] and sadness [Peretti, 1974]. While, neutral colors are considered to have a neutral affect like black, white, grey and brown. Based on such connections we have created an online friendly agent, the Emotion Recognition Agent (ERA), which is able to predict a user’s emotion from his choice of a sequence of three colors; the use of this agent can replace online questionnaires. This paper is organized as follows. Section 2 briefly describes some related work. In Section 3, we present in detail the architecture of ERA. Section 4 shows the experiments we conducted and how we interpret and apply their results. Finally, Section 6 discusses pending problems and suggests future projects.

2. Related Work Many researchers have developed methods for determining the emotional state of an individual. One of the most natural approaches is to use the various physical manifestations of emotions: the facial expression, physical parameters such as heart rate, blood pressure, respiration frequency, voice tonality [Murray & Arnott, 1996], and skin conductivity [Picard et al., 2000; John et al., 2003]. These methods, which allow for distinguishing the six basic emotions defined in [Ekman, 1993], all have in common that they use sophisticated technologies and sensors. Hence, they suffer from two problems: First, people do not like when their emotions are measured with physical sensors; secondly, the methods are useless in the context of online systems. In the same sense, other researchers have proposed systems based on text analysis. Their application has, however, been limited to the context of online chatting [Boucouvales & Zhe, 2002; Lieberman et al., 2003]. The latter consists of analyzing the sentence written by the subject using natural language processing techniques, to determine whether he is happy or sad. This method is reliable in restricted online context like chatting but isn’t useful for other environments or contexts because users, in general, don’t like to write messages. As we mentioned above, the colors are useful to represent emotions. For example, the color red is related to anger, whereas cool colors such as green, blue, or purple are associated rather to unexcited emotional states [Schachtel, 1943]. It is important to note, however, that some colors can have different and even opposite emotions related to them; for instance, yellow and blue are associated with both sadness and cheerfulness. This ambiguity is the reason for our approach which is based not only in a single color, but a sequence of three colors.

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

257

Figure 1. The Architecture of ERA.

3. The Emotion Recognition Agent (ERA) Architecture We have constructed an intelligent agent, the Emotion Recognition Agent, able to associate to a sequence of colors an emotion. This association is made by ID3 machine learning methods (see Section 5). The ERA is designed as a generic process for Webbased tutoring systems. ERA is able to recognize learners’ emotions using their ordered choice of three colors. The architecture of ERA consists of three tiers: User interface, application server, and data base (see Fig. 1). The user interface tier is where the learner’s services (such as the colors chosen) reside. The application server tier provides the ERA’s process management services (such as process data base and process ID3 tree). The third tier provides the database management functionality and is dedicated to data that are acquired during the experiments (for more details about them, see Section 4). When a user connects to the web-based system, he will interact with the animated Emotion Recognition Agent through the user interface. The Emotion Recognition Agent asks him to choose a sequence of colors representing his current emotions by clicking onto colored bubbles in a certain order (see Fig. 2). The Emotion Recognition Agent then determines the emotional state of the user by connecting to the database and searching for the emotion associated to this sequence of colors. In the following section, we show how the experiments were conducted. 4. Experiment and Results 4.1. Experiment Our objective in these experiments was to determine whether and emotions can be associated with colors, and, if so, which color corresponds to which emotion. More precisely, we try to find out how people use colors to express their emotions.

258

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

Figure 2. Screenshot of the User Interface.

Several researchers have described categories of emotions with the objective of classifying different emotional states. We use in our study the classification of Ekman [Ekman, 1992] (see also [Picard, 1997]) which is based on facial expressions and distinguishes six basic emotions: happiness, anger, sadness, surprise, fear, disgust. We have added the category normal, which corresponds to the neutral state of lacking any of the six basics emotions. On the other hand, we have chosen eleven colors: black, purple, brown, khaki, orange, pink, green, gray, and the three primary colors red, yellow, and blue. In order to determine the links between these emotions and colors, we have set up a Web site with a test where people are asked to give their description of emotions by colors. For the sake of attractiveness and user-friendliness, we have created an interactive game, using the software Flash. For the user, the goal of the game is to create a room according to his emotions (Fig. 3). As a first step, the user indicates how he feels at the moment by choosing one of the seven basic emotions (Fig. 4). In the second step, we ask the user to express his current emotion by a color. He can choose the color corresponding to his emotion among the eleven colors listed above. In the third step, the user is asked to express each one of the basic emotions by a color (Fig. 5). This step allows us to verify the consistency of the user’s responses: We check if the color chosen to express his current emotion in the preceding step has been the same as the one used to describe the same emotion in this step; if not, we do not take into account this user’s responses.

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

Figure 3. The Beginning of The Game.

Figure 4. Giving the Current Emotion.

259

260

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

Figure 5. Associating Colors to Emotion.

Figure 6. Painting the Room.

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

261

Figure 7. The User’s Emotion’s Room.

In a last step, we analyze the relation between an ordered sequence of three colors and an emotion: the user chooses three colors to paint, in a certain order; the three parts of a room (see Fig. 6). Finally, the user is shown the room colored according to his emotion and to his responses in each step. The parts of the room are colored by the colors chosen at the different steps, the facial expression of the character on the screen and the picture on the wall of the room show the user’s emotion (see Fig. 7). Before quitting the game, each user was asked to give his comments on the game as well as his opinion on whether there exists a relation between emotions and colors. The set of entries and choices of the user (emotion, colors, sequence of colors, opinion) is then stored in a MySQL database. 4.2. Results The users’ feedbacks of the two experiments have been accumulated through a period of two months. Table 1 represents the number of the users who selected a specific emotion-color pair. There were 322 participants in the experiments. The question whether there is a connection between emotions and colors was answered by 58% of the people in an affirmative way, whereas 22% did not agree and 20% gave no comment. Figure 8 shows the relation between colors and emotions. For example: we found that more than 27% of the participants prefer blue color to represent Normal emotion, however about 25% of the participants choose yellow for expressing joy and nearly 22% prefer the same color to represent surprise. There are also more than 41% of the participants who opt for red color to represent the emotion of anger. For expressing

262

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

Table 1. A Collection of Users’ Feedback. Normal

Happy

Anger

Surprise

Disgust

Sad

Fear

green

28

44

8

63

22

9

18

red

7

30

134

31

3

10

34

blue

87

38

3

19

8

16

16

pink

33

54

0

35

12

8

5

khaki

31

8

8

5

32

18

19

orange 21

21

13

23

21

8

20

yellow 21

81

7

70

16

9

12

purple

25

6

18

27

25

19

38

brown

11

2

12

2

76

22

14

black

7

5

57

5

26

62

72

gray

18

1

22

2

36

102

29

users numbers

Emotions and Colors 160

green

140

red

120

blue

100

pink

80

khaki

60

orange

40

yellow

20

purple

0

brown Normal

happy

anger

surprise disgust Emotions

sad

fear

black gray

Figure 8. Learner’s feedback.

disgust, nearly 24% of the participants select brown, and more than 31% of the participants prefer gray color to express the emotion of sadness and about 22% of the participants choose black color to represent fear. In the second experiment, we have studied the relation between a sequence of three colors and an emotion. During the experiment, participants have associated color sequences with emotions. The resulting database was then used to solve the problem of estimating an emotion on the basis of a new sequence. To do that, we have used the ID3 machine leaning method to construct rules based on the pair of sequences of colors and emotions collected during the experiment. In the next section, we explain the theoretical and practical tools involved to finish this task.

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

263

5. Experiment Result and Implementation The ID3 algorithm [Luger 2002] induces concepts from examples. It represents concepts as decision trees, a representation that allows our system to determine the classification of emotions by testing its values for certain properties. ID3 measures the information gained for each property and then selects the one which gains the greatest information. Following Shannon’s method [Luger 2002], we define the amount of information in each color sequence as a function of the probability of occurrence of each possible choice. The information gain provided by making the experiment at the root of our goal tree is equal to the total information in the tree minus the amount of information needed to complete the classification after performing the experiment. The amount of information needed to complete the tree is defined as the weighted average by multiplying the information content of each sub-tree by the percentage of the examples present in that sub-tree and summing these products. Following Razek [Razek, Frasson, & Kaltenbach], given our set of training instances U, if we make property P, with 11 values, the root of our goal tree, this will partition U into subsets, {Black, Yellow, Green, Purple, Brown, Blue, Khaki, Red, Orange, Pink, Gray}. The expected information needed to complete the tree after making P the root is: 11 ⎡7 ⎤ E(p) = ∑ P(U j ) ⎢∑ − l (mi ) log 2 (l (mi ))⎥ j =1 ⎣ i =1 ⎦

(1)

Where •

P (U j ) Represents the probability for the occurrence of a certain color in

•

j th choice of the eight choices. l (mi ) The probability for the occurrence of each emotion mi .

The gain from property P is computed as: 11

Gain( p) = ∑ − l (U i ) log 2 (l (U i )) − E ( p)

(2)

i =1

As shown in Fig. 9, The first level is a start point, the second level represents the first color could be chosen by the user from a set of eleven colors, the third level represents the second color could be chosen by the user from the same set, the fourth level identifies the third color could be chosen and the fifth level identifies the emotion associated to the sequence of colors chosen by the user. For example, one user have chosen blue as the first color, than the second color is yellow after that he has chosen purple as the third color. For this sequence of color (blue, yellow, purple), the predicted emotion is surprise and so on. The decision tree provided by the ID3 method, allow us to determine a set of rule which are used to solve new instances. The first choice of color provides the greatest information gain. Therefore, ID3 will select it as the root of the tree. The algorithm continues to apply this analysis recursively to each sub-tree until it has completed the tree. Figure 9 shows the decision tree solution of our experiment by which the system

264

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

Figure 9. Example of Decision Tree.

can predict the emotion of a new learner based on a choice of a color sequence reflecting his current emotional state.

6. Conclusion and Future Work We have shown that a sequence of colors is related to a specific emotion in the following way: By giving a sequence of colors, a person can express his emotion reliably. It is therefore possible to determine someone’s emotion through his ordered choice of three colors with 57,6% accuracy, a fact that we use in a very simple interactive system : the Emotion Recognition Agent (ERA). Our proposed method constitutes the simplest and fastest reliable online method for assessing the emotional state of a user. Determining someone’s emotional state is of paramount importance for example in the context of online learning systems, because it offers them the possibility to adapt to the user’s emotions with the objective of improving his performance. In fact, the recognition of somebody’s emotions is known to be one of the central features of, and even a necessary condition for, Emotional Intelligence. Therefore, our system can be seen as a first step towards realizing online tutoring systems that are emotionally intelligent and can use this ability for the sake of improved learning efficiency.

References [Black, 2002] Black, J. Color Psychology for E-Book Cover Design, 2002. [Boucouvalas, 2002] Boucouvalas, A.C. Zhe, X. 2002. Text-to-Emotion Engine for Real Time Internet Communication. In International Symposium on CSNDSP, Staffordshire University, July 15–17, pp. 164–168. [Cacioppo, 2000] Cacioppo, G.G. Berntson, J.T. Larsen, J.T. Poehlmann, K.M. Ito, T.A. “The psychophysiology of emotion. Handbook of emotions”. New York: Guilford, pp. 173–191, 2000. [Levy, 1984] Levy, B.I. Research into the psychological meaning of color. American Journal of Art Therapy, 23, 58–62, 1984. [Peretti, 1974], P.O. Color model associations in young adults. Perceptual and Motor Skills, 39, 715–718, 1974.

M.A. Razek et al. / Using Machine-Learning Techniques to Recognize Emotions

265

[Razek, 2002] Razek A.M., Frasson C., Kaltenbach M. Using Machine Learning approach To Support Intelligent Collaborative Multi-Agent System. International Conference on Technology of Information and Communication in Education for engineering and industry, TICE 2002, 13,14, 15 November 2002 LYON, France, 2002. [Schachtel, 1943] E.J. On color and affect. Psychiatry, 6, 393–409, 1943. [Wexner, 1954] L.B. The degree to which colors are associated with mood-tones. J. appl. Psychol. 38: 432–435, 1954. [Eckman, 1993 ] Eckman, P. 1993. Facial Expression and Emotion. American psychologist. 48, 384–392. [Ekman, 1992] Ekman P. 1992. An argument for basic emotions. Cognition and Emotion, 6(3/4):169–200. [Faivre, 2002] Faivre J., Frasson C., Nkambou R. Gestion Émotionnel dans les Systèmes Tuteurs Intelligents. Technologies de l’Information et de la Communication dans les Enseignements d’ingénieurs et dans l’industrie. 2002. Goleman D. 1997. L’intelligence Émotionnelle 1. Edition Robert Laffont, 1997. [Hargreaves, 2000] Hargreaves. 2000. Mixed emotions: teacher’s perceptions of their interactions with students. [Isen 2000] Isen A.M. 2000. Positive Affect and Decision making – Handbook of emotions, second edition, Guilford Press. [John, 2003] John, D. Boucouvalas, A.C. Zhe, X. 2003. Expressive Image Generator for an Emotion Extraction Engine. 17th HCI 2003: University of Bath, UK. [Lieberman, 2003] Lieberman, H. Hugo, L. Selker, T. 2003. A Model of Textual Affect Sensing using RealWorld Knowledge. In International conference on Intelligent User Interface. [Luger, 2002] Luger F., Stubblefieled W. 2002. “Artificial Intelligence: Structures and Strategies for Complex Problem Solving” Addison Wesley, 3rd Ed. [Murray, 1996] Murray, I.R. Arnott, J.L. 1996. Synthesizing emotions in speech: Is it time to get excited? In Proc. ICSLP pp. 1816–1819. [Picard 2000] Picard, W.R Healy, J. 2000. SmartCar: Detcting Driver Stress. In Proceedings of ICPR’00, Barcelona, Spain. [Picard, 1997] Picard. eds. 1997. Affective computing. MIT Press. [Schachtel, 1943] Schachtel, E.J. 1943. On color and affect. Psychiatry, 6, 393–409.

266

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

A Framework for Emotional Agents as Tutoring Entities Bogdan Florin MARIN, Axel HUNGER and Stefan WERNER University of Duisburg-Essen, Germany {bmarin|hunger|swerner}@uni-duisburg.de

Abstract. Current distance education systems try to mitigate the difficulties encountered by learners when they try to follow a distance course. The use of emotional animated agents in such environments as a tutoring paradigm can be benefic and increase the learners’ motivation. This paper discusses the premises under which emotional agents can be pedagogically effective as tutors in a collaborative learning environment. This work highlights a framework for creating believable synthetic agents. Keywords. Emotional agents, human computer interaction

1. Introduction One of the most successful applications of life-like characters is computer based learning environments where synthetic agents can perform a variety of roles especially as tutors or trainers [1–3]. The use of animated agents in such environments as a tutoring paradigm can be benefic and increase the learners’ motivation. Lester [4] investigates the impact of animated agents along the dimensions of motivation and helpfulness in an interactive learning environment. Current distance education systems try to mitigate the difficulties encountered by learners when they try to follow a distance course. Then, it is necessary to take account of these difficulties when distance learning is set up, avoiding insulation and a lost of motivation by learners that are the cause of many giving up [18]. Several major inconveniences can be noticed in distance education like laboratory experimentation/ practice. Usually, during these experimentations students have to be physically present in the university laboratories. A solution to avoid this disadvantage is virtual experimentation: the experiments are simulated and visualized by means of virtual reality [19]. In local laboratory experiments, students usually work together in groups of two or more. This learning paradigm is often called collaborative learning. One solution for this problem is usage of virtual collaborative environments which bring together users who are geographically distributed but connected via a network. Therefore the students can be trained using the virtual lab concept to work in spatially distributed teams. Within the past years a synchronous groupware named PASSENGER was developed for this purpose at University of Duisburg-Essen. This groupware is composed from modules: the communication component which contains video screens of each participant and a cooperation component which allows students to interact together on

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

267

a common artefact. The participants can be: three students and one tutor. A more detailed description of the Passenger environment is provided within the next section. We integrated animated-agents in Passenger in order to overcome the lack of tutor when the students meet outside tutor working’s hours. We consider agents in a virtual society of learners as entities that occupy a social position and perform several roles. To exemplify the concept in our environment the tutor-agent should have besides the tutoring-roles like: interrogator, reviewer, monitor, or instructor which are later described. Our approach aims at showing the premises under which animated agents can be pedagogically effective tutors in distance learning environments. This paper shows the design principles for animated agents to enhance their effectiveness as tutors in a synchronous learning environment. Within this paper we propose a solution for other inconveniences of collaborative environments like awareness or communication. A partially solution for the Passengerparticipants’ awareness can be also found in [20]. The remainder of this chapter is structured as follows: next section describes the Passenger Learning Environment, Section 3 highlights the state of the art in this research field, Section 4 presents a detailed analysis of the requirements of a life-like character while the following three sections highlight: the roles, a conceptual model and the architecture of an emotional tutor agent. Last two sections provide the evaluation studies and the conclusions of our work.

2. Passenger Learning Environment A synchronous groupware named PASSENGER was developed at the University of Duisburg – Essen (UDE) throughout the last years. The system is designed for spatially distributed collaborative working and especially for the usage in Software Engineering – Education. A client-/server architecture has been chosen whereby the server is located at the university, due to the fact that the university plays a major role in this configuration. The specified requirements for a groupware used in a software-engineering-lab give a direction how the function- and application-classes of the Passenger-Client were defined. Thus the following points were included to the concept: • • •

The communication components for the synchronous, audio/visual communication. The cooperation components, respectively the common work area. The network components for a reliable data transmission over the Internet.

Therefore the Passenger-Client consists of a communication component, a cooperation component and several shared tools and resources to carry out Software Engineering tasks. The Passenger Client user interface (see Fig. 1) contains video screens of each member and a whiteboard that is divided in a public window for common process on the outline documents and a private window for individual process on the documents. The creative tasks within the collaborative group processes, respectively the modification of the common artifacts take place in the common work area. The PublicWindow area of the Whiteboard is managed by the Floor-Control [25].

268

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

Figure 1. Passenger Learning Environment.

Each member has the same view of the public window according to the What You See Is What I See (WYSIWIS) principle, but only one of them can alter the document at a certain time. Each member is also equipped with a private working window to try out own ideas and to work simultaneous on an individual solution. A Telepointer serves to elucidate and present the facts. The Telepointer is implemented as a collaborative service which can be used by the Floor-Holder. This user can lead the attention of the other participants to some objects or screen areas of the public window during a discussion. The Passenger Floor Control handles the access to shared resources and coordinates the course of communication via the administration of different kinds of permissions, e.g. permissions to speak, permissions to alter the documents etc. A floor control, which meets the requirements of the fairness and the avoidance of mutual exclusion, would be easier to fulfill as more knowledge about the parameters describing the group process are present. Closed groups offer the advantage that their number of participants is constant and thus ascertain fairness and allow examinable measured value to be determined. For further considerations, there will be a group of three students and a tutor, as an example, taken for granted. Next paragraphs will present the concept of fairness and how this concept was implemented in Passenger. The fairness of a system can be realizes in various ways. A technology-oriented solution would be assigned to all students in a discussion and work cycle for the same duration of speaking time [34]. These solutions can be misconceived as unfairness, i.e. if the current Floor-Holder was not able to conclude his work and accomplishments within the given period. Likewise it is also possible that a Floor-Holder does not have any contributions. In this case the working time would elapse unused.

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

269

An essential part of our floor control implementation is the permission list. This list is implemented on the server and has three possible entries. The actual spokesperson is followed by the next two clients, who have requested to speak. Each request for the permissions as well as each passing of the permissions results in an update of the permission list. The list is implemented in such a way, that neither a participant can be excluded from the session nor the course of discussion can end up in a blocking state. Therefore, the model guarantees a defined fairness by means of each student having an entry in the permission list or at least having the option of placing an entry in the permission list. Due to this reason for the synchronous groupware Passenger the use of time duration was not taken into consideration and was designed to be more problem oriented. This provides the possibility that each student is able to request the Floor and to receive the Floor within a discussion and working cycle. This also corresponds to the observed role behavior; that a passive participant would probably not make use of his right each cycle. On the other hand passive participants are not to be excluded from the discussion. For this, a simple but efficient procedure was implemented. In this procedure a work cycle is defined such that everyone of the n participants of a session has the possibility of taking over the Floor once. In order to prevent blockings, the takeover of the Floor is not explicitly demanded by everyone the n participants. If n–1 participants had once possessed the Floor and up to the time tn–1 the Floor is released, then the last participant would lose his right to hold the Floor when he failed to request for the Floor at time tn–1. In this case a new work cycle begins at the time tn–1. Another important point in our groupware design is the definition and implementation of suitable methods for group-awareness [26]. Their implementation is done by visualizations in the communication windows and as well by a common context. In the next paragraph a description of the visualizations within communication windows is given, followed by a description of the common context. For the aspects of group awareness, each participant is always placed into the same video screen. Each participant appears always in the left corner of its own screen. The Tutor is always in the right corner. All other participants are associated to a fixed position with the help of an ID-number assigned by the Passenger Server. If a participant is missing, the associated position is empty. Furthermore, the video screens can neither be changed in size nor in position and none of the video screens can be covered by another one. This is to avoid that one specific member, conscious or unconscious, could be given visual priority over the others. Also, no video screen can be covered by other elements (e.g. shortcuts) or other windows of the desktop. This should avoid that some participants can have a lower position within the group in comparison to other participants. Group awareness functions are implemented by means of providing all needed information for a late coming in participant to discover the actual conference state. This is implemented by highlighting the video screen of the person who has access to the shared resources. Also the Passenger Client interface (see Fig. 1) contains a series of buttons to support basic modeling functions of Ward & Mellor [24] (e.g. draw a control transformation) or standard functions like cut, copy, paste. The client interface has also a function which permits a member to transfer the content of the public window into its own private window. The handling of the above mentioned buttons is closely oriented to other

270

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

usual applications and will not be described any further. A more detailed description of Passenger groupware can be found in [20]. The Software-Engineering lab at UDE is conducted as a project setup of student teams, each consisting of four participants: 3 students and one tutor, where the same tutor can be in several virtual teams. That can cause problems in terms of availability if the virtual teams meet at the same time but also if the teams meet at times outside the tutor consultation hours. To make sure that at least a virtual tutor is always available agent technology shall be used. The agent can be noticed in the right-upper corner of Fig. 1. Therefore, within the next sections, a novel idea and an intelligent tutor-agent’s architecture are presented.

3. Related Work The common research trend in designing animated agents is to make the life-like or believable [9]. Life-likeness is supposed to provide the user with the illusion of life and believability should allow users to suspend their disbelief. Due to the fact that characters can be life-like in a “human-like” or an “animal-like” way an ongoing debate concerning whether the life-likeness of characters is more effectively by a realistic or by a cartoon style agents. The answer to this debate can eventually be given empirically with respect to specific application scenario. For instance, while Blumberg in his thesis [10] conducts a series of investigations on animal like characters, especially dogs, Thalmann [11] aims to create virtual humans typically following the realistic approach, even strives for photorealism. In her thesis [12] Koda created a Web-based poker game in which a human user could compete with other personified computer characters including a realistic image, cartoon male, female characters, smiley face, no face and a dog. She gathered people’s impressions of the characters and she discovered that people’s impressions of a character were different in a task context than in isolation and were strongly influenced by perceived agent competence. In his work Lester [4] provides the result of his investigation on the impact of animated agents along the dimensions of motivations and helpfulness in an interactive learning environment. He coins the notion of ‘persona effect’ as “[…] the presence of a life-like character in an interactive learning environment – even one that is not expressive – can have a strong positive effect on student’s perception of their learning experiences”. The work of Nass & Reeves [13–15] focuses on the study of computers as social actors. They have conducted a series of experiments that examined how people react to computer systems and the applications that incorporate certain personified characteristics. They demonstrated that users like computers more when the computer flatters them [13]. Their most important discovery is that people interact with and characterize computers in a social manner, similar as they do with other people. More exactly, they found that existing, accepted sociological principles (e.g. participants with similar personalities tend to get alone better that do those with different personalities) apply even in the case when one of the participants is a machine.

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

271

There have been a number of efforts to introduce agents into learning environments in order to create better and more human-like support for exploratory learning, and social events that support tutor-tutee interactions and collaborative learning [27,28]. According to Aroyo and Kommers [29], agents can influence different aspects in educational systems. They supply new educational paradigms, support theories and can be very helpful both for learners and for teachers in the task of computer-aided learning. Guizzardi et al. [30] investigate the nonhierarchical relationship between teachers and students in an environment where everyone can teach and learn. They gather two perspectives: one from an implementation point of view and the other one from a software engineering perspective and propose an agent-based system to support extra-class discussions between students and teachers. According to Kay [31], in the first computer-assisted teaching environments the idea was to build “teachers” who could transmit knowledge to the learners. Currently, these types of environments are more geared up for exploration on the part of the learners, designing, building and using adaptive systems as tools. These environments also are being built to give greater responsibility to the learners regarding aspects of the learning process, and especially regarding control of its model, which is the central aspect in the adaptability of the tools. Finally, in [16] Prendinger and Ishizuka present a survey on life like characters. They gather a lot of papers concerning XML-based agent behaviour description languages which will expose readers to some practical designs they can choose if they want to implement such characters. Other papers covers how to model social or psychological logic or behaviours, for example, how to model a character’s expectations to drive learning.

4. Design Requirements Analysis for Emotional-Animated Agents Since human-human communication is an effective and efficient way of communications, life-like characters are promising candidates to improve HCI. Synthetic agents may use multiple modalities such as voice, gestures and facial expression to interact with humans. A related empirical question concerns the benefits of displaying life-like characters as facial agents (talking heads), full-body, or upper-body plus hands agents. Unfortunately, there is no predefined standard for designing synthetic agents due to the fact that the application context decides, by and large, what characteristics an animated agent should have. Still there are several major aspects one should keep in mind when designing animated agents: •

•

The look: is the agent going to represent a person or some other living creatures (e.g. animals)? Is the agent meant to be realistic (human-like character) or it is artistic (may be an exaggerated cartoon-like)? In case of a human-like character should the agent have a gender: male or female, if yes why? Physical details: what parts of the body are covered by the agent: head, torso or full body?

272

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

Figure 2. Generic architecture for a synthetic agent model.

• • •

Communication modality: how should the agent communicate with the human user(s) from its environment? Should the agent have the ability to use speech or only text? Mental model: should the agent have its own mental model? Should the agent posses the ability to express emotions? Should the agent have its own personality or not? Interacting with the agent: how does the user interacts with the agent? Who is able to control the agent: the user directly (e.g. avatars), some applications (e.g. presentation agents) or both (e.g. tutoring agents)?

Furthermore, the answers to the above questions are strongly dependent on the application context which is the major decision factor in the design process of such agent. Even though there are several requirements which every synthetic agent must satisfy: • • • • • •

•

A graphical representation: it is a fundamental requirement for every life-like character Awareness: it refers to the ability of the agent to perceive its environment and the changes that occur in it. Communication: ability to communicate with users or other agents through different ways using different communications protocols Roles: a role defines those behaviors, characteristic of one or more persons in a context [17]. In our case each agent should play different roles that are associated with different rights in different situation to enforce. Goals: a goal defines a special state that agent is pursuing to achieve in a current situation. There can be two types of goals: collaborative goals in case of multi-agent system, or individual goals for itself. Reactivity: ability to respond quickly to environment’s changes. This ability is generally defined as a series of event-condition-action rules like: ON event IF conditions THEN action which means: when event occurs and if the evaluated conditions are true then action in rule are executed. Planning: ability to elaborate an action sequence to achieve their goal in a situation.

Based on careful analysis of the above requirements a generic architecture can be proposed for synthetic agents (see Fig. 2). The model is composed from two components: the body part or the graphical representation of the agent in its environment and the brain part – the part which is responsible for all agent’s actions. We chose a human character instead of an animal in order to impose learners a degree of realism due to the fact that the environment where our prototype acts is virtual environment for distance education system. Next section explains in detail those two parts and the connections between them.

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

273

Figure 3. BNF specification of Role class.

5. Roles of an Emotional Tutor Agent A major problem in this research is not the design of a tutor agent but to determine how this agent can self-improve its instructional plan with respect to the learners’ group organization and also to the needs of different categories of learners. Categorization, instructional planning and the ability of agent-tutor to self improve are significant issues in the development of such tutoring system. In order to tackle these issues we propose a conceptual model based on role theory. We view agents in a virtual society of learners as entities that occupy a social position and perform several roles. Giddens [32] defines social position within a group as the social identity an individual has in a given group or society. Biddle [33] defines roles as those behaviours, characteristic of one or more persons in a context. In our case a role specifies a characteristics pattern of behaviour for the interactions of the agent so that the agent plays that role behaves in a specific way under certain situations involving other learner(s) or agents. Figure 3 shows the Role class written in BNF specification. Role ID is used to distinguish a role from other roles. A skill can be defined as the ability to carry out a task at a pre-defined level of competence. In our concept Skills of a role describe the properties (or the abilities) that the agent will need to possess in order to perform successfully the role. Skills should be linked together with roles: if an agent knows what role it has to play then it also knows the skill(s) required to successfully accomplish that role. As far as the procedures (the agent is developed in Borland Delphi, and a class in Delphi has procedures similar with the methods from a Java class) of the Role class there are a group of procedures which associate Role class with skill class: • • •

TAddSkill() – this procedure is used to bind a particular role with a particular skill TRemoveSkill() – destroys the link created by TAddSkill TGetSkill() – returns all skills relevant to the role under consideration.

Roleset refers to a set of roles that agent interacts with given this role. Prerequisites of the role refer to the credentials an agent needs in order to occupy the social position under that role. Responsibilities of a role refer to the duties of an agent undertaken within the context of the actual role.

274

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

To exemplify the concept in our environment the tutor-agent should have the following roles within a group of students: • • • • •

Interrogator – poses questions and the students of a collaborative group then provide answers. The questions should provide help for the students to reach a common learning goal. Reviewer – analyzes the students’ answers, including whether it is correct or not. Monitor – records the answers from all the students and the communications among students during the collaborative learning process. Instructor – gives individualized instructions and helps those students who cannot keep up with the progress of their group-mates. Group Manager – has the ability to control the coherence of the group.

Let’s take for example the Reviewer role: the skills required for this role are: agent should be able to understand the student’s answer (natural language processing) then it should analyze whether it is correct or not. An important skill of this role can be considered the ability to display emotions and gestures (animations) to students’ answers. Our prototype responds to the students’ answers (also to questions) by synthetic speech, facial display and gestures. The facial display of our tutor –agent (see Fig. 1) is limited to a predefined set of animations (e.g. happy, sad). In order to extend the animations for our model we implemented also gestures to express emotions like confused: agent is lifting shoulders or don’t recognize the answer/ or question (this skill is also available for other roles – Instructor): put a hand to ear. The Roleset for this role can be considered as the set composed from the roles = {Interrogator, Instructor}. Of course the Prerequisites for this role can be simply deducted: if the students answer to the question posed by agent (during the Interrogator role – here is also the link between these 2 roles: Interrogator and Reviewer). In the case of the Monitor role the agent needs the skill to create a student profile [8] from the students’ interaction. The student profile includes student’s goals, plans, capabilities, attitudes and knowledge. This profile is based on student’s activity during a learning session. Let’s consider a trait called Activity_Level which specifies level of a student activity during a session, as a numerical integer value from the interval [–5, 5]. –5 means that the student is very lazy or he or she is not interested in participating in the common learning, 0 that our student is neither lazy nor energetic, and 5 to define that the student is very active, energetic. The default value is 0 and it is assigned by the agent to each student at the beginning of a new Passenger learning session. This value can be incremented or decremented based on student’s behaviour during the learning process in the following manner: if the student has the initiative and takes the floor from the agent and performs successfully a sub task his activity level is increased by one, if he/she performs the sub task guided by agent without being able to make own decision the activity level is not modified. The activity level is decreased only when the student refuses to perform its task. Of course the most active student within a learning session can be considered as a leader for the group. In case that the final value of this trait is 0 this can characterize either a lazy or a cautious learner. The tutor-agent has the responsibility to deliver after each session a report with the students’ profile and its own beliefs and conclusions to a human tutor. This profile is based on several parameters like Activity_Level and the human teacher is capable to

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

275

deduct student’s behaviour during a session. Let’s assume that after a successful session the levels of activity are 5, 0, –2 for Jan, Mary and Robert. The human teacher can deduct the following things: Assessing collaboration: it is obvious that Jan dominated all the phases of the activity, Mary did everything under tutor supervision and Robert didn’t want to participate. Assessing contribution: Jan did almost everything while Robert did nothing. Based on the activity level agent can regulate its future teaching strategies and also create a report with its own mental beliefs and actions during a learning session: Selecting tasks(subtasks) that need focus, a more detailed analysis or explanation: in case the students reached a deadlock the agent should be able to provide more hints, detailed theoretical demonstrations or other similar examples which can help students to overcome the current situation. Assessing its own interventions: agent should be able to decide whether to intervene or not (students must learn to work in a team and as long as they make progresses do not intervene) and choose the right time when to intervene in the learning process. Planning the group structure: decide based on the students profile whether the group structure is the optimal one or not. In our example Jan is too active for his group maybe moving him into another group and bringing someone else into his place can motivate also Mary and Roberts to involve more in the learning process. Group Manager – controls the coherence of the group. For this role agent’s requirements are: to monitor the owners of the Passenger Floor Control (PFC), to control the PFC list: grant/take the PFC to/(from) the inactive users. Also it must assure a fair distribution of PFC among participants in a learning session. This role and the afferent skills try to solve one of the open problems in the collaborative virtual environments: communication issues among participants.

6. Conceptual Framework for a Life-Like Tutoring Agent The success of a life-like character in terms of user appreciation depends on factors like characters’ role, competence and communicative skills relative to an application and its ability to present itself as a believable virtual personality. Our model includes the following concepts: personality, emotions, sensations and attitudes. Personality is a pattern of behavioral, temperamental, emotional, and mental traits that distinguish people from one another. Personality has been studied by psychologists interested in the behavior and in mental processes [21]. Enhancing personality and dynamic behavior to pedagogical agents can develop a new social-psychological model for animated tutoring agents similar with a human one. In particular humans can easily adjust their behavior based on their role in a socio-organizational setting, where their actions tend to be driven emotions, attitudes and personality. Personality traits correspond to patterns of behavior and modes of thinking that determine a person’s adjustment to the environment [21]. Traits are basic tendencies that remain stable across the life span, but characteristic behavior can change trough adaptive processes. Trait theories assume that an individual’s personality profile can be described in terms of psychological traits that influence that person’s behavior. In other words, it is assumed that traits predispose people to behave consistently, indifferent to the situation. Thus the personality profile can be used to predict future behaviors.

276

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

Opposite to trait theories are social learning theories which assume that a personality is modified by each situation viewed as a learning experience. A person behavior may vary depending on the specific characteristics of the situation in interaction with the individual’s appraisal of the situation and reinforcement history [22]. Our agent’s personality conceptual model uses both of these theories. As an example of trait theories in our model let’s consider the previous trait: Activity_Level. Moffat [6] highlights the close relationship between personality and emotion, although they seem very different: emotions are short-lived and focused while personality is stable and global. He also considers mood rather short-live like emotion and not focused like personality. As a product of evolution, emotions have a particular purpose: they have helped humans become the most successful species on earth. Emotions bypass the need for deliberative thought by providing biases toward the behaviors with better chances of survival—short-circuiting time-wasting rationalization. Other kinds of mammals also exhibit emotional capabilities with very similar reactions to humans. Psychoevolutionary scientist Robert Plutchik shares such theories [23]. According to Plutchik, there are eight primary emotions—associated in complementary pairs: anticipation and surprise, joy and sorrow, acceptance and disgust, fear and anger. These primary emotions can be observed in varied intensities (for instance, rage, anger, annoyance, terror, fear, apprehension). His theory states that it’s not possible for humans to experience two complementary emotions at the same time; they balance out to provide diversity in the behaviors. Though, primary emotions can combine together into complex moods; acceptance and joy can be understood as love, fear and acceptance lead to submission, sadness and surprise form disappointment, and so forth. In psychoevolutionary terms, each emotion serves its purpose by triggering a reactive behavior that’s appropriate for survival. Psychoevolutionary theory succeeds at explaining the reasons for emotions and provides a basic understanding of their roles as evolutionary tricks to improve survival rates. However, Plutchik’s approach fails to take into account the cognitive process associated with emotions. Elliott [7] defines emotions as valence reactions to events, agents’ actions, and objects, qualified by agents’ goals (what an agent desires), standards (what the agent considers acceptable), and attitudes (what agent considers appealing). Ekman’s [5] basic emotion approach distinguishes those emotions that have different facial expressions associated with them: fear, anger, sadness, happiness, disgust and surprise. Our approach investigates the reasons for emotions’ appearance: all emotions in embodied creatures are initiated by sensations. The notion of sensation can be defined as an immediate reaction to a creature’s current status. By definition, sensations are experienced practically based on changes in the current situation. Two factors may cause sensations: the current perceptions (that is, stimuli from the environment), or cognitive activity (that is, thinking). Considering humans sensations are typically triggered by perceptions: the body detects stimuli from the environment, and the information causes an immediate reaction in the brain. Applying this concept to our model sensations are triggered when a pattern is matched in the brain. With perception module, this pattern is matched instantly based on sensory information. On the other hand, some cognition is necessary before a pattern develops in the brain (by thinking), which eventually engenders a sensation instantly when a pattern is matched. For example, the tutor – agent may experience a

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

277

Figure 4. Emotions’ display: sadness and joy.

Figure 5. Animations for confuse and deny.

sensation of surprise when a student appears suddenly in the middle of a learning session. For instance, the same sensation of surprise can be also caused by a student not being present, when the agent thought it should be there. We represent emotions as our agent-prototype’s response to students’ questions by synthetic speech, facial display and gestures. Verbal and non-verbal behavior is synthesized in agent’s mental model and interpreted in a learning-session (Fig. 1). The facial display of our tutor –agent is limited to a predefined set of animations like happy, sad, etc. (Fig. 4). In order to extend the animations for our model we implemented also gestures to express emotions like confused: agent is lifting shoulders or don’t recognize the question: put a hand to mouth. Figure 5 shows confuse and deny animations samples. Another important concept included in our model represents the agent’s attitudes which characterize a relationship between an agent and a student. We defined attitudes as a complete set of emotions that constitutes someone’s mental state at a particular time. The attitudes included in our model are based on degree of sympathy and trust. Those attitudes are: like and dislike, trust or don’t trust. For example: if an agent likes a student it can offer him a second chance and hints in case that the student made a

278

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

mistake or doesn’t know the answer for a certain question. If the agent trusts a student it will give him the privilege to continue solving the problem in his/her manner even if the agent cannot foresee if the outcome will be the right solution or not. We represented attitudes as numerical integers with values from the interval [–5, 5], where above 0 means trust and bellow 0 means don’t trust. 0 value corresponds to indifferent state: agent has no attitude toward the student. The default value for each parameter is assigned to 0 at the beginning of each semester. How these parameters are modified are shown in the following example: the following scenario can be assumed: during a session the students are asked whether they need help/hints or not after taking more time on a topic than the time allocated. When the students refuse the help and choose a different solution/action than the suggested one, the agent-tutor records this behavior as indirect feedback. If that solution proves to be wrong this is materialized as decrease by 1 of the parameter’s value. In case that the solution is good the parameter will be increased by 1. After each session those values are stored in a students’ profile database. The attitude like or dislike is correlated with the trait which defines the activity of a student. The agent likes a student with a high level of activity and dislikes a lazy student. In a current running Passenger-session if the activity level of a student is already 5, and that student is the current floor holder or he/she has always the tendency to take action/initiative the agent tries to temperate the student by taking the floor from him/her and passes it explicitly to the inactive users. After each session the tutor-agent realizes an individual report for each student based on the student behavior during that session. The attitude of like is materialized as a bonus on the evaluation report: the agent recommends that the student should have a high mark at the final exam. In case of dislike the human tutor is informed of the negative behavior of a student during the learning sessions. The human tutor will investigate the reasons which conducted to such of a negative behavior and he will try to motivate the student to adjust his behavior to a proper one. The agent will be subject to four independent emotions: pity, hatred, attraction, and disgust. These are relatively easy to portray in the behaviors (especially in a learning session), and are sufficiently distinct from each other. Each emotion is recognized by a finite-state automaton. The automaton uses the data collected about the participants to decide what the agent’s emotions are. For example, attraction is triggered for students which agent likes and trusts; disgust instead is felt for particularly not-trusted and not-liked students. Instead of keeping each finite-state automaton (FSA) separate, these are grouped into one large nondeterministic FSA (NFSA) for convenience. Non-determinism allows the different automata to be merged together very simply, using (epsilon) ε transitions, as shown in Fig. 6. The FSA for the new emotion can be modeled separately, and grouped with the NFSA during the design. The emotions will mainly be used to generate new animated responses to students in different situations—for example, selecting which student to ask to continue the exercise. Technically speaking, all the concepts are modeled using finite-state techniques: emotions are expressed as a nondeterministic automaton, sensations as fuzzy automata, while attitudes as nested states.

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

279

Figure 6. Emotions recognition module.

7. Architecture of the Tutor-Agent Each agent involved in a Passenger learning session is assumed to have its own mental model. A mental model contains besides affective mental states like: emotions, personality, attitudes or goals, also different kinds of entities like global knowledge (plans, beliefs) or coaching knowledge (topics which will be presented during a session). Before going deeper into architectural layers of an agent we will describe the global knowledge. A plan is represented as a sequence of actions. Each plan may have preconditions and outcomes, and it is associated with a defined goal. An action can be seen as a set of propositional preconditions and effects, which are represented by probabilistic effects. Also to represent the success of action execution, each action has an execution probability. In a hierarchical plan representation an action might be primitive (e.g. an action directly executable by an agent) or complex. A complex action can be decomposed in multiple ways and each decomposition can be either a primitive action or a complex sub-action. In a plan structure a simple node consists of a primitive action while a decision node can be decomposed in multiple ways (primitive or complex actions) and the agent has to decide amongst those options. The options of a decision node are called action alternatives. Also a plan if it contains only primitive actions is called primitive plan. There might be more than one primitive plan available to achieve a goal therefore those optional plans are called plan alternatives. During the semester the student-teams will experience the entire life-cycle of Software Engineering. The students start with a requirement analysis following the Ward & Mellor [24] approach during the modeling phase. The given problem/task for the practical training is chosen in such a way, that it cannot be solved by one student on its own.

280

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

Therefore, each topic is divided in sub-topics which can be assigned by the tutor agent to one of the participants. The learning material is represented by a set of topics, T = {Ti | 1 ≤ i < n}, where Ti represents a topic for a Passenger-session and n is the number of topics/lectures of Software Engineering course per semester. There is a binary relationship between very two different topics Ti and Tj called Precedence(Ti,Tj) to highlight the fact that Ti is a prerequisite for Tj, where i < j. A student can learn a new topic only after finishing all its prerequisites. Also, there is a unary relationship Time(t), t є T, which represents the time allocated for each topic. The value of Time(t) is a number in time units and it differs from a topic to other. Topics are designed to attract participants into an interactive dialogue and to avoid the “silence” during a Passenger session. Thus each topic has a tree structure, with nodes that are: first question for the participants, possible answers by participants, agent response to each of these answers. Also each topic has one or more associated goals, an outcome and consequences. Each topic T can be defined as T = {Qk, Pj | 1 ≤ k ≤ n, 1 ≤ j ≤ m}, where Qk represents a question and Pj represents possible answers by participants to this question, and also agent’s plan/reply to each of these answers, n is the number of questions per topic, m is the number of possible plans per question. Agent’s goals might be locally achievable which means that tutoring material can be retrieved from a local knowledge database, or require interaction with another tutoragent from a different session or a human tutor in case of a word pattern for which it cannot find it in its knowledge database. When students pose a question, the tutor-agent can provide immediate proper answers or hints by matching a regular pattern in its knowledge database. A word pattern refers to a unique collection of words. A regular pattern is a word pattern that appears with high frequency and reflects the relationships within or among topics. Regular patterns are the skeleton of topics and the agent uses these regular patterns to provide students proper answers or future steps in modeling. Here is an example to explain this approach. The following scenario can occur: one of the students doesn’t know how to draw a control memory thus he/she asks the tutor-agent the following question: “How (1) can I draw (2) a control memory (3)?” where (1) (2) and (3) ~How… draw… control memory…~ represent a pattern example. After recognizing a pattern agent will search its knowledge database for a proper answer and will provide this answer to the student. However words like “how” or “where” cannot be regarded as regular patterns because they lack of the essential structure of a question. Therefore in order to realize this pattern algorithm three functions were implemented: • Pattern recognizer: recognize a pattern • Pattern miner: search the recognized pattern in the knowledge database • Pattern adder: adds a new pattern to the knowledge database. The system consists of a number of tutor-agents, representing workflow participants, and a coordinator agent which provides directory service to tutoring-agents. All agents are logically dispersed and arranged in an interconnected network of Passenger platforms.

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

281

Figure 7. Tutor-agent’s architecture.

Tutoring-agents register their services with the coordinator agent and when a nonlocal activity like one tutor agent needs help for an unknown pattern/situation, needs to be performed the coordinator can be looked up to find the appropriate provider in our case can be another tutor agent or a human one. To make tutor-agent’s design process easy and accessible, we proposed a framework where the agent is composed of a set of categories of components (see Fig. 7). A message handler composes the communication heart of the tutor-agent. Messages may be transmitted either to the session or through the interconnection network directly in case help is required. The message handler processes incoming messages and processes appropriate logic to handle them to the core engine. The core engine is responsible for executing one or more learning tasks. Also this component composes the agent decision system, the brain of the agent – agent can choose and plan its future actions according to its goal. Here are implemented pattern recognition and learning algorithms. This component realizes also the evaluation of student – student profile database, based on his behavior during a learning session: it is the one who decided to increase or decrease the parameters (e.g. Activity_Level) which characterize the students’ behavior. The global knowledge database contains information about agents’ beliefs or plans while the coaching knowledge contains information concerning the learning material which is represented by topics. Agent interface is an animated cartoon with human like gestures. Our agent responds to the students’ questions/actions by synthetic speech, facial display and gestures. We choose to design and implement our own animated agent instead of using Microsoft Agent to ensure Passenger platform independence and extensibility. We choose Borland Delphi environment to realize the Passenger environment and also for our prototype implementation.

282

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

Navigational component: that is responsible for tutor-agent’s travel itinerary when it needs to communicate with other tutors (humans or agents) in order to receive help for accomplishing its task. In order to realize the agent’s mobility we implemented the following algorithm: get_hosts_list(CAG); search_hosts(n, Pj, Ak) begin for i= 0 to n-1, i•k do if KL.Ak > KL.Ai then jump to next host; else if ask_help_pattern (Ai, Pj)=NULL then jump to next host; else return(ask_help_pattern (Ai, Pj)); break; return(NULL); end search_hosts; if search_hosts(n, Pj, Ak)=NULL then ask_human(Pj);

where CAG represents the coordinator agent, n is the number of active hosts (we define as an active host a current running Passenger learning session), A k the tutor agent which needs help, Pj the unknown pattern, and KL.Ak the knowledge level of Ak. In case that agent Ak needs help for the unknown pattern Pj, it first connects to coordinator agent and gets the list of the current running Passenger sessions. After this operation is successful it starts migrating from a host to another to get an answer for pattern Pj. For each agent we defined as a knowledge level the binary relationship between very two different topics Ti and Tj called Precedence(Ti,Tj) which also highlights the fact that Ti is a prerequisite for Tj, where i < j. Thus if between agent Ak and agent Aj the following statement KL.Ak > KL.Ai is true that means that agent Ai is teaching a topic Ti, which is a prerequisite for Tk, topic of Ak. In other words, Ai cannot provide answers to Ak, therefore Ak has to move to the next host. In case that the agent Ak cannot find adequate help from the other tutor-agents it has to communicate with a humantutor and get from him/her proper answer.

8. Evaluation The intended evaluation study for this prototype concerns two levels: • •

Usefulness level: the usefulness of the agent facilities within Passenger groupware needs to be evaluated by human teachers. User friendliness level: this level highlights how the agent was accepted by students.

Several experiments took place in the local area network of our institute. Only the second part of the evaluation study was conducted among 25 first year Masterstudents. Each session consisted of three students and one tutor (human or agent). The student experienced the traditional lab with the human tutor and also with the agent feature of the Passenger system. After these experiments, students had to answer to questionnaires files. A sample of questions concerning the second level that were asked to the students is the following: 1.

Do you consider the application attractive? If yes, what did you like about it?

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

283

Figure 8. Acceptance of Agent Tutor.

2. 3. 4.

Do you think that the “agent” features prevented you from understanding the educational process better? Do you prefer the agent tutor instead of the human tutor? Please justify your answer. How do you rate the presence of the Tutor-Agent within the Passenger environment? Please justify your answer.

Based on these questionnaires several statistics could be made. Some results concerning the agent integration and acceptance are shown in the Fig. 8. 20 of the students rated the presence of the agent as “good” motivating that it was a new and unexpected challenge for them, while the rest of 5 were more skeptical stating that they would prefer a more “rich” agent (as animations and emotions). Although the number of participants in the evaluation test was rather small for a quantitative evaluation, the trends seem to be unambiguous: we can say that the agent was accepted by majority of students. We plan to realize the full-evaluation test including an evaluation result for the first level and also to increase the number of student participants.

9. Conclusion Enhancing social roles to animated pedagogical agents can develop a new socialpsychological model for life-like characters similar with a human one. In particular humans can easily adjust their behavior based on their role in a socio-organizational setting, where their actions tend to be driven emotions, attitudes and personality. This paper goal was to show how to integrate agent technology to support collaborative learning in distributed environments. The aim of this research is to provide the first steps to define a method for creating a believable tutor agent which can partially replace human-teachers and assist the students in the process of learning. The outcome of such research is relevant both for basic research into the nature of social minds as well as for design and development of systems in application areas that requires agent to show and use aspects of human intelligence like: ability to learn, recognize or express emotions.

284

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

References [1] C. Conati, Probabilistic assessment of user’s emotions in educational games, Applied Artificial Intelligence, 16 (2002), 555–575. [2] F. de Rosis, B. de Carolis, and S. Pizzulito, Software documentation with animated agents, Proceedings 5th ERCIM Workshop on User Interfaces For All, 1999. [3] Y. Kitamura, H. Tsujimoto, T. Yamada and T. Yamamoto, Multiple characters – agents interface: an information integration platform where multiple agents and human users collaborate, Proceedings of AAMAS-02, New York, 2002. [4] J.C. Lester, S.A. Converse, S.E. Kahler, S.T. Barlow, B.A. Stone, and R.S. Bhogal, The Persona effect: Affective impact of animated pedagogical agents, Proceedings of CHI-97, 1997, 359–366. [5] P. Ekman, An argument for basic emotions, Cognition and Emotion, 6(3–4) (1992), 169–200. [6] D. Moffat, Personality parameters and programs, Creating Personalities for Synthetic Actors, eds: R. Trappl and P. Petta, Springer, 1997, 120–165. [7] C. Elliott, The Affective Reasoner. A process model of emotions in a multi-agent system, PhD. Thesis, Institute for the Learning Sciences, Northwestern University, 1992. [8] B. Marin, A. Hunger, S. Werner, S. Meila and C. Schuetz, An Intelligent Tutor-Agent to Support Collaborative Learning within a Virtual Environment, Proceedings of IEEE International Conference on Systems, Man and Cybernetics, Hague, Netherlands, 2004. [9] J. Bates, The role of emotion in believable agents, Communications of the ACM, 37(7), 1994, 122–125. [10] P.M. Blumberg, Old Tricks, New Dogs: Ethology and Interactive Creatures, PhD Thesis, MIT, 1996. [11] D. Thalmann, H. Noser, and Z. Huang, Autonomous virtual actors based on virtual sensors, Creating Personalities for Synthetic Actors, eds: R. Trappl and P. Petta, Springer, 1997, 25–42. [12] T. Koda, Agents with faces: a study on the effect of the personification of software agents, Master thesis, MIT Media Lab, Cambridge, MA, 1996. [13] B. Reeves and C. Nass, The Media Equation: how people treat computers, televisions and new media like real people and places, Cambridge University Press, Cambridge, 1996. [14] B. Reeves, and R. Rickenberg, The effects of animated characters on anxiety, task performance, and evaluations of user interfaces, Proceedings of CHI 2000, The Hague, 2000, 329–336. [15] C. Nass, J. Steuer, and E. Tauber, Truth is the beauty: Researching embodied conversational agents, Embodied Conversational Agents, eds. J. Cassell, S. Prevost, J. Sullivan, and E. Churchill, MIT Press, Cambridge, 2000, 374–402. [16] H. Prendinger and M. Ishizuka, Life-Like Characters: Tools, Affective Functions, and Applications (Cognitive Technologies), Springer, New York, 2004. [17] B.J. Biddle and E.J. Thomas, Role Theory: Concepts and Research, K.E. Kriger Publishing Company, New York, 1979. [18] L. Rene-Boullier, Pedagogical coordination: a global vision of the personalized accompaniment of DESS DICIT, Revue Sticef, 10 (2003). [19] C. Schmidt, A remote laboratory using virtual reality on the web, Simulation, 73:1 (1999), 13–21. [20] B. Marin, A. Hunger, S. Werner, S. Meila and C. Schuetz, A synchronous groupware tool to conduct a spatially distributed collaborative learning process, Proc. of the 5th Int. Conf. on Information Technology Based Higher Education and Training, Istanbul, 2004, 269–274. [21] R.L. Atkinson, R.C. Atkinson and E.R. Hilgard, Introduction to Psychology, Harcourt Brace Jovanovich Inc. 1983. [22] A. Bandura, Social Learning Theory, Prentice-Hall, Englewood Cliffs, 1977. [23] R. Plutchik, A General Psychoevolutionary Theory of Emotion, Emotion: Theory, Research, and Experience 1:3 (1980), 33, New York: Academic. [24] P.T. Ward and S.J. Mellor, Structured development for real time systems, Prentice-Hall International, 1985. [25] H.-P. Dommel and J.J. Garcia-Luna Aceves, Group coordination support for synchronous Internet collaboration, IEEE Internet Computing, March-April (1999), 74–80. [26] P. Dourish and V. Bellotti, Awareness and coordination in shared workspaces, eds: J. Turnier and R. Kraut, Proc. Of CSCW’92 – Sharing Perspectives, ACM Press Toronto Canada, 107–114. [27] M.T.H. Chi, S.A. Siller, et al., Learning from Human Tutoring, Cognitive Science 25, 4 (2001), 471–533 [28] G. Clarebout, J. Ellen, W.L. Johnson and E. Shaw, Animated Pedagogical Agents: An Opportunity to be Grasped?, Journal of Educational Multimedia and Hypermedia, 11 (2001), 267–286. [29] L. Aroyo and P. Kommers, Preface – Intelligent Agents for Educational Computer-Aided Systems, Journal of Interactive Learning Research, 10 (3/4) (1999), 235–242.

B.F. Marin et al. / A Framework for Emotional Agents as Tutoring Entities

285

[30] R.S.S. Guizzardi, G. Wagner and L. Aroyo, Agent-oriented modeling for collaborative learning environments: a peer-to-peer helpdesk case study, Proceedings of the XIII th – Brazilian Symposium on Computers in Education, 2002. [31] J. Kay, Learner Control. User modeling and User Adapted Interaction, 11 (2001), 111–127, Kluwer Academic Publishers, Netherlands. [32] A. Giddens, Sociology, Polity Press, London, UK, 1997, 585. [33] B.J. Biddle and E.J. Thomas, Role Theory: Concepts and Research, K.E. Kriger Publishing Company, New York, 1979. [34] U.M. Borghoff and J.H. Schlichter, CSCW – An Introduction to Distributed Applications, Springer Verlag, Berlin, July 2000.

286

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees Cyrus F. NOURANI [email protected] Academia U California and [email protected] http://beam.to/af2im Multimedia Projects on Fraunhofer Institute IMK, Sank Augustin, Germany Netzspannung.org T

TU

TU

UT

UT

Abstract. A haptic logic and computing paradigm is presented with a basis for multiagent visual computing with the Morph Gentzen logic. The techniques since 1999 are the bases to a haptic logic and multiagent cognition where central affective computing questions might be addressed.The computing model is based on a novel competitive learning with agent multiplayer game tree planning. Specific agents are assigned to transform the models to reach goal plans where goals are satisfied based on competitive game tree learning. Affectice computing is addressed with a new haptic computing logic. Questions such as ‘Are intelligent decisions based on emotions?’ are addressed. Further questions on emotions and consciousness models studied. (The IM_BID model is introduced for planning and spatial computing. Visual intelligent objects are applied with virtual intelligent trees to carry on visual planning. New KR techniques are presented with G-diagrams and applications to define computable models and relevant world reasoning. G-diagrams are diagrams defined from a minimal set of function symbols that can inductively define a model. G-diagrams are applied to relevance reasoning by model localized representations and a minimal efficient computable way to represent relevant knowledge for localized AI worlds. Diagrammatic reasoning is defined in terms of inferences directed by the G-diagrams for models. The techniques show how computable AI world knowledge is representable. G-diagrams are applied towards KR from planning with nondeterminism and planning with free proof trees to planning with predictive diagrams. The IM Morph Gentzen Logic for computing for multimedia are new projects with important computing applications since the author’s and contemporary projects on diagrammatic computing. The basic principles are a mathematical logic where a Gentzen or natural deduction systems is defined by taking arbitrary structures and multimedia objects coded by diagram functions. The techniques can be applied to arbitrary structures definable by infinitary languages. Multimedia objects are viewed as syntactic objects defined by functions, to which the deductive system is applied. A basis to VR computing and computational illusion is presented. Keywords. Haptic logic, models, VR, intelligent syntax, agent linguistics, metacontextual reasoning, diagrams for models, morph gentzen computing, multiagent cognitive models, emotional intelligence

1. Introduction Affective computing expands human-computer interaction by including emotional communications together with appropriate means for handling affective information.

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

287

New models are suggested for computing recognition of human emotions, and both practical and theoretical applications are described for learning, human computer interactions, and perceptual information retrieval, creative arts and entertainment, human health and machine intelligence. Scientists have discovered many surprising roles for played by human emotion- especially in cognitive processes, such as perception, decision making, memory and more. Human intelligence includes emotional intelligence, especially the ability to accurately recognize and express affective emotions. The techniques here are the bases to a haptic logic and multiagent cognition where central affective computing questions might be addressed. Emotions might signal cognitive changes. The challenge remains for scientists to determine precisely how the any regulatory signaling cognitive functions and affective systems work their influences. Understanding the emulating human emotion might help in our efforts to understand intelligence. Are intelligent decisions based on emotions? If there is a Gestalt model for the world decided on, the answer might be affirmative. Furthermore, emotions might be independent of consciousness modes (Nourani 1998–1999). Agent computing is introduced to interactive intelligent multimedia. An overview to a practical agent computing model based on beliefs, intentions, and desire is presented and possible augmentation to intelligent multimedia is explored. (NilssonGenesereth 1987) introduces agent architectures. A specific agent might have internal state set I, which the agent can distinguish its membership. The agent can transit from each internal state to another in a single step. With our multi-board model agent actions are based on I and board observations. There is an external state set S, modulated to a set T of distinguishable subsets from the observation viewpoint. A sensory function s:S → T maps each state to the partition it belongs. Let A be a set of actions which can be performed by agents. A function action can be defined to characterize an agent activity action: T → A There is also a memory update function mem: I x T→ I. Worlds, epistemics, and cognition for androids are introduced with precise statements. The foundations are applied to present a brief on Computational Illusion, affective computing, and Virtual Reality. KR for AI Worlds, and Computable Worlds are presented with diagrams. Cognitive modeling is briefed. A preview to computational epistemology with cardinality and concept descriptions is introduced. Deduction models and perceptual computing is presented with a new perspective. Intelligent multimedia interfaces are an important component to the practical computational aspects. Visual context and objects are presented with multiagent intelligent multimedia. Context abstraction and meta-contextual reasoning is introduced as a new field where multimedia context can be treated with the author’s Morph Genzen logic. Mulltiagent visual multi-board planning is introduced as a basis to intelligent multimedia with applications to spatial computing.

2. The Agent Models and Desire Let us start with the popular agent computing model the Beliefs, Desire, and Intentions, henceforth abbreviated as the BID model (Brazier-Truer et al.). BID is a generic agent computing model specified within the declarative compositional modeling framework for multi-agent systems, DESIRE. The model, a refinement of a generic agent model, explicitly specifies motivational attitudes and the static and dynamic relations between motivational attitudes. Desires, goals, intentions, commitments, plans, and their relations are modeled. Different notions of strong and weak agency are presented at

288

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

(Wooldridge and Jennings, 1995). (Velde and Perram, 1996) distinguished big and small agents. To apply agent computing with intelligent multimedia some specific roles and models have to be presented for agents. The BID model has emerged for a “rational agent”: a rational agent described using cognitive notions such as beliefs, desires and intentions. Beliefs, intentions, and commitments play a crucial role in determining how rational agents will act. Beliefs, capabilities, choices, and commitments are the parameters making component agents specific. Such bases are applied to model and to specify mental attitudes (Shoham 1993), (Rao and Georgeff, 1991; Cohen and Levesque, 1990; Dunin-Keplicz and Verbrugge, 1996). A generic BID agent model in the multiagent framework DESIRE is presented towards a specific agent model. The main emphasis is on static and dynamic relations between mental attitudes, which are of importance for cooperative agents. DESIRE is the framework for design, and the specification of interacting reasoning components is a framework for modeling, specifying and implementing multi-agent systems, see (Brazier, Dunin-Keplicz, Jennings, and Truer, 1995, 1996; Dunin-Keplicz and Truer, 1995). Within the framework, complex processes are designed as compositional architectures consisting of interacting task-based hierarchically structured components. The interaction between components and between components and the external world is explicitly specified. Components can be primitive reasoning components using a knowledge base, but may also be subsystems which are capable of performing tasks using methods as diverse as decision theory, neural networks, and genetic algorithms. As the framework inherently supports interaction between components, multi-agent systems are naturally specified in DESIRE by modeling agents as components. The specification is sufficient to generate an implementation. Specific techniques for such claims might be further supported at (Nourani 1993a, 99a). A generic classification of mental attitudes is presented and a more precise characterization of a few selected motivational attitudes is given. The specification framework DESIRE for multi-agent systems is characterized. A general agent model is described. The framework of modeling motivational attitudes in DESIRE is discussed. 2.1. Mental Attitudes Agents are assumed to have the four properties required for the weak notion of agency described in (Wooldridge and Jennings, 1995). Thus, agents must maintain interaction with their environment, for example observing and performing actions in the world: reactivity; be able to take the initiative: pro-activeness; be able to perform social actions like communication, social ability; operate without the direct intervention of other (possibly human) agents: autonomy. Four main categories of mental attitudes are studied in the AI literature: informational, motivational, social and emotional attitudes. The focus is on motivational attitudes, although other aspects are marginally considered. In (Shoham and Cousins, 1994), motivational attitudes are partitioned into the following categories: goal, want, desire, preference, wish, choice, intention, commitment, plan. Individual agents are assumed to have intentions and commitments both with respect to goals and with respect to plans. A generic classification of an agent’s attitudes is defined as follows: 1. 2. 3.

Informational attitudes: Knowledge; Beliefs. Motivational attitudes: Desires; Intentions- Intended goals and Intended plans. Commitments: Committed goals and Committed plans.

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

289

In planning, see Section 6, the weakest motivational attitude might be desire: reflecting yearning, wish and want. An agent may harbor desires which are impossible to achieve. Desires may be ordered according to preferences and, as modeled in this paper, they are the only motivational attitudes subject to inconsistency. At some point an agent must just settle on a limited number of intended goals, i.e., chosen desires. 2.2. Specifying BID Agents The BID-architectures upon which specifications for compositional multi-agent systems are based are the result of analysis of the tasks performed by individual agents and groups of agents. Task (de)compositions include specifications of interaction between subtasks at each level within a task (de)composition, making it possible to explicitly model tasks which entail interaction between agents. The formal compositional framework for modeling multi-agent tasks DESIRE is introduced here. The following aspects are modeled and specified: 1. 2. 3. 4. 5.

a task (de)composition information exchange sequencing of (sub)tasks subtask delegation knowledge structures

Information required/produced by a (sub)task is defined by input and output signatures of a component. The signatures used to name the information are defined in a predicate logic with a hierarchically ordered sort structure (order-sorted predicate logic). Units of information are represented by the ground atoms defined in the signature. The role information plays within reasoning is indicated by the level of an atom within a signature: different (meta)levels may be distinguished. In a two-level situation the lowest level is termed object-level information, and the second level meta-level information. Some specifics and a mathematical basis to such models with agent signatures might be obtained from (Nourani 1996a) where the notion had been introduced since 1994. Meta-level information contains information about object-level information and reasoning processes; for example, for which atoms the values are still unknown (epistemic information). Similarly, tasks that include reasoning about other tasks are modeled as meta-level tasks with respect to object-level tasks. Often more than two levels of information and reasoning occur, resulting in meta-meta-information and reasoning. Information exchange between tasks is specified as information links between components. Each information link relates output of one component to input of another, by specifying which truth-value of a specific output atom is linked with which truth value of a specific input atom. For a multiagent object information exchange model see, for example, (Nourani 1996a). The generic model and specifications of an agent described above, can be refined to a generic model of a rational BID-agent capable of explicit reasoning about its beliefs, desires, goals and commitments. 3. Dynamics and Situations 3.1. Worlds and a Robot’s Touch Starting with the issues raised by Heidegger in 1935–36, starting with the notion of “What is a thing” as put forth in (Heidegger 63). The author’s reaction when presented

290

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

Person ------Cognition----Real Worlds and Objects | | | | | | |--------------------\ \-------------------------------| | The Sensory Illusion Gap | | | | | Robot------World Description---RealWorlds Objects

Figure 1. The Sensory Illusion Gap. T

with such challenges to computing applications with philosophical epistemics, while visiting INRIA, Paris around 1992, was to start with “first principles”, not touching such difficult areas of philosophy and phenomenology, and only present views to what they could imply for the metamathematics of AI. However, since the author’s techniques were intended for AI computations and reasoning, rather than knowledge representation from observations. Heidegger’s definitions had to be taken further. The common point of interest is symbolic knowledge representation. However, the research directions are two essentially orthogonal, but not contradicting, views to knowledge representation. 3.2. Computational Illusion and Virtual Reality der Vielleicht Vorhandenen objects are the Perhaps Computable and might be a computational illusion, as further illustrated by the following figure on the human intelligence and artificial intelligence comparison. Thus the robot’s senses are not always real. The important problem is to be able to define worlds minimally to have computable representations with mathematical logic thus the ability to make definitive statements. Heidegger’s Die Frage nach dem Ding will prove to be a blessing in disguise. Could it have computing applications to things without. Heidegger had defined three sorts of things 1. 2. 3.

Things in the sense of being “within reach”, des Vorhandenen. Things which “unify” things of the first kind, or are reflections on, resolution and actions. Things of kind 1 or 2 and also any kind of things which are not nothing.

To define a logic applicable to planning for robots reaching for objects, the der Vielliecht Vorhandenen computational linguistics game is defined. To start, let us explore Heidegger’s views of the “des Vorhandenen”, having to do with what object is within “reach” in a real sense. In AI and computing applications notion of des Vorhandnen is not absolute. As an AI world develops the objects that have names in the

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

291

Figure 2. Objects and Competitive Games.

world are at times des Vorhandnen and as defined by a principle of Parsimony only des Vorhandnen in an infinitary sense of logic (Nourani 1984,91). The logical representation for reaching the object might be infinitary only. The phenomenological problem from the robot’s standpoint is to acquire a decidable descriptive computation for the problem domain. Thus what is intended to be reached can stay always out of reach in a practical sense, unless it is at least what the author called der Vielliecht Vorhandenen (Nourani 1994a,94b). The computing issues are the artificial intelligence computation and representation of real objects. That is, we can make use of symbolic computation to be able to “get at” a real object. At times, however, only infinite computations could define real world objects. For example, there is a symbolic computation for an infinite ordinal, by an infinite sequence of successor operations on 0. The same problem might arise when the robot tries to actually get at elementary objects, where the robot finds what is called a paradox in (Didday 1990): that elementary objects have to be defined by comprehension. Comprehension is a closure with respect to properties that are essential and cannot be dropped without loss to the enclosed. Since the paper in its theory that is presented in part here, does not restrict Heidegger’s definition, it can be further developed for AI applications. I might suggest ways of incorporating the above for computing applications. The problems with words, objects and symbols have been there since Quine (1950’s). Furthermore, the present notion of der Vielliecht Vorhandenen is not intend to be the sense in which a robot cannot reach a particular object. The intent is that the language could have names for which the corresponding thing is not obvious in the AI world and there is incomplete information until at some point the world is defined enough that there is a thing corresponding to a name, or that at least there is a thing by comprehension, which only then becomes des Vorhandnen as the AI world is further defined or rearranged. These issues are examined in the computational context in the sections below. For example, the der Vielleicht Vorhandenen game has a winning strategy if the world descriptions by G-diagrams define the world enough to have a computation sequence to reach for an intended objective. This implies there must be a decidable descriptive

292

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

computation (Nourani 1994,96) for the world applied. The immediate linguistics example of these concepts from natural languages is a German child’s language in which to “vor” and “handenen” are some corresponding things in the child’s language world and mind, but “vorhandenen” is not a thing in that child’s world and only becomes a thing as the linguistics world is further defined for the child. When can the child reach for the stars? as Heidegger alludes on ‘children and stars’. 3.3. Representing AI Worlds Diagrams are the “basic facts of a model”, i.e. the set of atomic and negated atomic sentences that are true in a model. Generalized diagrams are diagrams definable by a minimal set of functions such that everything else in the model’s closure can be inferred, by a minimal set of terms defining the model. Thus providing a minimal characterization of models, and a minimal set of atomic sentences on which all other atomic sentences depend. However, since we cannot represent all aspects of a real world problem, we need to restrict the representation to only the relevant aspects of the real world we are interested in. Let us call this subset of relevant real world aspects the AI world. Our primary focus will be on the relations amongst KR, AI worlds, and the computability of models. Truth is a notion that can have dynamic properties. Interpretation functions map language constructs (constants, function and predicate symbols) onto entities of the world, and determine the notion of truth for individuals, functions and relations in the domain. The real world is infinite as the AI worlds are sometimes. We have to be able to represent these ideas within computable formulations. Even finite AI worlds can take an exponential number of possible truth assignments. Thus the questions: how to keep the models and the KR problem tractable, such that the models could be computable and within our reach, are an important area (Nourani 1991,93a,93b,94,96), (Lake 1996). 3.4. Computable World Models To prove Godel’s completeness theorem, Henkin defined a model directly from the syntax of the given theory. The reasoning enterprise requires more general techniques of model construction and extension, since it has to accommodate dynamically changing world descriptions and theories. The techniques in (Nourani 1983,87,91) for model building as applied to the problem of AI reasoning allow us to build and extend models through diagrams. We apply generalized diagrams to define models with a minimal family of generalized Skolem functions. The minimal set of function symbols are those with which a model can be built inductively. The models are computable as proved by (Nourani 1984,93a,95b). The G-diagram methods applied and further developed here allow us to formulate AI world descriptions, theories, and models in a minimal computable manner. Thus models and proofs for AI problems can be characterized by models computable by a set of functions. 3.5. AI Model Diagrams An AI world consists of individuals, functions on them, and relations between them. These entities allow us to fix the semantics of a language for representing theories about AI worlds. We take the usual model-theoretical way, and assign via an interpretation function individuals to constants, functions to function symbols and relations to

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

293

predicate symbols. Let us define a simple language L = <{tweedy},{a},{bird}, predicate letters, and FOL>>. A model may consist of {bird (tweedy), ¬ penguin(tweedy) → bird(tweedy), bird(tweedy) v ¬ bird(tweedy), ...}, others may consist of {p(a), ¬ p(a) → p(a), p(a) v p(x), p(a) v p(x) v p(y), ...}. Because we can apply arbitrary interpretation functions for mapping language constructs into AI worlds, the number of models for a language is infinite. Although this makes perfect sense from a theoretical and logical point of view, from a practical point of view, this notion of model is too general for AI applications. For AI we want effective and computable models. Thus, it is useful to restrict the types of models that we define for real world applications. Primarily, we are interested in models with computable properties definable from a theory. In order to point out the use of the generalized method of diagrams we present a brief view of the problem of planning from (Nourani 1991) within the present formulation. The diagram of a structure in the standard model-theoretic sense is the set of atomic and negated atomic sentences that are true in that structure. The generalized diagram (G-diagram) is a diagram in which the elements of the structure are all represented by a minimal family of function symbols and constants. It is sufficient to define the truth of formulas only for the terms generated by the minimal family of functions and constant symbols. Such assignment implicitly defines the diagram. This allows us to define a canonical model of a theory in terms of a minimal family function symbols. Models uphold to a deductive closure of the axioms modeled and some rules of inference, depending on the theory. By the definition of a diagram they are a set of atomic and negated atomic sentences. Hence a diagram can be considered as a basis for defining a model, provided we can by algebraic extension, define the truth value of arbitrary formulas instantiated with arbitrary terms. Thus all compound sentences build out of atomic sentences then could be assigned a truth value, handing over a model. This will be made clearer in the following subsections. The following examples would run throughout the paper. Consider the primitive first order language (FOL) L = {c},{f(X)},{p(X),q(X)}. Let us apply Prolog notation convention for constants and variables) and the simple theory {for all X: p(X) → q(X),p(c)}, and indicate what is meant by the various notions. (model) = {p(c),q(c),q(f(c)),q(f(f(c))),...},{p(c) &q(c), …. p(c) & p(X), p(c) &p(f(X)), ...}, {p(c) v p(X), p(c) v p(f(X)), p(c) → p(c)...}. (diagram) = {p(c),q(c),p(c),q(f(c)),q(f(f(c))),...},...,q(X)}; i.e., the diagram is the set of atomic formulas of a model. Thus the diagram is (diagram) = {p(c),q(c),q(f(c)),q(f(f(c))), ..,q(X)}. There are various notions of diagram from the author’s papers (see references) applied here. The term generalized diagram refers to diagrams that are instantiated with generalized Skolem functions. The generalized Skolem functions were defined in by the author, for example, (Nourani 1991) as functions with which initial models are defined inductively. We can define generalized diagrams based on the above.. The term generalized is applied to indicate that such diagrams are defined by algebraic extension from basic terms and constants of a language. The diagrams are completely defined from only a minimal function set. Generalized diagrams is (generalized diagram) = {p(c),q(c),p(f(t)),q(f(t))} for t defined by induction, as {t0 = c , and tn = {f(t(n – 1))} for n > 0. It is thus not necessary to redefine all f(X)’s since they are instantiated.

294

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

3.6. Cognitive Modeling Cognitive modeling can be enhanced with diagrams since our G-diagram techniques imply automatic models from basic functions. A systematic methodology for Cognitive modeling can be considerably assisted by the G-diagram modeling. The area has been emphasized by (Cooper et al. 1996). The notion of a symbolic object is put forth in (Didday 1990) by considering some individuals as elementary things and then defining symbolic objects from the elementary objects by a comprehension technique with some descriptor functions. Thus comprehension and descriptor make the jump from elementary objects to symbolic objects functions. The diagram of a structure is the set of atomic and negated atomic sentences that are true in that structure. The generalized diagram (G-diagram) (Nourani 1987,91) is a diagram in which the elements of the structure are all represented by a minimal family of function symbols and constants, such that it is sufficient to define the truth of formulae only for the terms generated by the minimal family of functions and constant symbols. Such assignment implicitly defines the diagram. It allows us to define a canonical model of a theory in terms of a minimal family of function symbols . Generalized diagrams are precisely what allow us to build models from the syntax of a theory, thus allow for symbolic computation of models and theories. The author had defined the notion of generalized diagram ever since (Nourani 1984) for AI reasoning. Since the author has shown generalized diagrams for models capture the possible worlds formulation in a concise and elegant manner. In a possible world approach one focuses on the “state of affairs” that are compatible with what one knows to be true. We have shown in the above papers how the approach with G-diagrams to possible worlds. 3.7. Kant’s Idealism and Illusion Logic Kant compares his innovations to the first thoughts of Copernicus. It involves reversing the usual way cognition is viewed. Instead of taking our knowledge conforming to a real object, we think of objects as conforming to our ways of knowledge. The latter include “forms of sensibility” through which objects are given to the mind as sensory experience, and pure concept categories, through which they are thought. Since objects must appear to us in sensible forms is order to be known, it follows that we can know only them as they appear, not as they may be themselves. Accordingly for Kant human knowledge is limited to appearances, whereas things in themselves are “noumena” – are thinkable but not actually knowable. Kant termed the doctrine Transcendetal Idealism. Given the idealism is the possibility of synthesizing a priori knowledge to possible description and experience is easily explainable, since each object must necessarily conform to the conditions under which they can become objects for us. It assumes the human mind possesses such condition and demonstrating it is transcendental Aesthetics. Creative pure mathematics as a synthetical cognition a priori, is only possible by referring to no other object than those of the senses. At the base of this empirical intuition lies a pure intuition which is a priori. Yet the faculty of intuition a priori affects not the matter of the phenomenon, but its form viz. space and time. Frege’s basic logical ideas and Hilbert’s program separate carrying out pure mathematics from the physical cognition perceptions of what is carried out as an end. Frege’s “concept and object” and on “sense and meaning,” is where carrying out logic for objects named by a language had started being distinguished from the object sense perception. Hilbert’s program, aside from its being left to reconcile with transcendental idealism on concepts,

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

295

were to arithmatize the entire mathematics. Where are we with descriptive computing Heidegger objects? We are at the language, model, arithemtatization trichotomy. The objects are described with languages as Frege intended, modeled by structures, which can be examined by Kan’t transcendental Idealism, and their computability and reducibility areas Hilbert arithmetized. Hence there is a systematic basis to carryout conceptobject descriptions for machine discovery. The correspondence of modalities to Possible Worlds and the containment of the possible worlds approach by the generalized diagrams approach of this author implies that we can present a model-theoretic formulation of the concept of modal symbolic objects (Didday 1990, Nourani 1992,93b, 1997). Objects with varying properties, with cross product of modes formed from various generalized diagrams corresponding to each mode. Also the notion of language L has some consequences as far as the model theory to be developed is concerned. Then all the notions for the various modes could be defined and perhaps open new views of computation on generalized diagrams allowing us to represent views of cognition and computation with modes of thought in artificial intelligence.

4. Competitive Models and Games Planning is based on goal satisfaction at models. Multiagent planning, for example as (Muller and Pischel 1994, Bazier et al. 1997), in the paper is modeled as a competitive learning problem where the agents compete on game trees as candidates to satisfy goals hence realizing specific models where the plan goals are satisfied. When a specific agent group “wins” to satisfy a goal, the group has presented a model to the specific goal, presumably consistent with an intended world model. For example, if there is a goal to put a spacecraft at a specific planet’s orbit, there might be competing agents with alternate micro-plans to accomplish the goal. While the galaxy model is the same, the specific virtual worlds where a plan is carried out to accomplish a real goal at the galaxy via agents are not. Therefore, Plan goal selections and objectives are facilitated with competitive agent learning. The intelligent languages (Nourani 1996, 1998) are ways to encode plans with agents and compare models on goal satisfaction to examine and predict via model diagrams why one plan is better than another, or how it could fail. Virtual model planning is treated in the author’s publications where plan comparison can be carried out at VR planning (Nourani 1999d). Games play an important role from economics to mathematic, computing and strategy to decision science. The author has presented new game tree planning techniques and brought forth agent game tree planning to AI, decisions and economic games. Competitive models with game trees are applied to manage processes and achieve goals in multidisciplinary arenas. Example competitive models are in multiplayer games. For example, when arranging team playing, there are many permutations on where the players are positioned. Every specific player arrangement is a competitive model. There is a specifc arrangement that does best in a specific game. What model is best can be detemined with agent player competitive model learning. Intelligent tree computing theories the author defined since 1993 can be applied to present precise strategies and prove theorems on multiplayer games. Game tree degree with respect to models is defined and applied to prove soundness and completeness. The game is viewed as a multiplayer game with only perfect information between agent pairs. Upper bounds on determined games are presented in (Nourani 1997–1999). The author had presented a chess-playing basis in 1997 to a computing conference. For each chess piece a designating agent is defined. The player

296

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

P makes its moves based on the board B it views. might view chess as if the pieces on the board had come alive and were autonomous agents carrying out twoperson games as in Alice in Wonderland. Game moves are individual tree operations. 4.1 Intelligent AND/OR Trees and Search AND/OR trees Nilsson (1969) are game trees defined to solve a game from a player’s standpoint. n / | \ m /__|__\ / | \

an OR node. an AND node

Formally a node problem is said to be solved if one of the following conditions hold. 1. 2. 3.

The node is the set of terminal nodes (primitive problem- the node has no successor). The node has AND nodes as successors and the successors are solved. The node has OR nodes as successors and any one of the successors is solved.

A solution to the original problem is given by the subgraph of AND/OR graph sufficient to show that the node is solved. A program which can play a theoretically perfect game would have task like searching and AND/OR tree for a solution to a oneperson problem to a two-person game. An intelligent AND/OR tree is and AND/OR tree where the tree branches are intelligent trees. The branches compute a Boolean function via agents. The Boolean function is what might satisfy a goal formula on the tree. An intelligent AND/OR tree is solved iff the corresponding Boolean functions solve the AND/OR trees named by intelligent functions on the trees. Thus node m might be f(a1,a2,a3) & g(b1,b2), where f and g are Boolean functions of three and two variables, respectively, and ai’s and bi’s are Boolean valued agents satisfying goal formulas for f and g. g is on OR agent / | \ | b1 | b2 f f is an AND agent /__|__\ / | \ The chess game trees can be defined by agent augmenting AND/OR trees (Nilsson 69). For the intelligent game trees and the problem solving techniques defined, the same model can be applied to the game trees in the sense of two person games and to the state space from the single agent view. The two person game tree is obtained from the intelligent tree model, as is the state space tree for agents. To obtain the two-person game tree the cross-board-co board agent computation is depicted on a tree. Whereas the sate-space trees for each agent is determined by the computation sequence on its side of the board-co board. Thus a tree node m might be f(a1,a2,a3) & g(b1,b2), where

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

297

f and g are Boolean functions of three and two variables, respectively, and ai’s and bi’s are Boolean valued agents satisfying goal formulas for f and g. g is on OR agent / | \ | b1 | b2 f f is an AND agent /__|__\ / | \ a1 a2 a3 A tree game degree is the game state a tree is at with respect to a model truth assignment, e.g. to the parameters to the Boolean functions above. Let generic diagram or G-diagrams be diagrams definable by specific functions. Intelligent signatures (Nourani 1996) are signatures with designated multiplayer game tree function symbols. A soundness and completeness theorem is proved on the intelligent signature language (Nourani 1996). The techniques allowed us to present a novel model-theoretic basis to game trees, and generally to the new intelligent game trees with applications to game tree planning. 4.1. Cardinality and Concept Descriptions Let us present what we refer to as Descriptive Computation applying generalized diagrams, following our earlier papers Nourani (1988,91). We define descriptive computation to be computing with G-diagrams for the model and techniques for defining models with G-diagrams from the syntax of a logical language. G-diagrams are diagrams definable with a known function set. Thus the computing model is definable by G-diagrams with a function set. The analogous terminology in set theory refers to sets or topological structure definable in a simple way. Thus by descriptive computation we can address artificial intelligence planning and theorem proving, for example. The author in (Nourani 1984) pursues the latter computational issues. The logical representation for reaching the object might be infinitary only. We show in Nourani (1994a,b, 96) that the artificial intelligence problem from the robot’s stand point is to acquire a decidable descriptive computation for the problem domain. (Nourani 1996) proves specific theorems for descriptive computing on diagrams. A compatibility theorem applies descriptive computing to characterize situation compatibility. Further, a computational epistemic reducibility theorem is proved by the descriptive computing techniques on infinitary languages by the author in (1994b). A deterministic epistemics is defined and it is proved not reducible to known epistemics. Cardinality restrictions on concepts are important areas explored by AI. The concept description logics systems allow users to express local cardinality on particular role filers. Global restrictions on the instances of a concept are difficult and not possible. Cordiality restrictions on concepts can be applied as application domain description logic ( Baader et al. 1996). The concept definitions with G-diagrams for localized KR and its relations to descriptive computable sets can be applied to concept cardinality restriction. By applying localized functions to define G-diagrams models for languages as defined by (Baader et al. 96) can be generated with cardinality restrictions.

298

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

4.2. Deduction Models and Perceptual Computing It might be illuminating to compare the G-diagram techniques and computational epistemology to the (Konolige 1984) starting with the consequential closure problem for artificial intelligence and the possible worlds. What Konologie starts with is the infeasibility premise for consequential closure, i.e. the assumption that an agent knows all logical consequences of his beliefs. The deductive model is defined for situations where belief derivation is logically incomplete. The area had been voiced since (Fodor 75) and (Moore 80). Konolige applies a model where beliefs are expressions in the agent’s “mind” and the agent reasons about them by manipulating syntactic objects. When the process of belief derivation is logically incomplete, the deduction model does not have the property of the consequential closure. Konolige defines a saturated deduction model and claims a correspondence property: For every modal logic of belief based on Kripke possible world models, there exists a corresponding deduction model logic family with an equivalent saturated logic. In (Nourani 84,87,91,95,96) and the present paper it is shown there is a minimal characterization of AI reasoning models with generic diagrams from which models can be defined for belief revision and automatically generated. The G-diagrams are defined for incomplete KR, modalities, and model set correspondence. What computational epistemology defines is a model theoretic technique whereby without the consequential closure property requirements on agents a model-theoretic completeness can be ascertained via nodeterministic diagrams. Specific modal diagrams were defined for computational linguistics models by (Nourani 1993,95). From the practical view point the KR problems for first order logic formalisms as it is implied by Konolige’s deductive view implies defining ways to apply links (Woods 75). In (Nourani-Lieberherr 1985) we showed how to define KR for automatic modeling with abstract objects for links in semantic nets (Schubert 76). Hence the deductive view might benefit from our computational applications.

5. Affective Computing Picard’s (1999) assertion indicates that not all modules in a designed AI system might pay attention to emotions, or to have emotional components. Some modules are useful rigid tools, and it is OK to keep them that way. However, there are situations, where human-machine interaction could be improved by having machines naturally adapt to their users. Affective computing expands human-computer interaction by including emotional communications together with appropriate means for handling affective information. R Picards’s projects addresses reducing user frustrations, enabling comfortable communication of user emotions, developing infrastructures to handle emotions, and building tools that help develop special emotional skills. Since neurological studies indicate that the role of emotion in human cognition is essential; emotions are not a luxury. Instead, emotions play a critical role in rational decision making, in perception, in human interaction, and in human intelligence. These facts, combined with the abilities computers are acquiring in expressing and recognizing affect, open new areas for research. The “key issue” in affective computing that arises is what deliberately influences emotions. New models are suggested for computing recognition of human emotions, and both practical and theoretical applications are described for learning, human computer

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

299

interactions, and perceptual information retrieval, creative arts and entertainment, human health and machine intelligence. Scientists have discovered many surprising roles for played by human emotionespecially in cognitive processes, such as perception, decision making, memory and more. Human intelligence includes emotional intelligence, especially the ability to accurately recognize and express affective emotions. Picard suggests that affective intelligence, the communications and management of affective information in humancomputer interaction is, a key link that is missing in telepresence environments and other technologies that mediate human-computer communications. Picard-Cossier (1997) discuss the new research in affective intelligence, and how it can impact upon and enhance the communications process, allowing the delivery of more natural interaction that is critical for a true telepresence. The techniques here since (Nourani 1999) are the bases to a haptic logic and multiagent cognition where central affective computing questions might be addressed. The serious questions raised and addressed in part by Picard are as follows. Emotions might signal cognitive changes. Picard suggests there are a small set of analogue signaling mechanisms. Increasing neurologist, cognitive scientists, and psychologists are finding that human emotions biases memory formation, which in turn appears to bias decision making, creativity, planning, perception, judgment, and mood-congruent memory memory retrieval, and a host of specific phenomena, The challenge remains for scientists to determine precisely how the any regulatory signaling cognitive functions and affective systems work their influences. Picard argues that once we have duplicated in the machine all the important biasing, regulatory, motivational, behavioral, and other expressive phenomena associated with human emotions, and done so in a general computable and flexible way, that we will have grated the machine emotions. Understanding the emulating human emotion might help in out efforts to understand intelligence. Are intelligent decisions based on emotions? If there is a Gestalt model for the world decided on, the answer might be affirmative. Further more, emotions are independent of consciousness models (Nourani 1998–1999, Chalmers 1999). As we have seen thus far there is new progress in intelligent (knowledge-based) user interfaces that exploit multiple media text, graphics, maps – and multiple modalities – visual, auditory, gestural to facilitate human-computer interaction. The areas addressed are automated presentation design, intelligent multimedia interfaces, and architectural and theoretical issues. For example, (Mayberry 1997) is an edited volume on some of the original contributions in the area. There are three sections that address automated presentation design, intelligent multimedia interfaces. The knowledge underlying multimedia presentations using “Live Information” in a multimedia framework, where multilayered empirical techniques with graphical interfaces are applied, were presented in in the author’s projects on spacecraft navigation (Nourani 1996d) and terrain logics at IV-98, DaimlerBenz, Stuttgart 1998. Intelligent active multimedia databases are treated in the author’s projects since 1998 and are amongst the areas.

6. Context A preliminary overview to context abstraction and meta-contextual reasoning is presented. Abstract computational linguistics (Nourani 1996b) with intelligent syntax, model theory and categories is presented in brief. Designated functions define agents, as in artificial intelligence agents, or represent languages with only abstract definition

300

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

known at syntax. For example, a function Fi can be agent corresponding to a language Li. Li can in turn involve agent functions amongst its vocabulary. Thus context might be defined at Li. An agent Fi might be as abstract as a functor defining functions and context with respect to a set and a linguistics model as we have defined. Generic diagrams for models are defined as yet a second order lift from context. The techniques to be presented have allowed us to define a computational linguistics and model theory for intelligent languages. Models for the languages are defined by our techniques in (Nourani 1996a, 1987b) KR and its relation to context abstraction is defined in brief. A computational linguistics with intelligent syntax and model theory is defined by (Nourani 1996b,97a). Intelligent functions can represent agent functions, as artificial intelligence agents, or represent languages with definitions know at syntax. Since the languages represented by the agent functions can have arbitrary grammars not known to the signatures defined amongst the agent set, nondeterministic syntax computing is definable by the present linguistics theory. Form and context are definable by viewing computational linguistics by agent function sets. An agent FI might be as abstract as Functors defining functions and context with respect to a set and a linguistics model as we have defined. To address the issues raised the role of context in KR and Natural Language systems, particularly in the process of reasoning is related to diagram functions defining relevant world knowledge for a particular context. The relevant world functions can transfer the axioms and relevant sentences for reasoning for a context. Further, by passing context around trees via intelligent syntax trees the locality burden is lifted from the deductive viewpoint. A formal computable theory can be defined based on the functions defining computable models for a context and the functions carrying context around. For the VAS (Nourani 1997b) context foundations it is indeed possible to decrease the computational complexity of a formal system by the means of introducing context? Context localizes relevant worlds and specific computable functions define the world. Thus extraneous deductions are instant credits reducing complexity. The “what is context” question is reviewed in Section 3.3 from Section 4 on we explore relations between contexts. Decontextualization is possible and might be necessary to address structural deductions. Meta-contextual reasoning and a brief view to defining inter-context relations are introduced further on Intellligent languages were presented in brief ib the author;s pubrlications. Since the function symbols appearing might be invented by an activated agent without being defined in advance, intelligent Syntax allows us to program with nonderministic syntax. The parsing problems are quite challenging. Trees connect by message sequences hence carry parsing sequences with them. Thus the present computational linguistics theory is a start to Programming with VAS (Nourani 1997b ) and Nondeterminitic Syntax. Other agent language projects are reported at(Finin, Fritzon, et al. 1997). We have defined intelligent context free grammars in (Nourani 1997b) as follows. A preliminary parsing theory might be defined once we observe the correspondence between String Functions and context. In the papers referred to, for example, the starting refrence on Slalom computing, the author presented computing with intelligent trees and objects, where intelligent tree rewriting as a formal algebraic and model-theoretic computing technique might be defined from the abstract syntax trees and language constructs. The generalized diagrams were defined by this author to encode the model-theoretic semantics of a language from its abstract syntax. The techniques present language designs with linguistics constructs that make it easier to identify G-diagram models and define automatic implementations from abstract syntax. There is a theory in principle for building models from syntax for first order logic. However, the computing enterprise requires more general techniques

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

301

of model construction and extension, since it has to accommodate dynamically changing world descriptions and theories. The models to be defined are for complex computing phenomena for which we define. 6.1. Meta-Contextual Reasoning What is context? Is context an inherent characteristic of natural language that ultimately decides the formal power of natural language? The abstract linguistics put forth by our linguistics abstraction has surprising implications. Utterance rich with abstractions, metaphors and string intelligent functions, i.e., functions and functors transcending context, is definable by a context free grammar. Abstract syntax and intelligent models are further presented. Computing with intelligent trees (Nourani 1996a), G-diagrams for their models, and D categories are introduced in our mathematics projects published at ASL 1996-on, and applied to meta-contextual reasoning. Meta-contextual reasoning is defined by lifting from syntax and clausal theories to proof theory with G-diagrams for intelligent trees D categories- categories for models definable by G-diagrams. Proof abstraction and planning with free proof trees (Nourani 1995c, Nourani-Hoppe 1994) are another technique for meta-contextual reasoning (Nourani 1999b). Relations between contexts can be defined by what context relevant functions are applied as to the context they correspond to and the context in which they appear. Intelligent signature functions transferring context around also define inter-context relations. A computer system can automatically infer the relation between some given set of contexts from the inter-context relevant functions. 6.2. KR, Models, and Context Defining a category from the generalized diagram below is a second order lift from context. The G-diagram D defines a linguistics abstraction from content from which a linguistics model might be defined for reasoning. Abstract model theory as a second order lift is defined by a category D. The D category is the category for models definable form D. Knowledge representation has two significant roles: to define a model for the AI world, and to provide a basis for reasoning techniques to get at implicit knowledge. Diagrams are the set of atomic and negated atomic sentences that are true in a model. Generalized diagrams are diagrams definable by a minimal set of functions such that everything else in the models closure can be inferred, by a minimal set of terms defining the model. Thus providing a minimal characterisation of models, and a minimal set of atomic sentences on which all other atomic sentences depend. Our primary focus will be the relations amongst KR, AI worlds, and the computability of models. To keep the models which need to be considered small and to keep a problem tractable, such that the models could be computable and within our reach, are important goals (Nourani 1994). We show that we can apply G-diagram functions to localize reasoning to the worlds affected by some relevant functions to a specific reasoning aspect. 6.3. Diagrams and Incomplete Knowledge In this section we extend the notion of generalized diagram (G-diagram) to include plausibility and nondeterminism for planning and for representation of possible worlds. An extended notion of G-diagram can encode possible worlds to capture the “maxi-

302

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

mally complete” idea and can be used for model revision and reconstruction. By assigning a plausibility ranking to formulas one can set a truth limit ordinal t as the truth threshold. These notions of diagram are applied by way of example to planning such that the notions of computations with diagrams and free proof trees can be illustrated. A nondeterministic diagram is a diagram with indeterminate symbols instead of truth values for certain formulas. For example, (nondeterministic diagram) = {p(c),q(c),p(f(t)),q(f(c)), q(f(f(c))),I_q(f(s)))}, t is as defined by induction before; and I_q(f(s)) = I_q for some indeterminate symbol I_q, for {s = t sub n, n >= 2}. Formulas with plausibility ranking less than t would be assigned ‘T’ and the other formulas would be assigned ‘F’. Thus (Nourani 1988,91) defined the notion of a plausible diagram, which can be constructed to define plausible models for revised theories. In practice, one may envision planning with plausible diagrams such that certain propositions are deliberately left indeterminate to allow flexibility in planning. In (Nourani 1991) nondeterministic diagrams were defined by assigning an undefined “X” symbol to predicates in the diagram whose truth values are not known at each stage of planning. Such extensions to the usual notion of diagram in model theory are put forth in (Nourani 1988, 1991). That approach was one method of avoiding the computational complexity and computability problems of having complete diagrams. Truth maintenance and model revision can all be done by a simple reassignment to the diagram. The canonical model of the world is defined directly from the diagram. Generalized diagrams are shown to be an encoding for a minimal efficient knowledge representation technique applied to define relevant world models and implement reasoning trees. We have further shown how by defining predictive diagrams, partial deduction and abduction could be represented model-theoretically. We have also applied the techniques to proof abstraction and other related problems elsewhere.

7. Multiagent Visual Planning 7.1. Visual Context and Objects The visual field is represented by visual objects connected with agents carrying information amongst objects about the field, and carried onto intelligent trees for computation. Intelligent trees compute the spatial field information with the diagram functions. The trees defined have function names corresponding to computing agents. The computing agent functions have a specified module defining their functionality. The balloons are visual objects, the squares agents, the dotted lines the message paths. Multiagent spatial vision techniques are introduced in (Nourani 1998a,b). The duality for our problem solving paradigm (Nourani 1991a,95a,95b) is generalized to be symmetric by the present paper to formulate Double Vision Computing. The basic technique is that of viewing the world as many possible worlds with agents at each world that compliment one another in problem solving by cooperating. An asymmetric view of the application of this computing paradigm was presented by the author and the basic techniques were proposed for various AI systems (Nourani 1991a). The double vision computing paradigm with objects and agents might be depicted by the following figure. For computer vision (Winston 1975), the duality has obvious anthropomorphic

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

303

Figure 3. Agents and Visual Objects.

Figure 4. Multiagent Multi-board Computing.

parallels. The object co-object pairs and agents solve problems on boards by cooperating agents. 7.2. Multiagent Visual Planning The co-operative problem solving paradigms have been applied ever since the AI methods put forth by Hays-Roth et al. (1985). See (Nii 1986). The muliagent multiboard techniques due to (Nourani 1995a), see next section. There are object co-objects depicted with circles on a rectangle, the boards are depicted further down. The agents are squres with dotted message indications. The BID model has to be enhanced to be applicable to intelligent multimedia. Let us start with an example multi-board model where there multiagnt computations based on many boards, where the boards corresponds to either virtual possible worlds or to alternate visual views to the world, or to the knowledge and active databases. The board notion is a generalization of the Blackboard problem solving model (HaysRoth 1985), (Nii 1986). The blackboard model consists of a global database called the blackboard and logically independent sources of knowledge called the knowledge sources. The knowledge sources respond opportunistically to the changes on the blackboard. Starting with a problem the blackboard model provides enough guidelines for sketching a solution. Agents can cooperate on a board with very specific engagement rules not to tangle the board nor the agents. The multiagent multi-board model, henceforth abbreviates as MB,

304

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

is a virtual platform to an intelligent multimedia BID agent computing model. We are faced with designing a system consisting of the pair , where IM-BID is a multiagent multimedia computing paradigm where the agents are based on the BID model. The agents with motivational attitudes model is based on some of the assumptions described as follows. Agents are assumed to have the extra property of rationality: they must be able to generate goals and act rationally to achieve them, namely planning, replanting, and plan execution. Moreover, an agent’s activities are described using mentalistic notions usually applied to humans. To start with the way the mentalistic attitudes are modulated is not attained by the BID model. It takes the structural IM-BID to start it. The preceding sections on visual context and epistemics have brought forth the difficulties in tackling the area with a simple agent computing model. The BID model does not imply that computer systems are believed to actually “have” beliefs and intentions, but that these notions are believed to be useful in modeling and specifying the behavior required to build effective multi-agent systems, for example (Dennet 1996). The first BID assumption is that motivational attitudes, such as beliefs, desires, intentions and commitments are defined as reflective statements about the agent itself and about the agent in relation to other agents and the world. These reflective statements are modeled in DESIRE in a meta-language, which is order sorted predicate logic. At BID the functional or logical relations between motivational attitudes and between motivational attitudes and informational attitudes are expressed as metaknowledge, which may be used to perform meta-reasoning resulting in further conclusions about motivational attitudes. If we were to plan with BID with intelligent multimedia the logical relations might have to be amongst worlds forming the attitudes and event combinations. For example, in a simple instantiation of the BID model, beliefs can be inferred from meta-knowledge that any observed fact is a believed fact and that any fact communicated by a trustworthy agent is a believed fact. With IM_BID, the observed facts are believed facts only when a conjunction of certain worlds views and evens are in effect and physically logically visible to the windows in effect. Since planning with IM_BID is at times with the window visible agent groups, communicating, as two androids might, with facial gestures, for example (Picard 1998). In virtual or the “realworld” AI epistemics, we have to note what the positivists had told us some years ago: the apparent necessary facts might be only tautologies and might not amount to anything to the point at the specifics. Philosophers have been faced with challenges on the nature of absolute and the Kantian epistemtics (Kant 1990) (Nourani 1999a) for years. It might all come to terms with empirical facts and possible worlds when it comes to real applications. A second BID assumption is that information is classified according to its source: internal information, observation, communication, deduction, assumption making. Information is explicitly labeled with these sources. Both informational attitudes (such as beliefs) and motivational attitudes (such as desires) depend on these sources of information. Explicit representations of the dependencies between attitudes and their sources are used when update or revision is required. A third assumption is that the dynamics of the processes involved are explicitly modeled. A fourth assumption is that the model presented below is generic, in the sense that the explicit meta-knowledge required to reason about motivational and informational attitudes has been left unspecified. To get specific models to a given application this knowledge has to be added. A fifth assumption is that intentions and commitments are defined with respect to both

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

305

Figure 5. Virtual Trees – from the author’s abstract art gallery, January 2005.

goals and plans. An agent accepts commitments towards himself as well as towards others (social commitments). For example, a model might be defined where an agent determines which goals it intends to fulfill, and commits to a selected subset of these goals. Similarly, an agent can determine which plans it intends to perform, and commits to a selected subset of these plans. Most reasoning about beliefs, desires, and intentions can be modeled as an essential part of the reasoning an agent needs to perform to control its own processes. The task of belief determination requires explicit meta-reasoning to generate beliefs. Desire determination: Desires can refer to a (desired) state of affairs in the world (and the other agents), but also to (desired) actions to be performed. Intention and commitment determination: Intended and committed goals and plans are determined by the component intention_and_commitment_determination. This component is decomposed into goal_determination and plan_determination. Each of these subcomponents first determines the intended goals and/or plans it wishes to pursue before committing to a specific goal and/or plan. 7.3. Virtual Trees, VR Computing, and Computational Illusion The preliminaries to VR computing logic are presented since Summer Logic Colloquium, Prague (Nourani 1997). Theorem 6.1 Soundness and Completeness- Morph Gentzen Logic is sound and complete. Proposition Morph Gentzen and Intelligent languages provide a sound and complete logical basis to VR. A virtual tree, or virtual proof tree is a proof tree that is constructed with agent languages with free Skolem functions. In the present paper we also instantiate proof tree leaves with free Skolemized trees. Thus virtual trees are substituted for the leaves.

306

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

In the present approach, as we shall further define, leaves could be virtual trees. By a virtual tree we intend a term made of constant symbols and Skolem functions terms A plan is a sequence of operations in the universe that could result in terms that instantiate the truth of the goal formulas in the universe. That’s what goes on as far as the algebra of the model is concerned. It is a new view of planning prompted by our method of planning with GF-diagrams and free Skolemized trees. It is a model-theoretic view. Proof-theoretically a plan is the sequence of proof steps that yields the proof for the goal formula. The proof theoretic view is what the usual AI literature presents. The planning process at each stage can make use of GF-diagrams by taking the free interpretation, as tree-rewrite computations, for example, of the possible proof trees that correspond to each goal satisfiability. The techniques we have applied are to make use of the free Skolemized proof trees in representing plans in terms of generalized Skolem functions. In planning with GFdiagrams that part of the plan that involves free Skolemized trees is carried along with the proof tree for a plan goal. Proofs can be abstracted by generalizing away from constants in the proof. Thus, such a generalized proof can be defined by a whole class of minimal diagrams. This process is usually realized via partial deduction, which can be regarded as the proof-theoretical way of abducing diagrams whose littorals are necessary conditions for the proof. We want to present a formal relation between partial deduction and abduction from a model-theoretical point of view. However, it is not clear yet how PD can be given a model-theoretical semantics. This is one reason why the formulation of nonmonotonic reasoning presented at the author’s publications could be applicable here. In the present approach, as we shall further define, leaves could be free Skolemized trees. By a free Skolemized tree we intend a term made of constant symbols and Skolem function terms. By dropping the assumption that proof-tree leaves get instantiated with atomic formulas, we get a more general notion of a proof, which is usually called “partial deduction”. Partial deduction usually computes from a formula and a theory an existential quantified diagram. Existentially quantified diagrams carry a main deficit- the Skolemized formulas are not characterized. Hilbert’s epsilon symbol may be applied to solve this problem. Our projects apply notion of a predictive diagram to provide a practical model-theoretic characterization for practical computations with incomplete knowledge with virtual proof trees. We presented a model-theoretic counterpart to the above with predictive diagrams (Nourani-Hoppe 1994), (Nourani 2003). We then define Hilbert models to handle the proof-model problems further on. The idea is that if the free proof tree is constructed then the plan has a model in which the goals are satisfied. The model is the initial model of the AI world for which the free Skolemized trees were constructed. Thus we had stated the Free Proof Tree Sound Computing Theorem. Theorem 6.2 For the virtual proof trees defined for a goal formula from the G-diagram there is a standard model satisfying the goal formulas. It is the canonical model definable from the generic diagram. The IM Morphed Computing Logic for computing for multimedia are new projects with important computing applications. The basic principles are a mathematical logic where a Gentzen or natural deduction systems is defined by taking arbitrary structures and multimedia objects coded by diagram functions. Multimedia objects are viewed as syntactic objects defined by functions, to which the deductive system is applied. Thus

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

307

Figure 6. Morphs and Trans-Morphs.

we define a syntactic morphing to be a technique by which multimedia objects and hybrid pictures are homomorphically mapped via their defining functions to a new hybrid picture. Functorial topological structures can be defined without difficulty. The deduction rules are a Gentzen system augmented by Morphing, and Trans-morphing. The logical language has function names for hybrid pictures. The MIM Morph Rule – An object defined by the functional n-tuple can be Morphed to on object defined by the functional n-tuple , provided h is a homomrphism of abstract signature structures (Nourani 1993c). The MIM TransMorph Rules- A set of rules whereby combining hybrid pictures p1,...,pn defines an Event {p1,p2,...,pn} with a consequent hybrid picture p. Thus the combination is an impetus event. By trans-morphing hybrid picture’s corresponding functions a new hybrid picture is deduced. The techniques can be applied to arbitrary topological structures structures. The languages and MIM rules are applied to hybrid picture’s corresponding functions a new hybrid picture is deduced. The techniques can algebraic structures. The deductive theory is a Gentzen system in which hybrid pictures are named by parameterized functions; augmented by the MIM morph and transmorph rules. The Model theory is defined from Intelligent syntax languages. A computational logic for intelligent languages is presented in brief with a soundness and completeness theorem. The idea is to do it at abstract models syntax trees without specifics for the shapes and topologies applied. We start with a wel-kown infinitary logic language Lω1,ω and further on might apply well-behaved infinitary languages.

8. Conclusion T

We have thus seen the development to bases for a Haptic logic addressing the computational questions on emotional intelligence that appear to effect decision making, creativity, planning, perception, and mood-congruent memory retrieval, with precise computing and cognitive models. Competitive game models and multiagent learning were introduced Then emulating human emotion can be effective to help understand intelligence. Are intelligent decisions based on emotions? If there is a Gestalt model for the

308

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

world decided on, the answer might be affirmative. From our published perspective we are where the objects are described with languages as Frege intended, modeled by structures, which can be examined by Kan’t transcendental Idealism, and their computability and reducibility areas Hilbert arithmetized. Hence there is a systematic basis to carryout concept-object descriptions for machine discovery. A premise to an illusion logic is developed. The foundations are applied to present a brief on Computational Illusion, affective computing, and virtual reality. Visual context and objects are imporant computational areas for the above where the morph Gentzen computing logic allows us to accomplish the above goals.

References [1] Nourani, C.F. 1996a Slalom Tree Computing – A Computing Theory For Artificial Intelligence, June 1994 (Revised December 1994), A.I. Communication Volume 9, Number 4, December 1996, IOS Press, Amsterdam. [2] Nourani, C.F. 1997, “MIM Logik, December 1997, Summer Logic Colloquium, Prague, July 1998. www.math.cas.cz/~lc98/abstracts/Nourani.html. [3] Badder, F., M. Buchheit, B. Hollunder, 1996, Cardinlity Restrictions on Concepts, AI, 1996. [4] Barwise 1985b, Barwise, J., Notes on Situation Theory and Situation Semantics, CSLI Summer School, Stanford, LICS, July 1985. [5] Bratman, M.A., 1987, Intentions, Plans, and Practical Reason, Harvard University Press, Cambridge, MA. [6] Brazier, F.M.T., Dunin-Keplicz, B., Jennings, N.R. and Treur, J. (1997) DESIRE: modelling multiagent systems in a compositional formal framework, International Journal of Cooperative Information Systems, M. Huhns, M. Singh, (Eds.), special issue on Formal Methods in Cooperative Information Systems, vol. 1. [7] Brazier, F.M.T., Treur, J., Wijngaards, N.J.E. and Willems, M. (1995). Temporal semantics of complex reasoning tasks. In: B.R. Gaines, M.A. Musen (Eds.), Proc. of the 10th Banff Knowledge Acquisition for Knowledge-based Systems workshop, KAW’95, Calgary: SRDG Publications, Department of Computer Science. [8] Brazier, F.M.T., Jonker, C.M., Treur, J., (1996). Formalisation of a cooperation model based on joint intentions. In: Proc. of the ECAI’96 Workshop. [9] Brazier, F.M.T., Treur, J. (1996). Compositional modelling of reflective agents. In: B.R. Gaines, M.A. Musen (Eds.), Proc. of the 10th Banff Knowledge. [9.1] Cohen, P.R. and Levesque, H.J. (1990). Intention is choice with commitment,Artificial Intelligence 42, pp. 213–261. [10] Dunin-Keplicz, B. and Treur, J. (1995). Compositional formal specification of multi-agent systems. In: M. Wooldridge and N.R. Jennings, Intelligent Agents, Lecture Notes in Artificial Intelligence, Vol. 890, Springer Verlag, Berlin, pp. 102–117. [11] Cooper et al. 1996, Cooper, R., J. Fox, J. Farrington, T. Shallice, A Systematic Methodology For Cognitive Modeling AI-85, 1996, 3–44. [12] Dennett, D. (1987). The Intentional Stance, MIT Press, Cambridge, MA. [13] AI 80, AI- Special Issue On Nomonotonic Logic, vol. 13, 1980. [14] Erman, D.L., B. Hays-Roth, F., V.L. Lesser, and D.R. Reddy 1980, The HEARSAY-II speacch understadnign system: Integrating Knowledeg to resolve uncertainty: ACM Computing Survey 12:213–253 [15] Didday, E. 1990, Knowledge Representation and Symbolic Data Analysis, NATO AIS series Vol. F61, edited by M Schader and W. Gaul, Springer-Verlag, Berlin. [16] Nourani, C.F. 2002, “Virtual Tree Computing, Meta-Contextual Logic, and VR”, ASL Spring 2002, Seattle WA, March, BSL Vol. 8. No. 3, ISSN 1079–8986. [17] Finin, T et al. 1994, Finin, T., R. Fritzson, D. McKay, and R. McEntire: KQML as an Agent Communications Language, Proceedings of the Third Internationla Conference on Information and Knowledge Management, ACM Press, November 1994. [18] Fodor, J.A. 1975, The Language of Thought, T.W. Cromwell Company, New York, N.Y. [19] Ford, K.M., C. Glymour, and P.J. Hayes 1995, Android Epistemmology, AAA/MIT Press.

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

309

[20] Sphon, W. 1988, Ordinal Conditional Functions: A dynamic Theory of Epistemic States, In Harper W.L. and Skyrms, B. (eds.) Causation, in decision, belief change, and statistics, Klawer Academic Publishers, 0105–134, 1988. [21] Gen-Nils 87, Genesereth, M. and N.J. Nilsson, 87, Logical Foundations of Artificial Intelligence, Morgan-Kaufmann, 1987. [21] Hays-Roth, B. (1985) Blackbaord Architecture for control, Journal of AI 26:251–321. [22] Heidegger 1962, Heidegger, M., Die Frage nach dem Ding, Max Niemeyer Verlag, Tubingen., 1962. [23] Hintikka 1961, Hintikka, J. 1961,Knowledge and Belief, Cornell University Press, Ithaca, N.Y. [24] Kant, I., 1990, Critique of Pure Reason, Trasnlated by J.M.D. Meiklejohn. [25] Kinny, D., Georgeff, M.P., Rao, A.S. (1996). A Methodology and Technique for Systems of BID Agents. In: W. van der Velde, J.W. Perram (Eds.), Agents Breaking Away, Proc. 7th European Workshop on Modelling Autonomous Agents in a Multi-Agent World, MAAMAW’96, Lecture Notes in AI, vol. 1038, Springer Verlag. [26] Kleene, S. 1952, Introduction to Metamathematics, 1952. [27] Koehler 1996- Koehler, J., Planning From Second Principles, AI 87. [28] Lambek, J. 1965 The Mathematics of Sentence Structure, American Mathematical Monthly, 65, 154–170. [29] Nourani, C.F. 197, “Syntax Trees, Intensional Models, and Modal Diagrams For Natural Language Models,” Revised July 1997. Uppsala Logic Colloquium, August1998, Uppsala University, Sweden. [30] Konolige, K. 1984, Belief and Incompleteness,” Stanford CSLI-84-4, Ventura Hall, March 1984. [31] Kripke, S.A. 1963, Semantical Analysis of Modal Logics, Zeitschrift fuer Mathematische Logik und Grundlagen der Mathematik, vol. 9: 67–69. [32] Lakemeyer, G.: Limited Reasoning in First Order KB with Full Introspection, AI, 84, 1996. [33] Nourani, C.F. 1999, Creating Art and Motion Pictures with Intelligent Multimedia, 1997, Published as a chapter in the Intelligent Multimedia Textbook the author wrote. Intelligent Multimedia New Computing Techniques, Design Paradigms, and Applications Prelimianry edition August 1999, Berkeley. Available from Lulu Press, North Carolina, http://www.lulu.com/CrisFN. [34] Mosetic, I. and C. Holsbaur, 1997, “Extedning Explanation Based Generalization by Abstraction Operators,” Machine Learning EWSL-91, Springer-Verlag, LNAI, vol. 482. [35] Maybury, M.T. 1998 Intelligent Multimedia Interfaces (Edited by) MIT Press, 1997–98, ISBN 0-26263150-4. [36] Moore, R.C. 1980, Reasoning About Knowledge and Action, AI Center Technical Note 191, SRI International Menlo Park, California, 1980. [37] Nii, P.H. 1986, Blackboard Systems: the Blackboard Model of Problem Solving and The Evolution of Blackboard Architectures, The AI Magazine, Summer 1986, 38–53. [38] Nourani, C.F. 1984, Equational Intensity, Initial Models, and AI Reasoning, Technical Report, 1983: A: Conceptual Overview, in Proc. Sixth European Conference in Artificial Intelligence, Pisa, Italy, September 1984, North-Holland. [39] Nourani, C.F. 1994a, “Dynamic Epistemic Computing,” 1994. Preliminary brief at the Summer Logic Colloquium, Claire-Mont Ferrand, France. [40] Nourani, C.F. 1997, “Computability, KR and Reducibility For Artificial Intelligence Problems,” February 25, 1997. Brief ASL, Toronto, May 1998. BSL, Vol. 4. Number 4, December 1998. [41] Nourani, C.F. 1993b, “Automatic Models From Syntax,” Proceedings XV Scandinavian Linguistics Conference, Oslo, Norway, January 1995. [42] Nourani, C.F. 1987, Diagrams, Possible Worlds, and The Problem of Reasoning in Artificial Intelligence, Proc. Logic Colloquium, 1988, Padova, Italy, Journal of Symbolic Logic. http://www.logic.univie.ac.at Abstract at author’s name. [43] Nourani, C.F. 1984, Equational Intensity, Initial Models, and Reasoning in AI: A conceptual Overview, Proc. Sixth European AI Conference, Pisa, Italy, North-Holland. [44] Nourani, C.F. 1991, Planning and Plausible Reasoning in Artificial Intelligence, Diagrams, Planning, and Reasoning, Proc. Scandinavian Conference on Artificial Intelligence, Denmark, May 1991, IOS Press. [45] Nourani, C.F. 1995c, “Free Proof Trees and Model-theoretic Planning,” Proceedings Automated Reasoning AISB, Sheffield, England, April 1995. [46] Nourani, C.F., 1999a Idealism, Illusion and Discovery, The International Conference on Mathematical Logic, Novosibirsk, Russia, August 1999. [47] Nourani, C.F. 1998d, “Syntax Trees, Intensional Models, and Modal Diagrams For Natural Language Models,” Revised July 1997, Proceedings Uppsala Logic Colloquium, August 1998, Uppsala University, Sweden.

310

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

[48] Nourani, C.F. and K.J. Lieberherr 1985, Data Types, Direct Implementations, and KR, Proc. HICSS KR Track, 1985, Honollulu, Hawaii. [49] Nourani, C.F. and Th. Hoppe 1994, “GF-Diagrams for Models and Free Proof Trees,” Proceedings the Berlin Logic Colloquium, Homboldt Universty, May 1994. [50] Nourani, C.F., 1995c, Free Proof Trees and Model-theoretic Planning, February 23, 1995, Automated Reasoning AISB, England, April 1995. [51] Nourani, C.F. 2003, “KR, Data Modeling, DB and KB Predictive Scheme Completion,” A version published at International Systems and Cybernetics, Florida, July 2003. [52] Nourani, C.F. 1996b, “Descriptive Computing, February 1996,” Summer Logic Colloquium, July 1996, San Sebastian, Spain. Recorded at AMS,April 1997, Memphis. [53] Nourani, C.F. 1993a , “Abstract Implementation Techniques for A.I. By Computing Agents, A Conceptual Overview,” Technical Report, March 3, 1993, Proceedings SERF-93, Orlando, Florida, November 1993. Published by the Univeristy of West Florida Software Engineering Research Forum, Melbourne, FL. [54] Nourani, C.F. 1995d, “Language Dynamics and Syntactic Models,” August 1994, Prceedings. Nordic and General Linguistics, January 1995, Oslo University, Norway. [55] Nourani, C.F. “Versatile Abstract Syntax Meta-Contextual Logic and VR Computing,” 36th Lingustische Kolloquium, Austria Proceedings of the 35th Colloquium of Linguistics, Innsbrucke. Europa Der Sprachen: Sprachkopetenz-Mehrsprachigeit-Translation, TIEL II: Sprache und Kognition, Sonderdruc 2003, Lew N. Zybatow (HRSG). [56] Nourani, C.F. 1998c, “Visual Computational Linguistics and Visual Languages,” 1998, Proceedings 34th International Colloquium on Linguistics, University of Mainz, Germany, September 1999. [57] Nourani, C.F. 1998d, Morph Gentzen, K.R., and World Model Diagrams April 2, 1998. Automated Deductions and Geometry, ETH Zurich, September 2000. [57.1] Nourani, C.F. 1999a, “Multiagent AI implementations an emerging software engineering trend,” Engeineering Applications of AI 12: 37–42. [58] Nourani, C.F. 1999b, “Functorial Syntax and Paraconsistent Logics,” ASL, New Orelans, May 1999, BSL 1999. [59] Nourani, C.F., 1995a, “Double Vision Computing,” IAS-4, Intelligent Autonomous Systems, Karlsruhe, April 1995, Germany. [60] Nourani, C.F. 1995b, “Multiagent Robot Supervision,” Learning Robots, Heraklion, April 1995. [61] Nourani, C.F., 1994b, “Towards Computational Epistemology-A Forward,” Proceedings Summer Logic Colloquium, July 1994, Clermont-Ferrand, France. [62] Nourani, C.F., 1996b, “Linguistics Abstraction,” April 1995, Brief Overview, Proceedings ICML96, International Conference on Mathematical Linguistics, Catalunya, Tarragona, Spain. [63] Nourani, C.F.1996b, “Autonomous Multiagent Double Vision SpaceCrafts,” AA99- Agent Autonomy Track, Seattle, WA, May 1999. [64] Picard, R.T. 1998, Affective Computing, TR#321, MIT Media Lab. 1998. [65] Picard, R.W. 1999a, Affective Computing for HCI, Proceedings of HCI, Munich, Germany, August 1999. [66] Picard, R.W. and G. Cosier 1997, Affective Intelligence – The Missing Link, BT Technology J. Vol. 14, No. 4, 56–71, October. [67] Quine 52, Quine, W.Van Orman,Word and Object, Harvard University Press. [68] Rao, A.S. and Georgeff, M.P. (1991). Modeling rational agents within a BID architecture. In: R. Fikes and E. Sandewall (eds.), Proceedings of the Second Conference on Knowledge Representation and Reasoning, Morgan Kaufman, pp. 473–484. [69] Schubert, L.K. 1976, Extending the Expressive Power of Semantic Nets, AI 7, 2, 163–198. [70] Shoham, Y. and Cousins, S.B. (1994). Logics of mental attitudes in AI: a very preliminary survey. In: G. Lakemeyer and B. Nebel (eds.) Foundations of Knowledge Reproham, Y. (1993). Agent-oriented programming, Artificial Intelligence 60. [71] Velde, W. van der and J.W. Perram J.W. (Eds.) (1996). Agents Breaking Away, Proc. 7th European Workshop on Modelling Autonomous Agents in a Multi-Agent World, MAAMAW’96, Lecture Notes in AI, vol. 1038, Springer Verlag. [72] Williams, M. 1994, “Explanation and Theory Base Transmutations,” Proceedings 11th European Conferenceon AI, Amsterdam, John Wiley and Sons Ltd. 346–350. [73] Woodesentation and Reasoning, Springer Verlag, pp. 296–309. [74] Nourani, C.F. 1999c, “TInfinitary Multiplayer Games, “March 8, 1999, ASL SLC, Utrecht. http://www.logic.univie.ac.at Abstracts on authtor’s name. www.folli.uva.nl/CD/1999/library/logic% 20colloquium%2099/abstracts/nourani2.ps. P

P

P

P

C.F. Nourani / A Haptic Computing Logic – Agent Planning, Models, and Virtual Trees

311

[75] Wooldridge, M. and Jennings, N.R. (1995). Agent theories, architectures, and languages: a survey. In: M. Wooldridge and N.R. Jennings, Intelligent Agents. (1993) 51–92. [76] Nourani, C.F. 2004, Intelligent Multimedia Computing Science Business Interfaces, Wireless Computing, Databases, and DataMines. ISBN: 1-58883-037-3, www.aspbs.com/multimmedia.html *Personale-mail [email protected].

312

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

ESF SCSS Exploratory Workshop on

Affective and Emotional Aspects of Human-Computer Interaction: Emphasis on Game-Based and Innovative Learning Approaches 23–25 September 2004 in Pörtschach, Carinthia, Austria Workshop Proposer Maja Pivec, Ph.D. FH JOANNEUM Gesellschaft mbH (University of Applied Sciences JOANNEUM) Alte Poststraße 152, A-8020 Graz Tel.: +43 316 5453 8623 Fax: +43 316 5453 8601 E-mail: [email protected] Homepage: http://www.fh-joanneum.at Workshop Co-Organisers Frank Thissen, Ph.D. University of Applied Sciences Stuttgart, Germany [email protected] Konrad Baumann, Ph.D. University of Applied Sciences JOANNEUM, Graz, Austria [email protected] About the Workshop The expected result from the intensive exchange of ideas in the workshop is a list of characteristics of affective and learner centred adaptive learning environments and contents. Selected workshop contributions that reflect the state of the art and future research trends, along with significant workshop results, are published in this book. The

ESF SCSS Exploratory Workshop Description

313

workshop contributions are available from the workshop web site: http://informationsdesign.fh-joanneum.at/images/esf_workshop_092004/index.htm. In order to provide opportunities for acquiring new personal contacts the workshop started with the introduction of participants. Most of the workshop’s activities were carried out in a homogenous group. The workshop was based on very intensive information exchange and discussions. On the one hand this was made possible because of the limited number of participants, and on the other hand the schedule provided a necessary structure for this to happen. Only half a day was spent in breakout groups, which had been previously defined by all participants. The workshop continued in three working groups and focused on the following topics: • • •

Design (of) emotions Games and learning Intercultural differences

The discussion relating to design of emotions concluded that design always has an emotional dimension (managing emotions). There are different applications and goals. The central question is What is the role of emotion?, with sub-questions in areas emotional demands, products, services, e.g. in education and art and performances where emotional feelings are in foreground. When we talk of emotional design we have the emotions of creators vs. the emotions of users, in the sense that the emotions of creators will be transformed into emotions of users. When we consider the design of products vs. the design of educational tools, we have to see what are the differences and requirements. Another issue is to look at emotions as additional channel of communication i.e. dynamics of emotions in communication. The group that discussed games and learning focused on games as motivators of learning. The discussion tackled questions as follows: 1. 2.

3.

4.

Does it? Do games motivate learning? The answer can be found in the market evolution. How do games motivate? Are there individual differences in acceptance of games? How do experience in game-playing and emotional attachment contribute to motivation for learning? Can games be applied to problem-based learning? Can games be used for approaches which “pull the knowledge” i.e. learning on demand versus pushing the knowledge to the learner? Can games contribute to innovation in learning in university and lifelong learning contexts? Could they de-motivate? It is very difficult to arrive at an optimum challenge so as to have fun when playing game. In case that the game is too difficult, it can soon be a frustration for the player. Games could be motivators due to the emotional attachment of the players. Who needs to be convinced? The acceptance of games for learning might vary in different cultures, one has to research that issue and produce guidelines that would be helpful for various populations, politics, teachers, pupils, the game industry, customers, etc. so that they can accept a game-based learning approach.

The group discussion on intercultural differences focused on topics such as interface design from the viewpoint of different culture, interfaces that apologize and how

314

ESF SCSS Exploratory Workshop Description

they are evaluated in different cultures. The next research steps will be directed toward trying out applications of game environments for collaboration and empirical evaluation studies. Ideas for further research and joint projects were directed toward questions of intercultural diversity of web sites, in particular: How do people from different cultures evaluate websites?, Are there culture-specific qualities of web sites?, Can you find cultural dimensions in web sites? Workshop Results • • • • •

A unique opportunity to acquire new personal contacts by bringing together leading researchers and developers from different fields of science focusing on related research questions. An exchange of research ideas from various scientific fields and outlining further research direction in the area of innovative human-computer interaction. The workshop is a networking platform that initiated new joint research projects within Europe. The workshop publication in book form that gathers and presents the workshop papers and significant workshop results along with related interesting research reports. A workshop web page, that provides essential information about the workshop (background, aims and objectives of the workshop, workshop schedule, discussions, and reported results), offers possibility to download the presentation slides as well as a list with full addresses of all participants.

Scientific Content of the Event Background, Goals and Topics Digital information technologies develop rapidly and play an important part in everyday life. They are actively implemented in the learning process, and used in formal (school, colleges, universities), informal (people to people or collaboration) and nonformal (online, distance) education. Emotions are essential for the success of the human learning processes. This wellestablished idea has been confirmed by modern cognitive science, but is neglected in most of the models and implementations of computer-based and web-based learning. Very often e-Learning means the presentation of information and material on a very rational basis. Communication between learners and trainers takes place in a much reduced form that ignores the fact that communication is always meta-communication as well. If human emotions are essential for human thinking and learning processes, virtual platforms and learning environments have to intensively consider this fact if they are to be successful. In particular, the graphical computer interface may not address humans as information processing machines, but needs to consider human emotion, passion and affects. Learning environments have to be more than simply a program, they have to be a form of rich media and to pursue whole didactic concepts. The field of computer games has become an important economic factor, comparable to the music and film industries. Analysis of the current situation in the field of the

ESF SCSS Exploratory Workshop Description

315

game industry shows that at the same time, new multiplayer environments give opportunities for interaction for thousands of simultaneous players. In January 2003 a record number of 120,000 users simultaneously played EverQuest, an online role-playing game. Its producer, SONY, “celebrated” a huge success when 430,000 subscribers joined the game last year (‘Times online’). EverQuest has a Command Centre at their San Diego office where 150 full-time customer service staff wander about the virtual game world assisting players, creating scenarios, conflicts, etc. It means that there is a huge interest in “collaborative” playing. The aim of this workshop is to bring together related research from different fields such as psychology, educational sciences, cognitive sciences, various areas of communication and human-computer interaction, user-interface design and computer science thus initiating interdisciplinary projects that will explore the fields of emotional learning settings, innovative human-computer interaction and game research combined with application development. This can assist the educational and games industry to focus on such research topics as game-based learning, game dramaturgy or interactive storytelling and to introduce these approaches into their products. The field of new HCI (human computer interface) paradigms as well as micro-games (UMTS and interactive set-top boxes) can be considered. The research reported in the workshop, on the workshop web page and in the workshop publication will play a role in future developments. Goals of the Workshop There is substantial effort and research within various disciplines that tackle communication issues, emotional aspects of human-computer interaction, communication agents and improvement of computer based learning settings and environments. Recently there have been several initiatives within the EU that focus on game development and facilitating and improving the learning process through the introduction of digital games into learning and the fostering of innovative learning paradigms such as gamebased learning. Because of the multicultural European space and the cultural differences present, there is a need to introduce an interdisciplinary and trans-European component into the research and future development of innovative learning approaches and novel ways of implementing affective and emotional human-computer interaction. Topics of the Workshop • • • • • • •

affective computing, emotional agents, emotional learning settings, game-based learning, human-computer interaction, human perception emulation, innovative learning approaches.

This page intentionally left blank

Affective and Emotional Aspects of Human-Computer Interaction M. Pivec (Ed.) IOS Press, 2006 © 2006 The authors. All rights reserved.

317

Author Index Akilli, G.K. Albert, D. Aylett, R. Bekebrede, G. Bopp, M. Brna, P. Burmester, M. Cagiltay, K. Chaffar, S. Dias, J. Dufner, A. Frasson, C. Hall, L. Heller, J. Hunger, A. Kearney, P.R. Kertz, M. Kickmeier-Rust, M.

93 165 246 136 8 237 217 93 255 246 217 255 246 165 266 38 165 165

Lian, A. Marin, B.F. Mayer, I. McKay, E. Nourani, C.F. Ochs, M. Paechter, M. Paiva, A. Parker, P.P. Razek, M.A. Schweizer, K. Sykes, J. Tüzün, H. Vala, M. Werner, S. Wilhelmsson, U. Woods, S. Zoll, C.

178 266 136 207 286 255 155 246 113 255 155 3 59 246 266 45 246 246

This page intentionally left blank