Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Intrinsic motivational effects and cognitive learning outcomes of an instructional microcomputer game Shaban, Abdullah 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1988_A2 S52.pdf [ 6.1MB ]
Metadata
JSON: 831-1.0098171.json
JSON-LD: 831-1.0098171-ld.json
RDF/XML (Pretty): 831-1.0098171-rdf.xml
RDF/JSON: 831-1.0098171-rdf.json
Turtle: 831-1.0098171-turtle.txt
N-Triples: 831-1.0098171-rdf-ntriples.txt
Original Record: 831-1.0098171-source.json
Full Text
831-1.0098171-fulltext.txt
Citation
831-1.0098171.ris

Full Text

INTRINSIC MOTIVATIONAL EFFECTS AND COGNITIVE LEARNING OUTCOMES OF AN INSTRUCTIONAL MICROCOMPUTER GAME by ABDULLAH SHABAN A DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF EDUCATION in THE FACULTY OF GRADUATE STUDIES Mathematics and Science Education We accept this dissertation as conforming to the required standard THE UNIVERSITY -OF BRITISH COLUMBIA August 1988 ® Abdullah Shaban, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of MaUrmah'cs GduJLAb'siH The University of British Columbia Vancouver, Canada Date Dc^U^ It/, MM  DE-6 (2/88) ABSTRACT The current study addresses the questions of what determines intrinsic motivation, how do the factors that determine it work, and what kinds of cognitive learning may be achieved in an intrinsically motivating environment? A microcomputer game environment, involving one instructional and one noninstructional game, was selected for the study. Two game-specific parallel tests of motivation involving the factors of Challenge, Curiosity, Control, and Fantasy were constructed. An achievement test of algebra relating to the content of the instructional game and involving the learning of Concepts, Rules, and Procedures was also constructed. In an experiment involving 134 lOth-grade students, a test of divergent feeling, measuring how creative the students feel about themselves, was administered. The subjects were then randomly assigned by gender and class to either an experimental or a control group. Following a practice session, the experimental group played each game twice and answered a test of motivation each time, while the control group played the noninstructional game twice and used worksheets twice to practice the mathematical content of the instructional game. The test of algebra was administered to all subjects after the last playing session and in the fifth week following that. The results revealed that each of the four factors of Challenge, Curiosity, Control, and Fantasy played a role in determining the intrinsic motivational effects of the games. The games did not differentiate in motivation between boys and girls or among students with different levels of perceived creativity. There were no significant differences in achievement or retention between the experimental and control groups: the worksheets were just as effective as the game in enabling the learning of ii Concepts, Rules, and Procedure on both the post-test and retention test. Gender differences in mathematics achievement, favouring boys over girls, were accounted for, in part, by the level of perceived creativity. Challenge, Control, and Fantasy correlated positively with cognitive learning. For the instructional game, there was no significant change for the factors of Challenge, Curiosity, and Fantasy; but student motivation attributed to Control increased significantly. iii TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vi List of Figures vii Acknowledgements viii I. THE PROBLEM 1 1. Introduction 1 2. Background of the Problem 2 3. The Problem 4 4. Rationale 5 5. Questions to be Answered 6 6. Definition of Terms 7 7. Scope and Delimitations of the Study 8 8. Outline of the Remainder of the Dissertation 9 II. REVIEW OF LITERATURE 11 A. Motivation 11 1. Extrinsic and Intrinsic Motivation 12 2. Rewards and Learning 14 3. Student Motivation 15 B. Learning 17 1. Learning Theories 17 a. Learning Systems 18 b. Discovery Learning 19 c. Comparison of Learning Theories 20 2. Computer-Based Learning 22 C. Simulations and Games 27 1. Simulation Games 28 2. Non-Computer Games 31 3. Computer Games 36 D. Summary 42 III. RESEARCH METHODOLOGY AND DESIGN 45 A. Research Methodology 45 1. The Games 45 a. Lode Runner™ 46 b. Mission: Algebra™ 48 2. Tests of Motivation and Learning 50 a. The Tests of Motivation 50 b. The Test of Algebra 53 3. Pilot Project 53 a. Procedures 54 b. Results 54 c. Discussion 56 iv 4. The Test of Divergent Feeling 57 B. Research Design and Procedures 58 1. Procedures 58 2. Selection of Subjects 61 3. Laboratory Setting and Procedures 62 4. Hypotheses to be Tested 62 5. Methods of Data Collection 65 6. Analyses of Data 66 IV. Results 67 A. Intrinsic Motivational Effects 67 1. Total Motivation Scores 68 2. Intra-game Subscale Motivation Scores 69 3. Inter-game Subscale Motivation Scores 71 4. Motivation, Gender, and Divergent Feeling 72 B. Cognitive Learning Outcomes 74 1. Post-test Scores 74 2. Retention Scores 76 C. Motivation and Learning 78 V. Implications of the Study 81 A. Interpretations 81 1. Intrinsic Motivational Effects 81 2. Cognitive Learning Outcomes 85 3. Motivation and Learning 88 4. Limitations 89 5. Summary 90 B. Discussion 91 1. Motivation 91 2. Learning , 94 3. Motivation and Learning 95 C. Conclusions and Recommendations 96 1. Conclusions 97 a. Intrinsic Motivation 97 b. Learning and its Correlation with Motivation 98 2. Recommendations 99 REFERENCES , 100 Appendix A 104 Appendix B 106 Appendix C 108 Appendix D 112 Appendix E 114 v List of Tables Table 1. Mission: Algebra™ Estimates of Reliability 55 Table 2. Lode Runner™ Estimates of Reliability 55 Table 3. Test of Algebra Estimates of Reliability 56 Table 4. Motivation Scale Means and Standard Deviations 67 Table 5. Comparisons of Total Motivations 68 Table 6. Intra-game Comparison of Motivations 70 Table 7. Inter-game Comparison of Motivations 71 Table 8. Lode Runner™ Motivations 73 Table 9. Mission: Algebra™ Motivations 73 Table 10. Post-test Means (Standard Deviations) 74 Table 11. Summary Table for Post-test 75 Table 12. Interaction of Gender and LDF 76 Table 13. Retention Test Means (Standard Deviations) 76 Table 14. Summary Table for Retention Test 77 Table 15. Interaction of Gender and LDF 78 Table 16. Correlation of Motivation and Learning 80 vi List of Figures Figure 1. Starting Level of Lode Runner™ 46 Figure 2. End of Level 3 of Lode Runner™ 47 Figure 3. Mission: Algebra™ Screens 49 Figure 4. Challenge Test Items 51 Figure 5. Curiosity Test Items 51 Figure 6. Control Test Items 52 Figure 7. Fantasy Test Items 52 Figure 8. Research Design and Procedures 59 Figure 9. Mean Total Motivation Scores 68 Figure 10. Interaction of Gender and LDF on the Post-test 76 Figure 11. Interaction of Gender and LDF on Retention 79 vii ACKNOWLEDGEMENTS It is my duty and pleasure to express my deep appreciation and gratitude to my thesis supervisor, Dr. Marvin Westrom, for his continuous help, valuable suggestions, and encouragement during the preparation of this dissertation. I also wish to express my appreciation and gratitude to: • Dr. James Sherrill, for his professional reviews of the manuscript and insightful comments; • Dr. David Whittaker, for his amiability, encouragement, and enthusiastic help throughout the process of conducting this study; • Dr. Raymond Corteen, for his valuable remarks and helpful suggestions regarding the content and organization of the dissertation; • the teachers and students who participated in the study and helped make it a complete success; • all my family members who supported me wholeheartedly from the beginning; • my dear wife, Hassna, who was patient and encouraging and who took good care of our son, Omar. viii I. T H E P R O B L E M 1. I n t r o d u c t i o n Bruner (1962) stated: "It is sentimentalism to assume that the teaching of life can be fitted always to the child's interests just as it is empty formalism to force the child to parrot the formulas of adult society. Interests can be created and stimulated." Bruner's statement addresses the inability of school systems to enhance and maintain children's interest in exploration and learning that starts at an early age, yet begins to fade away as these children enter the school. The statement also addresses the need for an interesting and exciting classroom environment in which the child can act, as Holt (1964) argued, with a high degree of attention, concentration, and involvement and in which the child will care most about what he/she is doing. Cognitive theorists claim that children are equipped with powerful intellectual capabilities for thought, including the ability to acquire first language and a developing ability to handle logical operations. To enhance learning, the active participation of children in their interaction with the environment should be taken into account. Continual interaction with the environment is motivated by intrinsic motives. To enhance the intrinsic interest in learning, on the other hand, some cognitive processes should be developed. Bruner (1962), for example, encouraged teachers to create a classroom environment where the child will learn by discovering things rather than learn about them. Such an environment, he suggested, may be created by letting the children exercise competence motives, which in turn may strengthen the degree to which the child gains control over his/her behavior. 1 T H E P R O B L E M / 2 The current study investigates the potential of microcomputer game environments for enhancing learning and increasing the intrinsic interest in learning. In particular, the study examines motivational effects and learning outcomes of engagement with selected microcomputer games. 2. B a c k g r o u n d of the P r o b l e m Lepper (1985) isolated four classes of concepts that have been associated with the study of intrinsic motivation. The first class deals with challenge. An activity is intrinsically motivating to the extent that it engages individuals in a process of seeking to solve problems and accomplish goals that require the exercise of valued personal skills. The second class involves curiosity. The activity will be intrinsically motivating if it provides the individual with some optimal level of surprise, incongruity, complexity, or discrepancy as a function of his/her initial skills and expectations. The third class involves control and self-determination. An activity will evoke intrinsic motivation when it provides the individual with the opportunity to exert control, to determine his/her own fate, or at least to maintain the perception that he/she is doing so. The last class of concepts associated with intrinsic motivation involves fantasy. Intrinsic motivation may be enhanced through the evocation of a playful set, identification with fictional characters, and/or involvement in a world of fantasy. Malone and Lepper (1987) labelled these four factors as individual motivations. The authors contended that interpersonal motivations, which include the concepts of cooperation, competition and recognition, are also associated with intrinsic motivation. THE PROBLEM / 3 Malone (1981a) had earlier developed a rudimentary theory of intrinsically motivating instruction based on the results of his own research on the motivating factors of microcomputer games. The theory integrates the concepts of challenge, curiosity and fantasy as determinants of intrinsic motivation. The study of intrinsic motivation, however, has received little research attention in the past few decades (Lepper & Chabay, 1985). The way people learn has, on the other hand, long been a central issue in psychological research. Gagne (1984) suggested that studying what people learn is an essential step in understanding how they learn. He further indicated that intellectual skills, which include concepts, rules, and procedures, form one of five widely accepted categories of learning outcomes. There is agreement among various learning theories that motivation initiates and directs behavior, and leads to particular responses which are directed toward achieving a specified goal (Dubin & Okun, 1973). Yet, virtually no studies have been carried out to investigate the relationship between motivation and learning (Lepper & Malone, 1987). Computer games in general can have a wide range of capabilities for instruction. One such capability is the rapid diagnosis of student errors through practice or by monitoring each child's activities and generating a list of missing skills, the acquisition of which can then be fostered. A current research goal that a computer gaming environment can help achieve is to make diagnoses that differentiate the need of the child for conceptual acquisition from the need for practice of procedural skills. . THE PROBLEM / 4 Computer games, in addition, have potential as a device for studying various motivational effects and learning outcomes in ways that would be difficult to do otherwise. Such study can be conducted in a classroom environment as well as in a laboratory setting. Eysenck (1984) suggested that laboratory research may form a useful means for investigating phenomena that could, for example, support an important theoretical principle. 3. The Problem The current study deals with two related research issues that have potential both in theory and educational applications. The first issue deals with the study of intrinsic motivation. Specifically, what determines intrinsic motivation, and how do the factors that determine it work? The popularity of video games indicates that the ever-developing microcomputer technology forms a promising setting for conducting research studies that would help understand the concept of intrinsic motivation. The second issue deals with the way that intrinsic motivation affects learning. Specifically, what kinds of learning may be achieved in an intrinsically motivating environment? Again, the micrcomputer has potential as a device in studying the relationship between motivation and learning. The present study examines the challenge, curiosity, control and fantasy aspects of microcomputer games. The study also examines the types of cognitive learning outcomes that can be achieved in a microcomputer gaming environment and the relationship between intrinsic motivation and cognitive learning in such an environment. The types of cognitive learning outcomes investigated include concepts, rules, and procedures. THE PROBLEM / 5 4. R a t i o n a l e There is a need for investigating various aspects of intrinsic motivation and creating better learning environments for children. A set of factors that determine intrinsic motivation must therefore be stipulated. Malone and Lepper (1987) suggested that use of the individual motivations of challenge, curiosity, control, and fantasy, and the interpersonal motivations of cooperation, competition, and recognition is appropriate for designing intrinsically motivating instructional environments. Focusing on the features of two microcomputer games that appear to employ individual motivations, specific hypotheses relating to the motivations were formulated. To test the hypotheses, two parallel instruments that utilize the apparent motivational aspects of the games were constructed. It seems irrelevant however to isolate the study of motivation in an instructional environment from the study of possible learning. Moreover, in the highly structured classroom, it is important for the practitioner to be aware of which kinds of learning can better be achieved by which students in a motivating environment as opposed to a traditional one. Virtually nothing is known about the relationship between motivation and learning (Lepper & Malone, 1987). It becomes necessary therefore that a systematic investigation into this fundamental issue be started. Again, specific hypotheses relating to the types of cognitive learning outcomes of playing an instructional microcomputer game and the relationship between intrinsic motivational effects and cognitive learning outcomes were formulated. An achievement test relating to the educational contents of the instructional game was also constructed. T H E P R O B L E M / 6 The independent variables of gender and perceived creativity were considered for investigating intrinsic motivational effects and cognitive learning outcomes of playing the games. Perceived creativity is measured by a test of divergent feeling which involves the factors of complexity, curiosity, risk-taking, and imagination. The obvious similarity between these factors and the individual motivations of challenge, curiosity, control, and fantasy makes relevant a study of the variability in intrinsic motivation and cognitive learning over such a trait as perceived creativity. 5. Questions to be Answered The current study seeks answers to the following questions: 1. How do the intrinsic motivational levels of players of a microcomputer game vary with the playing experience of the players? 2. How do the intrinsic motivational levels of players of a microcomputer game differ from the corresponding motivational levels of the players on another microcomputer game? 3. Do microcomputer games differentiate the levels of intrinsic motivation between boys and girls having different levels of perceived creativity? 4. Do players of an instructional microcomputer game score higher on a post-test relating to the contents of the game than non-players practicing the same content, regardless of gender or level of perceived creativity? 5. Do players of an instructional microcomputer game score higher on a retention test relating to the contents of the game than non-players practicing the same content, regardless of gender or level of perceived creativity? 6. How does the motivational level of players of an instructional microcompter game correlate with the cognitive learning outcomes of playing the game? THE PROBLEM / 7 6. Definition of Terms Motivation is defined in the current study as "an intervening variable which is used to account for factors within the organism which arouse, maintain, and channel behavior toward a goal" (Chaplin, 1979). The current study, however, deals with only one type of motivation termed intrinsic motivation. In operational terms, an intrinsically motivated activity can be defined as one with which no apparent external reward is associated, thus the reward is said to be in the activity itself. Yet if deeper understanding of intrinsic motivation is sought, the psychological processes which go along with it should be considered. Deci (1975) contended that intrinsic motivation exists (humans do perform tasks for their own sake and in the absence of external rewards). Intrinsically motivating behaviors for him are "behaviors aimed at bringing about certain internal rewarding consequences that ... are intended to bring about the feeling of competence and self-determination". The definition of intrinsic motivation stated above will be extended to include the factors, suggested by Lepper and Malone (1987), of interpersonal motivation and individual motivation. Interpersonal motivation involves cooperation, competition and recognition. Individual motivation, on the other hand, involves challenge, curiosity, control and fantasy. The interpersonal and individual motivations were suggested by the authors as guidelines for designing effective instructional microcomputer games. The individual motivations and their definitions are accepted here because they fit the purpose of the study. Yet, the questions of which factors actually determine intrinsic motivation and how do such factors work is still an open one. Extensive empirical research is needed to support the theory of intrinsic motivation introduced by the authors. THE PROBLEM / 8 Challenge involves activities that engage individuals in a process of seeking to solve problems and accomplish goals that require the exercise of valued personal skills (Lepper, 1985). For an instructional environment, such activities may incorporate goals, uncertain outcomes, performance feedback, and self-esteem (Malone & Lepper, 1987). Curiosity involves activities that provide individuals with some optimal level of surprise, incongruity, complexity, or discrepancy as a function of their initial skills and expectations (Lepper, 1985). Such activities may incorporate sensory effects or cognitively intriguing factors such as paradoxes, incompleteness, and potential simplifications (Malone & Lepper, 1987). Control involves activities that provide individuals with the opportunity to determine their own fate, or at least to maintain the perception that they are doing so (Lepper, 1985). Such activities may incorporate the factors of contingency, choice, and power (Malone & Lepper, 1987). Fantasy involves activities related to the involvement in a world of imagination and identifying with fictional characters. Such activities may incorporate emotional aspects, cognitive aspects, or aspects of endogeny (Malone & Lepper, 1987). Cognitive learning outcomes refer to changes in a student's behavior at the intellectual level, and involve the learning of facts, concepts, rules, and procedures. 7. Scope and Delimitations of the Study As mentioned earlier, the current study investigates intrinsic motivational effects and cognitive learning outcomes of an instructional microcomputer game. The study also examines the relationship between intrinsic motivation and cognitive learning in a THE PROBLEM / 9 microcomputer gaming environment. However, the following delimitations to the study are identified: 1. Only one school participated in the study. The school is located in a suburban community in the Greater Vancouver Regional District in British Columbia. 2. Only one instructional game and one noninstructional game were used. Accordingly, the findings of the study apply only to the two games. 3. Only a few concepts, rules, and procedures related to linear algebraic equations in two variables were tested. 4. Only individual motivations were examined. 8. Outline of the Remainder of the Dissertation In Chapter II, a review of literature related to the study is presented. The review deals with convergent perspectives of research on motivation and learning. Various effects of games are also reviewed. Separate reviews of motivation, learning and games are presented and, where applicable, an examination of the ways the three concepts may be related is presented. The relationship between each of the three concepts and computers is also discussed. In Chapter III, a description of the research methodology and design of the study is presented. The chapter includes a description of the two games, the instruments used and the pilot project. Chapter III also includes a description of the research procedures, the selection of the subjects, the laboratory setting, the methods of data collection and the methods used in analyzing the data. A list of the hypotheses, operationally stated in the null form, is also included. In Chapter IV, the results obtained from analyzing the collected data are presented in concise terms, accompanied by tables and charts where appropriate. T H E P R O B L E M / 10 In Chapter V, implications of the study are presented. Interpretations of the results as they relate to the stated hypotheses are presented first. A discussion of the results as they relate to earlier results and theory follows. Finally, conclusions are drawn and recommendations for designers of instructional microcomputer games and suggestions for further research are presented. II. REVIEW OF LITERATURE The development of learning theories in the last few decades has not been paralleled by the development of a theory of motivation supported by empirical research. However, various pedagogical decisions have been made built on the assumption that motivation affects learning. Consequently, a review must consider the convergent perspectives of research on both motivation and learning. Microcomputer games have special capabilities as devices for studying motivation and learning. Thus, a review of games must also be considered. The present review examines separate literatures on motivation, learning, and games. The way the reviewed concepts may be related, as well as their relationship with computers, will also be examined. The review seeks a cohesive guide for theoretical work on the interaction of motivation and learning. A. MOTIVATION It is common belief that motivation affects performance. In school and in industry, especially, motivation has been considered basic for making various pedegogical decisions (Lepper & Chabay, 1985) and for increasing workers productivity. Thus, rewarding school children or workers is a common practice for motivation. Yet, due to many factors, according to Lepper and Chabay, this concept has received so little research attention in the last two decades that a sound theory of motivation and the way it interacts with other factors has not yet developed. The earlier view of Hull was that motivation can be explained within a framework of reinforcement and that drives are the motivational aspect of psychological needs (Cofer & Appley, 1964). Cognitive theorists, on the other hand, 11 REVIEW OF LITERATURE / 12 believe that, to satisfy their needs, humans set their own goals and make choices about what to do. Their choices, in turn, are based on the goals they set and their assessments of the likelihood that various behavioral alternatives will lead to these goals. The interest in intrinsic motivation accompanied the interest of cognitivists in internal processes, and research on intrinsic motivation has therefore received more attention recently. 1. E x t r i n s i c a n d I n t r i n s i c M o t i v a t i o n Bruner (1962) distinguished between behavior controlled by externally-administered gains and losses and behavior that operates by intrinsic, self-administered rewards and punishment. Extrinsic rewards, according to Bruner, lose their power at a very sharp rate the further they are separated in time from the acts that they are supposed to be controlling. Discovery, on the other hand, "with the understanding and mastery it implies, becomes its own reward, a reward that is intrinsic to the activity of working" (Bruner, 1960a). Curiosity and competence are two examples of intrinsic motives that have potential to initiate learning with discovery in children (Bruner, 1960b). According to Lepper and Malone (1987), however, intrinsic motivation is determined by the individual motivations of challenge, curiosity, control and fantasy, and the interpersonal motivations of cooperation, competition and recognition. The detrimental effects of externally-administered rewards on intrinsic motivation and performance have been studied. DeCharms (1984) predicted a paradoxical effect of an extrinsic reinforcer on an intrinsically motivated action. He stated: "... if an action is freely initiated through choice and then it is discovered that someone else desires the actor to do the action so much that the other will reward him or her for doing it, then the actor may lose the feeling of freedom, ownership and choice ... the actor should continue the action only as long as the extrinsic reinforcer is offered." REVIEW OF LITERATURE / 13 Deci (1975) reviewed a number of studies which confirm that external rewards can have detrimental effects on intrinsic motivation. Levine and Fasnacht (1974) also pointed to an impressive amount of data which indicate that giving a reward can have an effect counter to that desired. The authors argued that the apparent artificiality of administering rewards in the classroom may undermine the inherent interest to learn in children. Earlier experiments on motivation involved infrahumans. Later, however, similar experiments involving humans have been conducted. To demonstrate that giving a reward can have detrimental effects on intrinsic motivation, Lepper, Greene, and Nisbett (1973) conducted an experiment in which nursery school children were given the opportunity to play with felt-tip markers and paper. The children had demonstrated high intrinsic interest in this drawing activity during baseline observations. The subjects were assigned to one of three conditions: (1) expected reward condition, (2) unexpected reward condition, and (3) no reward condition. After the subjects played under the three conditions, they were given the opportunity to play with the markers in a free-play situation. The study revealed that the expected-reward group played with the previously-rewarded activity, in the free-play period, less than the other two groups. Furthermore, the quality of the pictures drawn by the expected-reward group was judged to be poorer than the quality of the pictures drawn by the other two groups. The results, according to the authors, indicate a decrease in intrinsic motivation under the expected-reward condition. REVIEW OF LITERATURE / 14 2. R e w a r d s a n d L e a r n i n g Studies that examined the effects of rewards on learning have also been conducted. McGraw (1978) reviewed research studies that examined the effects of extrinsic rewards on performance. The review revealed that the effects of rewards on performance are detrimental when the task is interesting for the subjects, and when the solution to the task is open-ended enough that the steps leading to a solution are not immediately obvious. The results were consistent among studies concerning discrimination learning, concept attainment, tasks requiring insight and creativity, and incidental learning. Condry and Chambers (1978), on the other hand, reviewed research studies that examined the effects of rewards on the process of learning. Relative to the initial engagement in a task, the review revealed that extrinsic motivation is associated with a preference for relatively simpler activities and with an orientation toward performance rather than progress. The degree of active self-involvement was found to be substantially less under the extrinsic than under the intrinsic contexts. Eysenck (1984) tried to reconcile discrepant views on the effects of incentives on performance. He noted that Skinner's view that incentives or reinforcers enhance performance was applied to simple, repetitive tasks. In contrast, the effects of incentives become detrimental when they involve performance on complex cognitive tasks like heuristics in problem-solving. The author further summed up the various different effects of an incentive on cognitive performance as follows: "It increases attentional selectivity^ it increases performance speed, decreases accuracy of performance, produces some cognitive inflexibility, and increases short-term storage capacity. Some of these effects may be REVIEW OF LITERATURE / 15 interdependent." Eysenck also suggested that there may be additional effects of incentive (and anxiety) on cognitive performance, and that, for a complete psychology of cognition to become a realistic possibility, a complete theoretical understanding of the way motivation, emotion, and cognition interact should develop. Various studies which attempted to examine various effects of extrinsic rewards have been criticized, in general, for their failure to investigate long-term effects or the quality of performance on the tasks involved. Another criticism is that the way the variables involved were measured might have affected the results and conclusions of the studies. For example, the change in time in the free choice period may not in the first place indicate a change in intrinsic motivation. 3. Student Motivation Student motivation is another area that has recently attracted the attention of researchers. Ball (1984) undertook a content analysis of articles published in the Journal of Educational Psychology about student motivation and found out that the topic ceased being a neglected one in recent years. Articles of the 1970's, the author concluded, "were primarily concerned with attribution theory and research and with drawing together the diverse elements of motivation that had characterized earlier research." For Ball, attribution theory is a suitable device by which some of the vast array of motivational constructs can be integrated. The guiding principle of attribution theory (Weiner, 1984) is that individuals search for understanding, seeking to discover why an event has occurred. The Freudian and Hullian theories of motivation, Weiner argues, do not provide the needed conceptual tools to explain classroom motivation. Thus, a theory of motivation must include the full range of cognitive processes and the the full range of emotions, REVIEW OF LITERATURE / 16 and must explain rational and nonrational actions, using the same concepts for both. Stipek (1984) also noticed that many researchers have begun to investigate the relevance of cognitive developmental change for achievement motivation. The author suggested that, on the average, children value academic achievement as they progress through school, but their expectations for success and self-perceptions of competence decline and their affects toward school becomes more negative. Stipek's review of the literature revealed that, until the second or third grade, children's achievement-related cognitions remain generally positive. Thereafter, their achievement cognitions, and probably their accompanying emotions, increasingly reflect their actual relative performance in the classroom. Attitudes toward school become more negative on the average, and children tend to shift their attention to extrinsic rewards for academic outcomes rather than intrinsic satisfaction in achieving greater competence. DeCharms (1984) attempted to study ways of enhancing motivation in educational settings. He developed the origin-pawn terminology to refer, respectively, to the positive experience of internal locus of causality and the negative experience of external locus of causality for behavior. For deCharms, motivation cannot be trained or taught, yet a classroom with the right amount of structure should enhance motivation. An unstructured or a rigidly structured classroom, on the other hand, should inhibit motivation. Teachers can pursue the optimum amount of structure to fit the needs of the class and, when possible, the individual student's needs. The desired structure can be achieved by giving the students conceived choices, and by creating an atmosphere that encourages responsible student-influence attempts and independent activity. REVIEW OF LITERATURE / 17 B. L E A R N I N G A central question in psychological research has been: how do people learn what they learn? Gagne (1984) suggested that studying what people learn is an essential step towards understanding how do they learn. Eysenck (1984) proposed a strategy for studying human cognition. The strategy involves laboratory as well as applied research on human learning. Results of such research should be compared and then integrated. 1. L e a r n i n g T h e o r i e s Currently, several learning theories exist. Behavioristic and neo-behavioristic theories of learning include operant learning, drive reduction, neuro-psychological, social learning and learning systems theories, and deal mainly with the effects of reinforcers on learning. Skinner (1961), for example, advocated the use of teaching machines for reinforcement, and predicted that the future electronic classrooms may incorporate small computers that can have the ability to analyze behavior. Cognitive and humanistic learning theories include discovery learning, meaningful reception learning, and the humanistic self-directed learning theories. Piaget's cognitive developmental theory has also been applied to learning. The Piagetian developmental stages can offer guidance as to the kind of lesson content most suited to children of a certain age (Piaget, 1952). On the other hand, the process of adaptation, coupled with the active interaction of the child with the environment, can be relevant to the teaching methods of teachers and to their organization of learning situations (Richmond, 1970). Gagne's learning systems and Bruner's discovery learning theories will be briefly reviewed, followed by a comparison of various learning theories. REVIEW OF LITERATURE / 18 a. Learning Systems The theory of learning systems is associated with Gagne's concern with handling conditions of learning. Initially, Gagne (1965) suggested a hierarchical classification of eight types of learning. These types include: (a) signal learning, -(b) stimulus-response learning, (c) chaining, (d) verbal association, (e) discrimination learning, (f) concept learning, (g) rule learning, and (h) problem-solving. According to the hierarchical classification, one cannot "master a more complex type of learning until the initial states (prerequisites) have been learned. For learning to be effective, then, task analysis procedures should be implemented. Task analysis starts by specifying elementary objectives, followed by utilizing the hierarchy to analyze the adequate subskills that will lead to more complex objectives. Later, Gagne (1984) introduced five categories of learning outcomes: (a) intellectual skills, (b) verbal information, (c) cognitive strategies, (d) attitudes, and (e) motor skills. Intellectual skills, also known as procedural knowledge, include concepts, rules, and procedures. The possession of an intellectual skill is shown when a person is able to apply a sequence of concepts representing condition and action to a general class of situations. For Gagne (1984), learning "must somehow be devoted to acquiring the sequence of the procedure in such a way that it can be retrieved readily." Simple rules involving only a small number of steps are acquired abruptly, complicated ones need continued practice. REVIEW OF LITERATURE / 19 b. Discovery Learning The discovery learning theory is attributed to Bruner's inductive approach to learning. Bruner (1966) developed the theory of structures from investigations into the processes of concept formation. Cognitive structures are combinations of acquired concepts and thinking abilities. Simple cognitive structures develop into more complex ones through the addition of new concepts. Complex structures are so easy to formulate that they can be transmitted at a lower level of cognition. Bruner (1960b) remarked: "... any subject can be taught effectively in some intellectually honest form to any child at any stage of development." The remark means, according to Bruner (1966), that any body of knowledge can be presented to any particular learner in a form that the learner can understand in a recognizable way. The process of cognitive growth, according to Bruner (1966), includes three stages or modes of thinking: the enactive, iconic, and symbolic modes. In the enactive mode, a child's motor behavior leads to motor representations. What the child does is what he or she thinks. Pure S-R theory may form an adequate account for learning that involves enactive representations, which are perhaps the only representaions involved in infant learning. The iconic mode includes visual imagery and other sensory representations that allow a child to make predictions, to extrapolate and to otherwise fill in missing information. Gestalt theory, based upon the analysis of experience and the manner in which perception and memory are linked, forms an adequate system for analysis of the iconic mode. Bruner (1966) stated: "Affective and motivational factors affect imagery and perceptual organization strikingly, particularly when impoverished stimulus material is used and linguistic categorization rendered ambiguous." REVIEW OF LITERATURE / 20 In the symbolic mode, a child can deal with reality in far more flexible ways than in other modes. Vast amounts of information are stored in and retrieved from our language and symbol systems of number and logic. Use of symbolic techniques account for the rational quality of behavior. In the symbolic mode, Bruner (1966) argues, information processing is performed by the mediation of language between stimulus and response and by freeing behavior from "immediate stimulus control by interposing little invisible s's and little r's between the stimulus and the final response." Thus, Bruner (1966) advocated the spiral curriculum as the most effective means of arranging material for the learner who should be assuming more responsibility for learning as his/her knowledge grows. In the spiral curriculum, a learner is given the opportunity to gain familiarity with the structures and to obtain a better grasp of their complexity which, in turn, will further the acquisition of new concepts. The most effective means to better learning is through discovery learning where the child is not given the subject matter in its final form, but is rather required to organize it himself. In doing so, the child will acquire and have full mastery of the structures. c. Comparison of Learning Theories The different learning theories reflect different explanations of behavior by different psychologists. In contrast to the behaviorists, the neo-behaviorists, for example, incorporate mediational processes into their theories, but only as a hypothetical construct to link observable events. Incorporating mediational processes into their theory enables the neo-behaviorists to account for more complex kinds of learning. REVIEW OF LITERATURE / 21 Cognitive learning theories are differentiated from behavioristic ones by the concern of cognitive psychologists with the ability of humans to transform stimuli, process information, and represent external events internally in the form of cognitive structures. There is disagreement, however, between Bruner and Ausubel on the arrangement of material for instruction and on the organization of material. Bruner - (1966) advocated hierarchical arrangement and allowing the learner to organize the material for himself by discovery techniques, while Ausubel (1979) believes that the material should be presented in its final form. However, there are several broad points of agreement among all learning theories (Dubin & Okun, 1973). First of all, there is an agreement that S-R relationships are complex, and that learning requires the identification of relevant stimuli in a complex field. Secondly, there is an agreement concerning individual differences. All theories acknowledge that individuals differ in their capacity to recognize or discriminate relevant stimuli. Thirdly, there is an agreement that motivation initiates and directs behavior and leads to particular responses which are directed toward achieving a specified goal. Finally, there is an agreement on the need of the learner to recognize whether his/her response is appropriate and to modify the response whenever it is inappropriate. The existence of different learning theories has been looked at by Scandura (1970) as a gradual shift in psychology from the study of simple to complex learning. Scandura argues that some theorists have chosen to elaborate on or to extend the S-R mediational language, while others have preferred more cognitive, or rule-based, formulations. Thus, basic formulations and theories grounded in empirical data appealed to the first group, while there was almost no precise language about behavior for the second group. Information processing theories, for example, have REVIEW OF LITERATURE / 22 been formulated in precise terms, but it is not clear how specific aspects of computer programs relate to human behavior. 2. Computer-Based Learning A broad spectrum of computer applications in education has been identified. According to Hall (1982), Computer-Based Education (CBE) includes Computer-Managed Instruction (CMI), Computer-Based Interactive Instruction (CBII) and Computer-Based Instructional Simulation (CBIS). CMI relies principally on record-keeping and summarizing, diagnosis, prescription and monitoring each learner's progress. CBII, Hall stated, "presents instructional material to the learner, accepts and judges responses from the learner, provides feedback, and alters the flow of subsequent instructional material based upon the learner's responses." CBIS, on the other hand, "causes learners to apply, analyze, integrate, and synthesize their knowledge as they solve 'real life' problems that have been stored in the computer." Based on the above classification, Hall (1982) suggested that the development of Computer-Based Education should proceed from CMI to CBII to CBIS, parallel to the three stages of learning (acquisition, transfer and integration). The locus of control thus shifts from the computer in CMI to the learner in CBIS as the learner advances through the stages of learning. Earlier efforts to implement CBE in the school demonstrated that the computer has the potential and power to improve instruction. The effectiveness of Computer-Based Education has been summarized by Berg (1983) as follows: 1. CBE leads to higher scores on criterion tests. It yields gains of one-half of a standard deviation at the elementary level, one-third at the secondary level, and REVIEW OF LITERATURE / 23 one-fourth at the college level. 2. CBE appears to be more efficient. Learning time is reduced forty to fifty percent when compared with traditional instruction. 3. CBE seems to improve student attitudes toward computers and subject matter. Although these results are widely reported, it is difficult to determine the source of this outcome. 4. CBE may increase retention. Some analyses claim that this outcome is substantiated, others suggest slight positive results. 5. CBE may be feasible for small groups of students as well as individuals. Earlier applications of computers in education utilized teletypes as the interactive device between the computer and the learner. The earlier efforts evolved into more sophisticated Computer-Based Learning (CBL) systems such as PLATO, TICCIT and LOGO. The different systems followed different philosophies and different approaches. The philosophy behind TICCIT, the acronym for Time-shared Interactive Computer-Controlled Information Television, is the Component Display Theory (CDT). CDT, according to Merrill (1980), assumes that the components of cognitive subject matter can be classified into four categories: facts, concepts, procedures, and principles. TICCIT also adopted a learning task called "use-generality." The idea behind use-generality is that, given a previously unencountered instance, the student will be able to apply the generality by classifying a concept, executing a procedure, or explaining a relationship (principle). Facts have no generalities and thus were not included in the TICCIT system. CDT, Merrill adds, prescribes an adequate presentation for the use-generality task in TICCIT. The presentation includes a rule display, a set of example displays and a set of practice displays accompanied by appropriate feedback. REVIEW OF LITERATURE / 24 The features of TICCIT described so far ensure learner control whereby a student becomes system-independent. The components of learner control, according to Merrill (1980), are: content selection, display selection, conscious cognition (learning strategies) and meta-cognition. To achieve the purpose of learner control in TICCIT, a special keyboard is utilized as the principal mode of interaction between the computer and the learner. The keyboard consists of an ordinary typewriter keyboard augmented by edit keys on the left and learner control keys on the right. The edit keys are used for such functions as moving the cursor on the screen, entering subscripts or superscripts and modifying verbal messages. The learner control keys are used by the student to indicate which instructional display he/she would like to study next. It includes a rule button, an example button and a practice button in addtion to nine other buttons that ensure learner control. Laurillard (1978) described efforts to implement CBL projects in the U.K. that are based on learner-controlled simulations. The approach of the U.K. projects is that CBL can be problem-see&mg - not only problem-oriented - in the sense that the types of learning and the subject matter that are assisted by the computer should be discovered. At the exploratory stage, rather than using the more traditional kind of test that prejudges the nature of learning that takes place, the evaluation of learning must itself be exploratory and open-ended. Such evaluation provides a basis for future studies with respect to the content of learning and with respect to the implementation of the CBL system. The Computers in the Undergraduate Science Curriculum (CUSC) simulates physical systems encountered in physics, chemistry and biology. Laurillard (1978) argued that, rather than teaching the students by establishing precise objectives, CUSC attempts to enrich learning by requiring that the students use their previous REVIEW OF LITERATURE / 25 knowledge productively and encouraging them to use their reasoning abilities. Evaluation of CUSC was done in two ways. One way was by observing the external manifistations of the students in what they they do and say as they worked on a package. The other way was by utilizing the students' capability to introspect about their experiences via open-ended interviews in which the students were asked to comment on their reaction to the package and what they felt the package helped them with. The type of learning activity evaluated involved "intuitive understanding" which requires the student to demonstrate his/her understanding by using it and by constructing the student's own procedures to control the simulated physical system. Learning activities were related to the design strategies of the computer packages. It was indicated that interpreting can be related to the use of graphics, experimenting to the interactive mode of computer use, and reasoning to the use of questions. Thus, according to Laurillard (1978), it was possible to "demonstrate that Computer Assisted Learning can provide a medium which will enable students to deepen and enrich their intuitive understanding of a subject." The Programmed Logic for Automatic Teaching Operations (PLATO) system is another CBL system that emphasized learner control. Learner control in PLATO is achieved, according to Bitzer, Lyman, and Easley (1966), through versatile "teaching logics." Through "tutorial logics," the student is lead through a fixed sequence of topics (the system presents facts and examples and asks questions) and is allowed to branch between problems (the student answers the questions and asks for judgement). Through "inquiry logics," on the other hand, the system permits dialogues between the student and the computer by presenting general problems to the student. REVIEW OF LITERATURE / 26 To solve the problems, the student must request and organize appropriate information from the computer. According to the authors, PLATO demonstrated "extremely encouraging" educational results, and teaching with PLATO can be "extremely varied" since laboratory, as well as classroom work, is possible. LOGO also emphasized learner control. According to Papert (1980), instead of making the computer teach the child, in LOGO "the child programs the computer." In doing so, the child acquires a "sense of mastery" over the computer and is placed in a position of feeling some identification with scientists. Papert took from Piaget a model of children as builders of their own intellectual structures. However, Papert extended Piaget's model by creating the LOGO environment in which children can communicate with the computer. By communicating with the computer in LOGO, the children would be able to gain a head-start. The children will be able to concretize formal operations even before they reach the age of formal operations. LOGO utilizes a Turtle which is a mechanical device, or a symbol on the surface of the screen of a computer monitor, that moves by obeying orders given by a programmer. Through the Turtle, according to Papert (1980), the children have the opportunity to imitate mechanical thinking, the opportunity to "master the art of deliberately thinking like a computer." Instead of the "got it" or "got it wrong" model of learning, children learn by searching for and correcting "bugs" in the programs they write for the Turtle. REVIEW OF LITERATURE / 27 C. SIMULATIONS AND GAMES The importance of making learning fun through games has been recognized since the earlier stages of history. Plato stated: "... while enforced bodily labours do no harm to the body, study (learning) forced on the mind will not abide there... train your children in their studies not by compulsion but by games (we must make learning fun), and you will be better able to see the natural abilities of each." (Republic, 536-7) Earlier this century, Dewey (1915) urged that games be introduced into the schools as an integral part of the curriculum. He argued that the games children play are "essential factors in their growth." The use of games in schools was also advocated more recently. Bruner (1965) argued that children have a need to devise "emotionally vivid, special games," and that play is a suitable device for intrinsic learning. Coleman (1968), on the other hand, contended that a game "induces, in a restricted and well-defined context, the same kinds of motivation and behavior that occur in the broader contexts of life where we play for keeps." He further wrote: "There are apparently certain aspects of games that especially facilitate learning, such as their ability to focus attention, their requirement of action rather than merely passive observation, their abstraction of simple elements from the complex confusion of reality, and the intrinsic rewards they hold for mastery. By the combination of these properties that games provide, they show remarkable consequences as devices for learning." Piaget (1962) views children's play as a process of evolution closely related to the process of intellectual development. Practice games start to appear with the first month of life, yet become less and less numerous with age. Practice games may evolve into symbolic games, that start during the second year of life and decline after the age of four years, or games with rules, that start between the ages of four and seven years and belong mainly to the concrete operational stage. REVIEW OF LITERATURE / 28 Practice games are found whenever a new skill is acquired. With practice games, behavior is practiced merely for the pleasure obtained from awareness of new powers, and instances of the transition from sensori-motor practice to practical and verbal intelligence can be detected. In symbolic games, on the other hand, the child is interested in the things symbolized, and the symbol merely serves to evoke them. Such games decline at the time when the child starts to adapt himself more to the natural world and progressively subordinate ego to reality. Games with rules evolve from adult practices or from practice or symbolic games which have become collective but in doing so lost all or part of their imaginative content or symbolism. Play, according to Piaget (1962), is therefore distinguishable by a modification, varying in degree, of the conditions of equilibrium between reality and ego. Thus, as the adapted activity or thought constitute an equilibrium state between the processes of assimilation and accomodation, play begins as soon as there is predominance of assimilation: play is assimilation of reality to the ego. 1. Simulation Games A number of reviews of research on the effectiveness of simulations and games as learning environments have been conducted. Fletcher (1971) found out that not much research on games has been carried out and thus proposed a program of research built on two relevant, sets of hypotheses. One set deals with the learning environment created by games. The other deals with what players learn from a game. Such research, Fletcher suggested, may involve variables from the following six categories: (a) the number of players and player's characteristics, (b) the number of alternative actions and the degree of prominence of alternatives, (c) goals or payoff materials, (d) types of conflict of interest, (e) capabilities of players to achieve goals REVIEW OF LITERATURE / 29 and their values to the goals, and (f) blocks to communications (e.g., time for moves, feedback, information flow, and amount of uncertainty). Greenblat (1975) reviewed research studies that investigated various effects of games on teaching and training and summarized the results as follows: 1. there is an increasing amount of positive data on the effects of teaching with games; 2. where the evidence does not reveal benefits of gaming techniques over other modes of teaching, neither does it show the reverse; that is, those taught with games do not prove to have learned less than those taught in traditional ways; 3. questions of retention and application have continued to receive scant attention; 4. results from studies with particular games cannot be generalized to learning with games in general; 5. the quality of evaluative research seems to be improving as researchers become more sensitive to the methodological conditions that yield valid and reliable results. Nonetheless, problems persist by virtue of the nature of the games and the conditions of their operation. Greenblat (1975) therefore concluded that games form an appropriate environment for fulfilling contemporary needs for learning large quantities of information and developing general comprehension of some domains rather than detailed information about them. Furthermore, games entail the active involvement of learners with the subject matter in autotelic activities that free the learners from dependence on authority and offer them feedback and ways of measuring their progress toward a goal. Games naturally incorporate simulation models that display dynamic processes. REVIEW OF LITERATURE / 30 In a later review, Bredemeier and Greenblat (1981) categorized results of research on simulation gaming under three titles: cognitive subject matter learning, affective subject matter learning, and motivation to learn. 1. Cognitive Subject Matter Learning: One claim of simulation gaming advocates is that the method experientially teaches facts, concepts, and procedures more effectively than conventional techniques. Review of claims and evidence suggests that games may be more effective in teaching principles, procedures, and concepts than in teaching facts. It is speculated that the possible efficacy of games lies mainly in removing learning blocks, facilitating preparation for examinations, and providing linkage for various course concepts. Hard evidence favours simulation gaming over conventional methods only with respect to retention of what is learned. In sum, the available evidence suggests that simulation games are at least as effective as other methods in facilitating subject matter learning and are more effective aids for retention. 2. Affective Subject Matter Learning: Simulation games are widely believed to have great potential in the area of affective learning. It seems plausible that experiencing the worlds of others would be more effective than traditional teaching methods for increasing empathy and might lead to changed perspectives and orientations. The available evidence suggests that, under certain circumstances and for some students, simulation gaming can be more effective than traditional methods of instruction in facilitating positive attitude change toward the subject and its purposes. 3. Motivation to Learn: In their course evaluations, students report that they perceive the experience of playing simulation games as having stimulated their motivation and REVIEW OF LITERATURE / 31 interest. Pierfy (1977) reviewed eight studies that compared student interest under simulation gaming treatments with that in standard control treatments. In seven of these eight studies, significantly greater interest was reported to have been stimulated by the simulation gaming approach. Little is reported about the "whys" of motivation and interest stimulation by simulation games. These whys appear to depend, according to Bredemeier and Greenblat (1981), on a variety of interactive variables about which further research is needed. The authors thus deduced that "we do not yet have (1) a theoretically based taxonomy of games with (2) clear theories about (a) what aspects of them are expected to have (b) what sort of distinct effects (c) on what sort of students (d) for what reasons." 2. Non-Computer Games Bright, Harvey, and Wheeler (1978) found out that there has never been a review of literature related to the cognitive effects of playing instructional mathematics games. Furthermore, the authors were surprised that so little attention has been given to the study of mathematics games. Their own review of the relevant literature revealed that research on mathematics games has been neither thorough nor systematic and was instead very fragmanted and based on weak or unstated foundations. Thus, no definitive conclusions, at that time, could be made about the use of games in teaching mathematics, making game playing in the school a promising avenue of research. Bright, Harvey, and Wheeler (1979, 1980a) then started a systematic effort to investigate the cognitive learning outcomes of instructional mathematics games. In-their research, the authors used board games and card games of the types often REVIEW OF LITERATURE / 32 found in mathematics classrooms and school learning centres. Based on the results of their research, Bright, Harvey, and Wheeler (1980b) concluded that, in general, instructional mathematics games can be used effectively to retrain and maintain skills with basic multiplication facts. In two of their studies, Bright, Harvey, and Wheeler (1979) examined the claim that games can be effectively used to improve computational skills. The studies attempted to answer the following questions: 1. can games be used to retrain skills with basic facts? 2. does pre-testing on the skills alter the post-test performance of students? 3. are retraining effects altered by the use of different drill and practice games? One multiplication and one division board game were used in the two studies, which were conducted during the first ten school days of each of two consecutive school years. A pre-test was administered to half of the subjects in the first study, and to all subjects in the second. A post-test of basic multiplication facts was administered to all subjects in both studies. An additional speed test was administered in the second study, both as part of the pre-test and the post-test. Fourteen intact classes (grades 4-6) participated in the first study, and ten others (grades 5-6) in the second. All subjects had the opportunity to master the basic facts in question prior to conducting the experiments. Wilcoxon matched-pairs T-values were calculated to test the first hypothesis. All T-values were significant, indicating increases in the means of all tests used in both studies. ANOVA techniques were used to test the remaining hypotheses. None of the calculated F statistics were significant at the 0.05 level, indicating no effects of pre-testing or choice of game on performance. REVIEW OF LITERATURE / 33 The authors concluded that the studies were successful in demonstrating that games can be an effective way to retrain skills with multiplication facts. However, a possible Hawthorne effect might have occurred as the studies conducted within the first two weeks of schooling, with the students exposed to a new environment of game playing and new teachers. The stability of the effects of games on the learning of mathematical skills was later examined by Bright, Harvey, and Wheeler (1980a) in a study that addressed the following questions: 1. can games be used effectively to help students maintain skills with multiplication basic facts over a school year? 2. if games are effective, how often must they be used? For the purpose of the study, four intact classes (grades 5-6) were used. The experiment took place between September, 1977, and April, 1978. Following a pre-test of multiplication facts, the subjects played the multiplication or division games in 14 playing sessions. The interval of time between one session and the next was increased from an initial 6 instructional days to a final 20 instructional days. Two days after each session, a speed test, sampled from the pre-test, was administered to all subjects. The whole test was administered, as a post-test, 6 weeks after the last playing session. To test the first hypothesis, t-tests were employed. The results revealed a significant gain in the mean score of the whole sample. The gain was significant for each individual class and also by gender. For the session tests, only two of the four classes demonstrated a significant decline in the mean score, and that occurred after the fourth session of play only. REVIEW OF LITERATURE / 34 The authors thus concluded that the game treatment is an effective way to help students maintain their skills with multiplication facts. Moreover, for the games to be effective, the interval between game-playing sessions can gradually be lengthened and can be long enough so that the time devoted to the maintenance of basic multiplication skills is not excessive. The study was criticized for failing to compare the effects of games with other treatments, for the sampling procedures, and for the fact that the effect of the instructional treatment was confounded by other mathematics instruction which took place during the school year. Spraggins and Rowsey (1986) tried to compare cognitive learning outcomes of simulation games with those of worksheets. The authors designed a research study that addressed the following questions: 1. will there be a difference between the achievement of the students taught by simulation games and those taught by worksheets? 2. will there be any significant interactions between the type of instruction and students' ability or gender on achievement? 3. will there be a difference between retention of the students taught by simulation games and those taught by worksheets? 4. will there be any significant interactions between the type of instruction and students' ability or gender on retention? Eighty-three students, 42 boys and 41 girls, forming four biology classes and taught by the same teacher, participated in the study. Each of the four classes was divided into 6 groups of 4 or 5 students, and each group was randomly assigned to either an experimental or a control condition. The experimental group played REVIEW OF LITERATURE 7 35 simulation games introducing the current lesson (experimental games) and was assigned worksheets related to a topic previously covered. The control group played simulation games pertaining to a past lesson (control games) and was assigned worksheets introducing the current lesson. The content ~ of the experimental and control groups' lessons was the same, with only the method of presentation differing. A test that measures the educational ability of the students had previously been administered and the results were used to place each student in one of four ability classes. The top 25% of the students were placed in the first class, the second highest in the second, and so on. The experiment was conducted at three intervals in the school year during the months of December, February, and April. An achievement test in biology was administered at the end of the experimental period, and a random sample of the test items was used in June as a measure of retention. Multivariate analysis of variance was employed to test for various differential learning effects, " while univariate analysis of variance techniques were performed to assess various effects on the retention test. Analyses of the data indicated the following: 1. There was no significant difference in mean achievement scores between students taught by the simulation game and those taught by the worksheets. 2. There were no significant interactions between the type of instruction and students' ability and gender on achievement. 3. There was no significant difference in mean retention scores between students taught by the simulation game and those taught by the worksheets. 4. There was a significant interaction between the type of instruction and students' ability and gender on retention. Low ability females utilizing simulation games REVIEW OF LITERATURE / 36 scored higher on retention than low ability females utilizing worksheets. Low ability males utilizing worksheets scored higher on retention than low ability males uitlizing simulation games. The authors deduced that the simulation games used in the study were as effective in teaching biological concepts as the more traditional worksheets. It was recommended then that simulation games be used as an alternative instructional strategy to the more traditional strategy of worksheets. 3. Computer Games During the last decade, interest in introducing instructional computer games in schools has been increasing. The military developed similar interest in introducing computer games and simulations for training purposes. Nawrocki and Winner (1983) reported on attempts, by the U.S. Navy and Army, to develop and use computer games and simulations for basic skills instruction, and other efforts that have been directed at developing performance measures which can be separated into identifiable psychomotor skill categories within a particular computer game environment. The authors observed players at large arcade game centres and themselves played such games in an attempt at identifying and developing a category classification of computer games. However, they noticed that color could act as a distractor from other more vital cues in a learning situation, that winning while remaining challenged is what motivates players of computer games, and that these players may also play to beat the program rather than the game itself. Moreover, "all a player learns in a game is how to play the game," and this may account for the difficulty in demonstrating learning transfer from games. REVIEW OF LITERATURE / 37 Finally, the authors suggested a set of generic military tasks (navigation, communication, weapon operation) and a set of task descriptors (perform, identify, analyze) that have potential instructional value and can be integrated and used in a computer game environment. They also suggested a set of skill classes which could be related to the generic tasks. The skills include: perceptual skills, psychomotor skills, memory, rules, and concepts. Manipulating proper combinations of the generic tasks, the task descriptors and the skills, the authors argued, would allow for selecting video games with possibly high learning transfer to specific skills, or from skills to training tasks. Loftus and Loftus (1983) attempted to analyze motivational and cognitive aspects of microcomputer and video games. The authors analyzed those cognitive skills and strategies that enter and accompany the playing of such games, and speculated on the learning outcomes of video game playing. Thus, partial reinforcement schedules are in part what makes video games so compelling and irresistible. Reinforcement can be provided immediately, and this instant reinforcement makes the behavior of playing video games so satisfying and therefore so prevalent. Loftus and Loftus also contrasted earlier CAI projects with current ones. Earlier, the computer was challenging but was not considered to be fun. Earlier projects engaged students at the verbal level (via a teletype), hardly at the visual level (graphics and sound). Finally, computers were large and expensive, making isolated, poor communities out of luck. Exploiting the characteristics of the new microcomputer technology makes it feasible to create effective learning environments. Computer games, specifically, have the power to intrinsically motivate children by incorporating elements of goal structures, challenge, and fantasy. REVIEW OF LITERATURE / 38 Malone (1981a) conducted a survey followed by two experiments in an effort at studying general motivational effects of computer games. The investigator attempted to answer the following research questions: 1. Why are computer games so captivating? and 2. How can the features that make computer games captivating be used to make learning - especially learning with computers - interesting and enjoyable? In the survey, 65 private school students (42 boys and 23 girls) from kindergarten through grade 8 were involved. The children were interviewed and asked to fill out a questionnaire pertaining to their preference for each of 25 selected games. Results from the survey showed that no single game received more than 17% of first place rankings. Moreover, individual differences in game preferences according to age, experience with computers, and gender were detected. Malone also studied features of the games and correlated these features with game preferences of the subjects. The results showed that the most important feature of popular games was whether or not the game had a goal. Other important features included scoring, audio effects, and randomness. Graphic games were liked and word games were significantly disliked. In the second study, Malone examined motivating aspects of a sensori-motor game, called Breakout. The game utilizes three features: (a) a bouncing ball, (b) breaking bricks out of a wall, and (c) scores given according to the number of broken bricks. Various versions of the game were created by excluding one or two features from the original version of the game. Breaking of bricks by the bouncing ball was found to be the most important motivating aspect of the game. Bouncing of the ball and score were approximately equal to each other in importance, but both were much less important than breaking out the bricks. REVIEW OF LITERATURE / 39 In the third study, Malone examined motivational aspects of a cognitive skills game called Darts. In the game three balloons appear at random places on a vertical number line on the screen. The player can guess the position of a balloon by typing in a mixed fraction. At each guess, an arrow shoots across the screen to the specified position and pops a balloon. Music follows if the response is correct. Eighty fifth-grade students were involved in the experiment. Eight different versions of the game were designed to allow for testing the effects of its various aspects. Students were allowed to choose and move between a version of the game and another game called Hangman in two twenty-minute sessions. Students' game preferences, their ratings of how well they liked the two games, and the time they spent on playing Darts in preference to Hangman were measured. Analysis of the collected data revealed that all three measures were significantly correlated with each other. Moreover, both the time spent on Darts and preference for the game revealed significant effects of condition (version of the game played). There was also a highly significant interaction between condition and gender in determining time spent on Darts. The original version was significantly less interesting for girls than the version in which the intrinsic fantasy of arrows and balloons was replaced by an extrinsic version of the same fantasy. Boys liked arrows and balloons when introduced as an extrinsic fantasy, and disliked being told in words that their guess was too high or too low. Girls liked the music. Finally, the version with no performance feedback was not significantly less interesting than the version with simple performance feedback for either boys or girls. As a result of the study, Malone (1981b) suggested ways of constructing effective and motivating computer games for learning purposes, utilizing certain REVIEW OF LITERATURE / 40 aspects of challenge, fantasy and curiosity. Challenge may be incorporated in the game by presenting clear, personally meaningful goals, a variable difficulty level determined automatically or by the student, multiple goal levels, randomness, and hidden information selectively revealed. Fantasy, on the other hand, should be emotionally appealing, intrinsically related to the skill learned in the activity, and should provide a useful metaphor. Sensory curiosity can be incorporated through the use of audio and visual effects as decoration, as a reward, as a representation system, and to enhance fantasy. Cognitive curiosity, on the other hand, can be incorporated by including surprises and constructive feedback in the program. As a follow-up study, Malone and Lepper (1987) conducted a research study that compared the learning outcomes of playing the Darts game with the learning outcomes of a parallel drill game in which various, game-like elements had been removed. Two groups of children were used in studying intrinsic motivational effects of the game. Two other groups were used in studying learning outcomes of playing the game. Results of the study revealed that children chose the activity presented in the form of a game for roughly 50% more time than the activity presented in the form of a drill. Compared to a control group, both learning groups demonstrated significant learning about fractions and number lines. Enhanced motivation, however, did not produce enhanced learning. Bright, Harvey, and Wheeler (1980b) investigated cognitive effects of non-computer mathematics games and concluded that such games can be used effectively to retrain and maintain skills with basic multiplication facts. Bright and Harvey (1984) argued that similar conclusions can be drawn regarding the REVIEW OF LITERATURE / 41 effectiveness of computer games as instructional tools. Instructional computer games, according to the authors, have strong motivational capabilities, have capabilities for providing immediate feedback, and can control the nature of instruction provided in the game to fit each individual player. Computer games, on the other hand, may distract students from the instructional objectives of the game, may inhibit transfer to non-computer settings, and may develop inflated, negative reaction in students to other modes of instruction. Realizing the possible advantages and disadvantages of computer games, Bright and Harvey concluded, teachers may choose for use in their classrooms instructional computer games "that have definite instructional objectives, that accurately teach those instructional objectives, and that employ the unique capabilities of the computer in order to achieve maximum effects." Bright (1985) then set up an experiment to study learning outcomes of instructional microcomputer games. In the experiment, a measurement game called Golf Classic, and a probability game called Jar Game were used. In Golf Classic, players estimate angles and lengths and verify the accuracy of their estimates through the visual display on the screen. In Jar Game, the players generate data and receive extra points when they choose the jar with the better ratio. Seventy-eight preservice elementary school teachers, 90% of whom were female, participated in the study. Each subject played one of the two games for a total of 40 minutes and acted as control for the second group. The subjects were grouped into two groups because they were taught by two different instructors. All subjects were asked to complete two achievement tests before the study and two at the end. Analysis of variance on the post-tests revealed no significant differences on the REVIEW OF LITERATURE / 42 probability unit for either group. On the measurement unit, however, the game group achieved significantly higher than the control group on the angle portion of the test. Bright argued that these negative results may be due to the elementary nature of the games, to the restricted size of the monitor screen, to the distortion of the screen, and to the difference in field vision between the screen and paper-and-pencil tests. However, he concluded, much more research is needed on the appropriate instructional uses of computers so that teachers know how to exploit the best features of such environments. D. SUMMARY Three major problems related to motivation have been identified. The first problem involves the existence of motivation as a measurable variable. Skinnerians believe that "linguistic fiction" may have been created in introducing the term motivation in the language of psychology (Orbach, 1979). Other psychologists accept motivation as a hypothetical construct but have difficulties in formulating a definition of the term. Most psychologists prefer to explain, rather than define, motivation. The second problem deals with building a theory of motivation supported by empirical research. Measurable variables that describe motivation have been devised, and two types of motivation, intrinsic and extrinsic, have been recognized. Studies that attempted to investigate various effects of motivation have been criticized for the lack of studying long-term effects and for the way the variables involved were measured, indicating a need for more research on motivation. The third problem deals with applications of motivation. In education, for example, developmental considerations of motivation, effects of the students' environment on motivation, and enhancing motivation in the classroom have been REVIEW OF LITERATURE / 43 discussed. Yet, the available evidence is not sufficient to make generalizations applicable to a classroom setting, again indicating a need for more research on motivation. Cognitive learning theories differ from behavioristic ones in the concern of cognitive psycholgists with internal processes and the ability of human to process information internally. There is disagreement within various cognitive learning theories. For example Bruner and Ausubel disagree on the arrangement and organization of the instructional material. There is also disagreement also within various behavioristic learning theories. For example, neo-behavioristic psychologists, in contrast to behaviorists, incorporate mediational processes in their theories. However, the various learning theories agree that S-R relationships are complex, that individuals differ in their capacity to recognize stimuli, that motivation initiates and directs behavior, and that the. learner needs to recognize and modify his/her responses if necessary. Since the introduction of computers, a number of studies have been carried out to investigate their effectiveness in educational settings. Hall (1982) pointed to the effectiveness of supplemental Computer-Based Education (CBE) at various levels, to considerable savings in time, and to positive effects on attitudes. Specifically, CBE can be effective in achieving higher scores on criterion tests and in increasing retention, in addition to its feasibility for small groups as well as individuals (Berg, 1983). Results of research on games have been categorized under three titles: cognitive learning, affective learning, and motivation. Bredemeier and Greenblat (1981) concluded, in their review of literature on simulations and games, that games can be at least as effective as other methods in facilitating subject matter learning, and REVIEW OF LITERATURE / 44 more effective as aids to retention. The authors also concluded that games can be effective in facilitating positive attitude change toward the subject, and in creating greater interest and motivation in a gaming enviroment than in a standard one. Very few studies, however, attempted to investigate motivational or learning effects of microcomputer games, and virtually no studies have been carried out that investigated both of these effects , and the relationship between them. It is a popular belief that motivation affects learning. Computer games have potential as a tool for investigating the relationship between motivation and learning, the main concern of the current study. III. R E S E A R C H METHODOLOGY AND DESIGN A. R E S E A R C H M E T H O D O L O G Y The current study utilizes an experimental research design approach. A control group and an experimental one have been used. At the beginning of the experiment, a test of divergent feeling that measures complexity, curiosity, risk-taking and imagination was administered to all subjects. Subjects in the experimental group played an instructional microcomputer game and a non-instructional one for equal periods of time in five sessions. For each game, the subjects answered a game-specific test of motivation at two predetermined stages of playing the game. The two parallel tests were designed to measure intrinsic motivational effects of playing the games. The players also answered questions pertaining to their skill at playing each game. Subjects in the control group played the non-intructional game and used worksheets to practice the mathematical content of the instructional game. The control group also sat for five sessions of play and practice. An achievement test of algebra, pertaining to the mathematical content of the instructional game, was administered to the two groups both as a post-test and a retention test. A description of the games, the instruments and the pilot project will be presented next. 1. T h e Games Two microcomputer games were used in the study. One is non-instructional and called Lode Runner™, and the other is instructional and called Mission: Algebra™. 45 RESEARCH METHODOLOGY AND DESIGN / 46 a. Lode Runner™ Lode Runner™ is an arcade-style strategy game. The elements of the game include a hero, treasury rooms, guards, and chests of gold. The hero is described as a Galactic Commando whose task is to infiltrate each of the 150 different treasury rooms by recovering every chest of gold stolen by leaders of a repressive Bungling Empire, evading deadly guards in the process (Figure 1). r < File Editor Game Options Scores Figure 1. Starting Level of Lode Runner™ To recover a chest of gold, the hero must simply touch the chest, evading nearby guards. To evade a guard, the hero must run away from the guard or trap him in a ditch which the hero can dig. The hero can drill ditches or passageways through fissured brick floors and use them to get down to a lower level or to trap guards. To infiltrate a treasury room, the hero must collect all chests of gold in the room. By collecting all chests of gold, a ladder appears which leads to the next room and the hero must make his way to the ladder and up the ladder to the next room (Figure 2). RESEARCH METHODOLOGY AND DESIGN / 47 r £ File Editor Game Options Scores EQSREBBB.iqOB MEMOES L.EMELBB13 Figure 2. End of Level 3 of Lode Runner™ A guard's mission is to protect the chests of gold from being recovered by the hero. A guard can carry one chest of gold whenever he passes by it and may drop the gold later. However, if a guard falls into one of the ditches drilled by the hero, the hero can run over him, taking away any gold which the guard must have dropped. A trapped guard can jump out of a ditch after a period of time, otherwise he will disappear as the ditch closes around him. If a guard disappears in a ditch, he will be replaced by another one at the top of the screen. The player, however, will lose one life if trapped in one of his own ditches which is only one brick wide. The game starts with five lives for the hero and is played by manipulating the movement of the hero by means of special buttons on the keyboard (the mouse is an alternative option on the Macintosh™ version of the game). The hero can move up, down, right or left on the screen of the monitor. The hero can also dig ditches to the right or to the left of where he stands. The player wins points for each chest of gold collected, for trapping guards and for infiltrating a treasury room. RESEARCH METHODOLOGY AND DESIGN / 48 b. Mission: Algebra"" Mission: Algebra™ is an instructional microcomputer game. The elements of the game include a hero and missions of saving disabled space ships through solving linear algebraic equations and plotting points on a graph. The hero is captain of a space ship who, while cruising in his ship, receives a message that a sister ship has been disabled and therefore gets the mission of finding and rescuing the lost ship. To do so, the hero is forced to retrace the path of the lost ship. He/she must use his/her skills in solving linear algebraic equations which help find the path of the lost ship, and plotting graph points which help locate the lost ship on a space map. Specifically, the player must first find the Course of the lost ship by simplifying a linear equation and bringing it to the y=mx+b form (Figure 3a). Then the player must find the new Destination of the lost ship by finding the value of Y, given a value for X (Figure 3b). Finally, the player must plot on a graph the new destination [point (X,Y)] which has already been found for the lost ship (Figure 3c). The three steps must be repeated until the lost ship is found. However, the player may be given the opportunity to make a guess about the next position of the ship. Making a correct guess depends on the ability of the player at recognizing the travelling pattern of the lost ship. The player has the options of choosing a path to play with from a "menu" of built-in paths, letting the computer choose the path, or even creating a new path to play with. The player also has the option of choosing the difficulty level (easy or hard) of the problems to be solved (Figure 3d). The player wins points for simplifying an equation, solving for Y, plotting a point on the graph, and making a correct guess of the next position of the lost RESEARCH METHODOLOGY AND DESIGN / 49 P r e s s : C*) f o r x C ) f o r * CF) to e n t e r a f r a c t i o n CtO to e rase CH] f o r h e l p CGI to d i s p l a y graph Y+7-7*-lX+18-7 Y--1X+U C O o u r s e ? - » Y ' - l X + U I C D ) e s t i n a t i o n ? The coMPuter p i c k s up the t r a i l : Y+7—1JU18 The t r a i l changes d i r e c t i o n a t X=8. 3a. Course Screen 1 < ) P r e s s : C O f o r x ( / ) f o r + CF) to enter a f r a c t i o n C4-) to erase CH) f o r ImIp CN) to d i s p l a y Horkspace ( / J \ C * , -91 C O o u r s e ? CD)es t ina t ion? Mo data i s coding i n - You Hay t ry to yuess the l o c a t i o n of the next point an the path. fPress RETOHHl Y+7=-lX+m P r e s s : C«) f o r x C/3 f o r r CF) to enter a f r a c t i o n €«•) to erase CH) f o r help CC) to d i s p l a y graph Y+7-7«-lX+l8-7 Y * - l X + H C O o u r s e ? Y * - X + l l CO)est inat ion? Go to C O a p h to p l o t the po in t * 3b. Destination Screen Use the I-M keys to p o i n t to teiat you want. Then p r e s s RETURN. > P lay gaMe Ccotcniter chooses path) P lay gaMe Cyou choose path) Create o r e d i t a path See a demonstrat ion Haue YOU Sound? Yes S k i l l l e v e l ? Easy Disk d r i v e s ? Tho Backgroieid? B lack 3c. Graphs Screen 3d. Main Menu Figure 3. Mission: Algebra™ Screens ship. The player loses points for each error, or for using the Help facility. The Help facility allows the player to skim through pages of algebraic rules, to look at useful hints, or to see the next step(s) in the solution of the problem. The types of errors include: (a) range errors (number larger than 255), (b) non-matching parentheses errors, (c) operator errors (wrong mathematical symbols), (d) no " = " sign error, (e) syntax errors, (f) division by zero errors, and (g) wrong form error (not in the y=mx+b form). RESEARCH METHODOLOGY AND DESIGN / 50 2. Tests of Motivation and Learning Two game-specific tests that measure intrinsic motivational effects of playing the games were designed by the investigator. An achievement test of algebra that measures cognitive learning outcomes of practicing the solution of linear algebraic equations was also designed. The three tests were pilot-tested for reliability. They are described next. a. The Tests of Motivation The factors of challenge, curiosity, control, and fantasy have been utilized to study intrinsic motivational effects of the two games. In the Lode Runner™ game, the elements of chest of gold, guard, and treasury hall may be used as goals that the player would target. The Load Runner™ test was thus designed using combinations of the factors of challenge, curiosity, control, and fantasy, and the game elements of chests of gold, guards, and frames (Appendix A). A similar procedure has been followed for the construction of Mission: Algebra™ test items. The game elements used were missions of saving a lost space ship, solving linear equations, and plotting points on a graph (Appendix B). A parallel was then drawn between missions and frames, between solving equations and guards, and between plotting points and chests of gold. Thus Mission: Algebra™ test items were designed built on this parallelism by replacing a Lode Runner™ game element by the corresponding one in Mission: Algebra™. The test items that have been constructed for the factors of challenge, curiosity, control and fantasy of both Lode Runner™ and Mission: Algebra™ are presented in Figures 4, 5, 6 and 7 respectively. RESEARCH METHODOLOGY AND DESIGN / 51 LODE RUNNER MISSION: ALGEBRA 1. I wanted to finish this frame 2. I wanted to get more chests of gold 3. I wanted to trap more.guards 4. I wanted to get a high score 5. I wanted to guess which guards carried concealed chests of gold 6. I wanted to know how well I was playing 7. I felt better about myself everytime I trapped a guard 8. I felt better about myself everytime I retrieved a chest of gold 9. I felt better about myself by finishing the last frame 1. I wanted to finish this Mission 2. I wanted to plot more points on the graph 3. I wanted to solve more equations 4. I wanted to get a high score 5. I wanted to make guesses about the position of the lost ship 6. I wanted to know how well I was playing 7. I felt better about myself everytime I solved an equation that helps locate the lost ship 8. I felt better about myself everytime I plotted a point to help locate the lost ship 9. I felt better about myself by finishing the last mission of saving a lost ship Figure 4. Challenge Test Items LODE RUNNER MISSION: ALGEBRA 1. I wanted to know what graphic and sound effects occur when a guard is trapped in a ditch I dig 2. I wanted to know what graphic and sound effects occur when I finish a frame 3. I wanted to know what graphic and sound effects occur when I retrieve a chest of gold 4. I wanted to know what graphic and sound effects occur when I make a mistake 5. I wanted to know how this game can be played better 6. I wanted to know what the next frame looks like 7. I wanted to know how the game ends with the last frame 8. I wanted to know where a new guard will appear when a guard disappears in a ditch I dig 1. I wanted to know what graphic and sound effects occur when I solve an equation 2. I wanted to know what graphic and sound effects occur when I finish a mission 3. I wanted to know what graphic and sound effects occur when I plot a point on the graph 4. I wanted to know what graphic and sound effects occur when I make a mistake 5. I wanted to know how this game can be played better 6. I wanted to know what the path of the next mission looks like 7. I wanted to know how the game ends with the last mission 8. I wanted to know what sort of equation comes after the one I have been solving Figure 5. Curiosity Test Items RESEARCH METHODOLOGY AND DESIGN / 52 LODE RUNNER MISSION: ALGEBRA 1. I wondered if the score 1 get really depends on how many chests of gold 1 retrieve 2. 1 wondered if the score 1 get really depends on how many guards 1 trap 3. 1 wanted to play with frames of my own choice 4. 1 wanted to control the movement of guards 5. 1 wanted to have the power to design my own frames 6. 1 wanted to have the power to control the number of guards 7. 1 wanted to have the power to control the number of chests of gold 1. I wondered if the score I get really depends on how many points I plot on the graph 2. I wondered if the score I get really depends on how many equations I solve 3. I wanted to play with missions of my own choice 4. I wanted to control the difficulty level of the equations that have to be solved 5. I wanted to have the power to design my own missions 6. I wanted to have the power to control the number of equations to be solved 7. I wanted to have the power to control the number of points to be plotted on the graph Figure 6. Control Test Items LODE RUNNER MISSION: ALGEBRA 1. I imagined I was the hero everytime I trapped a guard 2. I imagined I was the hero everytime I retrieved a chest of gold 3. I imagined I was the hero when I finished this frame 4. I thought of good people similar to the hero in the game, all doing right things 5. I related guards to actual people working for a wrong cause 6. I compared retrieving a chest of gold with doing good things for society 1. I imagined I was the hero everytime I solved an equation that helps save the lost ship 2. I imagined I was the hero everytime I plotted a point on the graph to help save the lost ship 3. I imagined I was the hero when 1 finished this mission 4. 1 thought of good people similar to the hero in the game, all doing right things 5. 1 related solving equations to solving actual daily-life problems 6. 1 compared plotting points to trace the lost ship with doing good things for society Figure 7. Fantasy Test Items For each item, a student could strongly agee, agree, disagree, or strongly disagree with the statement presented. Scoring was done by assigning a value of 5 for a strongly agree response, 4 for agree, 2 for disagree, 1 for strongly disagree. RESEARCH METHODOLOGY AND DESIGN / 53 Thus a maximum score of 45 could be obtained for challenge, 40 for curiosity, 35 for control, and 30 for fantasy. b. The Test of Algebra The Test of Algebra (Appendices C and D) consists of. two parts: a concepts and rules part (Part I), and a procedures part (Part II). The Test of Algebra measures which types of cognitive learning outcomes (intellectual skills) can be achieved by playing the instructional game as compared to using the worksheets. Algebraic expressions and graph concepts, and rules for order of operations and transposition have been used to construct the test items. Part I is a multiple-choice item test, and Part II is a step-by-step solution test consisting of questions taken at random from the instructional game. Initially, ten items for each concept and each rule, and ten procedure items, were constructed. A reliability study reduced Part I to 32 items that need 13 minutes to complete, and Part II to eight items that need 12 minutes to complete. Scoring for Part I of the test was done by assigning a value of 1 for a correct response to an item and 0 otherwise. A maximum score of 16 could thus be obtained on each of the concepts and rules subtests. Scoring for Part II was done by assigning a value of 1 for a right answer and 0 otherwise. A maximum score of 8 could be obtained on this part of the test. 3. Pilot Project The primary purpose of the pilot project was to study the reliability of the instruments which were devised by the investigator, and to adjust them to produce reliable results in the main study. RESEARCH METHODOLOGY AND DESIGN / 54 a. Procedures The pilot project took place in a secondary school in Richmond, British Columbia. A sample of twelve llth-grade students, 6 males and 6 females, was selected. The subjects were divided into two groups of 6 students each. Each group was asked to play Mission: Algebra™ and Lode Runner™ in two one-hour sessions. For each session, half the time was spent on Mission: Algebra™ and the other half on Lode Runner™. The experimenter observed the players and recorded the time taken by each player to finish each phase of each game. For each game, the appropriate motivation test was administered to each player individually when he/she finished the first phase for the first time, then again at the end of the second session of playing the game. The Test of Algebra was administered after the last playing session. Results of the first phase of the project were analyzed. Part I of the Test of Algebra was then adjusted by rewriting those items which were either too easy or too difficult. Part II and the revised version of Part I were administered five days later. Three days after that, Part I of the test was administered for the last time. b. Results Item analysis techniques were used in estimating the reliability of the two motivation tests and the concepts and rules parts of the achievement test. Test-retest correlation coefficients were also computed and their significance tested for all instruments used in the experiment. A Cronbach's alpha of 0.88 and a test-retest correlation coefficient of 0.77 (n=ll, p = 0.003) were obtained for the total motivation scores (mean=89.75, RESEARCH METHODOLOGY AND DESIGN / 55 S.D. = 30.29) of the subjects on the Mission: Algebra™ test. Hoyt estimates of reliability and test-retest correlation coefficients for the four motivation subscales are reported in Table 1. Table 1. Mission: Algebra™ Estimates of Reliability C h a l l e n g e C u r i o s i t y C o n t r o l F a n t a s y E s t i m a t e o f -a=0.93 a=0.95 a=0.88 a=0.93 R e l i a b i l i t y x=30.08 x=24.25 x=20.50 x=14.92 s=10.33 s=10.06 s=7.66 s=6.71 T e s t - r e t e s t r=0.58 r=0.93 r=0.40 r=0.92 c o r r e l a t i o n n = l l n = l l n = l l n = l l c o e f f i c i e n t p=0.031 p=0.000 p=0.U0 p=0.000 For the total motivation scores on Lode Runner™ (mean= 106.42, S.D. = 20.62), a Cronbach's alpha of 0.88 and a test-retest correlation coefficient of 0.94 (n=ll, p = 0.000) were obtained. Hoyt estimates of reliability and test-retest correlation coefficients for the four motivation subscales are reported in Table 2. Table 2. Lode Runner™ Estimates of Reliability C h a l l e n g e C u r i o s i t y C o n t r o l F a n t a s y E s t i m a t e o f a=0.67 a=0.89 a=0.77 a=0.93 R e l i a b i l i t y x=35.00 x=28.50 x=24.17 x=18.75 s=4.33 s=7.32 s=5.31. s=6.72 T e s t - r e t e s t r=0.75 r=0.98 r=0.74 r=0.80 c o r r e l a t i o n n = l l n = l l n = l l n = l l c o e f f i c i e n t p=0.004 p=0.000 p=0.005 p=0.002 Hoyt estimates of reliability for the original and final versions of each of the concepts and rules parts of the test of algebra, as well as the test-retest correlation coefficients for the final version of the three subtests are reported in Table 3. RESEARCH METHODOLOGY AND DESIGN / 56 Table 3. Test of Algebra Estimates of Reliability C o n c e p t s R u l e s P r o c e d u r e s E s t i m a t e o f a=0.54 a=0.58 R e l i a b i l i t y x=13.33 x=14.59 ( O r i g i n a l ) s=1.92 s=2.27 E s t i m a t e o f a=0.74 a=0.82 R e l i a b i l i t y x=9.83 x=9.75 ( F i n a l ) s=3.50 s=4.16 T e s t - r e t e s t r=0.60 r=0.59 r=0.83 c o r r e l a t i o n n = l l n = l l n = l l c o e f f i c i e n t p=0.027 p=0.029 p=0.001 c. Discussion Internal consistency and stability coefficient methods were used as estimates of reliability. Internal consistency was tested by computing Cronbach's alpha for total scores, and Hoyt estimate of reliability for subscale scores. A minimum of 0.70 is normally accepted as a reliability measure for each of the two statistics. The results indicate that the internal consistency of all scales and subscales, except for the challenge factor of the Lode Runner™ test, is established. Stability of the three instruments, on the other hand, was tested by computing test-retest correlation coefficients and testing for their significance. The results suggest that stability of all scales and subscales is established at 0.05 level, except for the control subscale of the Mission: Algebra™ test. Based on these results and the fact that the sample was small (12 subjects), all three instruments were accepted as suitable for use in the main study. Hoyt estimate of reliability for the challenge factor of the Lode Runner™ test was 0.67, a little lower than the normally accepted value. The test-retest correlation coefficient for the control factor of the Mission: Algebra™ test was 0.40, signficant at 0.110 level. RESEARCH METHODOLOGY AND DESIGN / 57 4. T h e Test o f D i v e r g e n t F e e l i n g The Creativity Assessment Packet (CAP) is a test packet that can be used to screen, identify, and evaluate the most important factors of creativity found in some degree among children (Williams, 1980). CAP consists of a Test of Divergent Thinking, a Test of Divergent Feeling, and a rating instrument for parents and teachers of the same test factors. The packet is suitable for boys and girls ages 8 through 18 (grades 3 to 12). The Test of Divergent Feeling is a 50-item multiple choice exercise asking children how curious, imaginative, complex and risk-taking they are. It yields a total weighted raw score and four subscores of curiosity, imagination, complexity and risk-taking, all affective in nature. A significant test-retest correlation coefficient for a sample of 256 students was obtained by Williams (1980), indicating moderate reliability for the instrument. A validity coefficient of 0.76, statistically significant beyond the 0.05 level of confidence, was also reported by Williams. A validity coefficient is the correlation coefficient obtained by correlating the test scores of a group of students with a related criterion measure administered at the same time or within a short interval of time. A mean score of 62.1 and a standard deviation of 18, derived from the groups upon which the instrument was validated, are accepted as norms for interpreting data obtained by the Test of Divergent Feeling. Based on these norms, divergent feeling scores were divided into three levels. A score. of less than 50 was considered low, a score between 50 and 74 was considered average, and a score greater than 74 was considered high. RESEARCH METHODOLOGY AND DESIGN / 58 B. R E S E A R C H D E S I G N A N D P R O C E D U R E S An experimental research designed was used in the study. Five mathematics classes, selected at random from a single school, participated in the study. All subjects were studying a unit of algebra related to the content of the instructional game. Prior to the experiment, the subjects were told by their instructors that the experiment forms part of the unit of instruction they were studying. No direct instruction related to the content of the instructional game, specifically, transforming a linear equation in two variables into the slope-intercept form, has been previously given to the students. A description of the procedures, selection of the subjects, laboratory setting, and methods of data collection and analyses is presented next. A statement of hypotheses is also presented. 1. P r o c e d u r e s First, the Test of Divergent Feeling (Williams, 1980) was administered to a sample of lOth-grade students. The scores obtained on this test were used (as an independent variable) in studying intrinsic motivational effects of the games and cognitive learning outcomes of the instructional game. Following the administration of the Test of Divergent Feeling, the subjects were randomly assigned by class and gender to either an experimental or a control group (Figure 8). Five mathematics classes participated in the experiment. Each subject in the experimental group was asked to sit for five playing sessions. The first session was used to introduce the subjects to the games and to let them practice playing each game. In the remaining four sessions, the order of playing was reversed each session in the sense that students who played Lode Runner™ in one session had to play Mission: Algebra™ in the next session, and vice-versa. Test of Divergent Feeling i Complexity Curiosity Risk-taking Imagination Control (37 girls, 30 boys) (37 girls, 30 boys) 1 Pract ice J Experimental Pract ice Lode Runner Lode Runner Challenge Curiosity Control Fantasy I.e.sJ..oj,.Motivation LRch1 LRcul LRcol LRfnl 3 ^Algebra Worksheets Lode Runner J Mission: Algebra i : Lode Runner J f \ / ' 5 ^Algebra Worksheetsj ^ Mission: Algebra ( Algebra Post Test l£5L2!J^ 9J^ tion MAchl MAcul MAcol MAfnl ISSL^ M o t i o n LRch2 LRcu2 LRco2 LRfn2 IS.?.!..?/. J.?lv.3i!9S MAch2 MAcu2 MAco2 MAfn2 ™L°!..£l9.™ Concepts Rules Procedures weeks Algebra Retention Test J l ! £ ° ! . ^ a Concepts Rules Procedures Figure 8. Research Design and Procedures 50 H CO W > S3 O X H X O a o t-* o o > a a M CO i—i O CO RESEARCH METHODOLOGY AND DESIGN / 60 Phases were assigned to Lode Runner™, each phase consisting of one frame. As mentioned earlier, a frame is a microcomputer screen graphic design representing a treasure hall in which a hero moves to collect all available chests of gold before moving to the next frame. Players were asked to answer the Lode Runner™ Test of Motivation (Appendix A) after they had successfully finished the first frame for the first time and at the end of the last session of playing the game. As with Lode Runner™, different phases were assigned to Mission: Algebra™, each phase consisting of a mission of saving a sister ship. The Mission: Algebra™ Test of Motivation (Appendix B) was also administered twice: after saving one ship, and at the end of the last playing session. The five playing sessions yielded four total scores (16 subscores) for each subject. The scores were used to study intra-game and inter-game differences in motivational effects of the two games. The scores were also used to study the relationship between intrinsic motivational effects and cognitive learning outcomes of playing the instructional game. The control group played Lode Runner™ as a placebo to control for a possible Hawthorne effect, and used worksheets to practice the same subject matter incorporated in Mission: Algebra™. In the first session, the control subjects were introduced to the game and worksheets and then practiced both. In the remaining four sessions, they played Lode Runner™ and practiced on the worksheets alternatively in the sense that subjects who played the game in one session had to practice on the worksheets in the next session, and vice-versa. Each equation in the worksheets was followed by a work space and spaces for the answers. A 2-dimensional grid at the bottom of each sheet allowed the students RESEARCH METHODOLOGY AND DESIGN / 61 to plot the points obtained by solving the equations (see Appendix E for an example). Students were given the worksheets at the start of a 50 minute period with instructions to solve all equations on the page. The supervisor went from student to student, marking all questions completed to that point as either right or wrong. Students corrected the wrong responses and then moved to untried questions, or to graphing the responses. The supervisor examined each student's paper about every five minutes, or about 10 times during the session. Students who completed the worksheets before the end of the period were instructed to note the elapsed time on their paper. The experiment ended with the Test of Algebra (Appendix C), administered as a post-test to both groups that participated in the experiment. The test was also administered as a retention test in the fifth week after the post-test. The Test of Algebra was designed to measure cognitive learning outcomes of playing the instructional game as compared to practicing the worksheets. 2. S e l e c t i o n of Subjects In the current study, 134 lOth-grade students were selected from a secondary school in North Vancouver, British Colmbia. Five mathematics classes were randomly selected from the school, and the subjects participated in the experiment as whole classes. Each student was randomly assigned to either the experimental or control group. There were 74 girls, half of them in the experimental group, and 60 boys, also half of them were in the experimental group. One girl from the experimental group was dropped at random from the study because the number of microcomputers was not enough to accomodate all subjects in her group. RESEARCH METHODOLOGY AND DESIGN / 62 3. Laboratory Setting and Procedures The experiment took place in two rooms, one containing 19 Macintosh™ microcomputers and the other 18 Apple® II ones. The control group played Lode Runner™ in the Macintosh™ laboratory and practiced the worksheets in the classrooms. The experimental group spent all five sessions in the Apple® II laboratory. The Apple® II microcomputers were arranged in the room in such a way that the experimenter could easily observe all subjects. Arrangement of the microcomputers, coupled with built-in graphic and sound effects that mark the end of each playing phase of a game, allowed the experimenter to determine when a game-specific test of motivation should be administered to each individual subject for the first time. 4. Hypotheses to be Tested The design of the study allowed for testing the following hypotheses: 1. For each game, there will be no significant difference between the means of the total motivation scores of the experimental group in the first testing session and their respective means in the second testing session. H ° i : "gl " wg2 = °' S = 1 ' 2 g=l for Lode Runner™, g=2 for Mission: Algebra™. 2. There will be no significant difference between the means of the total motivation scores of the experimental group on the Lode Runner™ test, at either the first or second administration, and their respective means on the Mission: Algebra™ test. Ho2: M 1 S " M 2 S = 0 , s=l,2 s=l for first testing session, s = 2 for second testing session. RESEARCH METHODOLOGY AND DESIGN / 63 For each game, and for each motivation subscale, there will be no significant difference between the means of the scores of the experimental group in the first testing session and their respective means in the second testing session. H ° 3 : V l " V 2 = °' 8 = 1 , 2 i = 1 > 2 ' 3 ' 4 g=l for Lode Runner™, g=2 for Mission: Algebra™; i= l for challenge, i = 2 for curiosity, i = 3 for control, i = 4 for fantasy. For each motivation subscale, there will be no significant difference between the means of the scores of the experimental group on Lode Runner™ test, at either the first or second administration, and their respective means on the Mission: Algebra™ test. Ho 4: tflig - u 2 i s = 0, i= 1,2,3,4 s=l,2 s = 1 for first testing session, s = 2 for second testing session. i= l for challenge, i = 2 for curiosity, i = 3 for control, i = 4 for fantasy. For each game, there will be no significant difference in mean total motivation scores of the experimental group between boys and girls, or among the three divergent feeling groups over the two playing sessions. Ho 5: M G L = M G 2 = Ui g=l,2 (gender) M G L = /ug2 = M G 3 = M, g=l,2 (level of divergent feeling) g=l for Lode Runner™, g=2 for Mission: Algebra™. For each game, there will be no significant interaction on total motivation scores between session and gender, between session and divergent feeling, or between gender and divergent feeling. H°«: a 0 g i j = 0' g =l,2 i= 1,2 j = l,2 (session by gender) a7"gik = 0, g=l,2 i=l,2 k=l,2,3 (session by level of divergent feeling) 0 7 g ^ j = 0, g=l,2 j=l,2 k= 1,2,3 (gender by level of divergent feeling) g=l for Lode Runner™, g=2 for Mission: Algebra™. For each game, there will be no significant 3-way interaction on total RESEARCH METHODOLOGY AND DESIGN / 64 motivation scores between session, gender and divergent feeling. Ho 7: a 0 7 g i j k = 0, g=l,2 i=l,2 j=l,2 k=l,2,3 g=l for Lode Runner™, g = 2 for Mission: Algebra™. There will be no significant difference in mean post-test scores between the experimental group and the control group, between boys and girls, or among the three divergent feeling groups. Ho 8: = A»2 = M (treatment) ix ^  = M 2 = M (gender) J U ^ = #2 = 1*2 = w ^ e v e^ °f divergent feeling) There will be no significant interaction on the post-test scores between treatment and gender, between treatment and divergent feeling, or between gender and divergent feeling. Ho ,: a|3^j = 0, i=l,2 j = l,2 (treatment by gender) = Or i = 1,2 k= 1,2,3 (treatment by level of divergent feeling) 07^j = 0, j = 1,2 k= 1,2,3 (gender by level of divergent feeling) 10. There will be no significant 3-way interaction on the post-test score between treatment, gender and divergent feeling. H o 1 0 : a 0 7 i j k = 0, i=l,2 j = l,2 k=l,2,3 11. There will be no significant correlation between any one of the concepts, rules, or procedures post-test subscores and any one of the challenge, curiosity, control, or fantasy subscores on the second testing session of Mission: Algebra™. H o x l : j ^ j = 0, i=l,2,3 j = l,2,3,4 i = l for concepts, i = 2 for rules, i = 3 for procedures. j = l for challenge, j = 2 for curiosity, j = 3 for control, j = 4 for fantasy. 12. There will be no significant difference in mean retention scores between the experimental group and the control group, between boys and girls, or among the three divergent feeling groups. RESEARCH METHODOLOGY AND DESIGN / 65 Ho 1 2: = 1*2 = M (treatment) M 1 " = M 2 = M (gender) = = M-j = M (level of divergent feeling) 13. There will be no significant interaction on the retention scores between treatment and gender, between treatment and divergent feeling, or between gender and divergent feeling. Ho 1 3: <*0^ j = 0, i= 1,2 j = 1,2 (treatment by gender) cry^ = 0/ i =l>2 k= 1,2,3 (treatment by level of divergent feeling) 07j_j = 0, j = l,2 k= 1,2,3 (gender by level of divergent feeling) 14. There will be no significant 3-way interaction on the retention scores between treatment, gender and divergent feeling. Ho 1 4: a07 i j k = 0, i=l,2 j= 1,2 k= 1,2,3 5. Methods of Data Collection To test the hypotheses stated above, appropriate data were collected and analyzed. The achievement and affective tests used in the experiment are described above. The Test of Divergent Feeling yields a total weighted raw score and four subscores of curiosity, imagination, complexity and risk-taking. The two tests of motivation are attached as Appendices A and B. Each one of them was administered two times during the experiment, yielding two total scores and eight subscores each. The players were observed to determine when each one of them finished, for the first time, the first phase of the game he/she was playing. At that time, a motivation test was handed to the player for the first time. The Test of Algebra was administered as a post-test and, in the fifth week after the post-test, as a retention test. The test yields a total score and subscores for concepts, rules, and procedures. RESEARCH METHODOLOGY AND DESIGN / 66 6. Analyses of Data Intrinsic motivational effects and cognitive learning outcomes of playing the two games were examined by analyzing the data collected, using t-test procedures, Pearson correlation coefficients, and univariate and multivariate analysis of variance techniques. The t-test procedures were used to examine intra-game and intergame differences in total and subscale scores of the experimental group on the motivation variables. Repeated-measures design analysis of variance was performed to examine differences in total motivation scores of the experimental group between boys and girls and among the three divergent feeling groups over the two testing sessions. Two-way and three-way interactions on total motivation scores of the factors of gender, divergent feeling and session were also examined. Multivariate analysis of variance was performed to examine differences in post-test and retention test scores between the experimental and control groups, between boys and girls, and among the three divergent feeling groups. Two-way and three-way interactions on post-test and retention test scores of the factors of treatment, gender and divergent feeling were also examined. Univariate analysis of variance was performed whenever a multivariate F-value had been detected. Pearson correlation coefficients were computed and tested for significance in order to examine the relationship between each one of the three parts of the post-test and the four motivation subscales of the Mission: Algebra™ Test of Motivation, second session. I V . R E S U L T S The data collected in the study were analyzed using t-test procedures, Pearson correlation coefficients, and univariate and multivariate analyses of variance. The computations were carried out using the SPSS:X™ Information Analysis System, Release 2.2. For hypotheses testing, a significance level of 0.05 was set as as acceptable risk of Type I error. For convenience, an asterisk (*) is inserted wherever a significant statistic with p<0.05 is obtained, two asterisks (**) for p<0.01, and three asterisks (***) for p<0.001. Where applicable, cases with missing values were excluded from the respective analyses. A. INTRINSIC MOTIVATIONAL EFFECTS The means and standard deviations of the total and subscale scores of the experimental group on the motivation variables are reported in Table 4. Table 4. Motivation Scale Means and Standard Deviations C h a l l e n g e C u r i o s i t y C o n t r o l F a n t a s y T o t a l Lode Runner™ x=32.71 x=26.14 x=21.45 x=14.74 x=95. 03 F i r s t S e s s i o n s=5.48 s=6.19 s=5.17 s=5.80 S=17. 85 Lode Runner™ x=30.59 x=22.42 x=20.83 x=12.95 x=87. 80 Second S e s s i o n s=6.98 S=7.48 s=6.15 s=5.48 S=21. 90 M i s s i o n : Algebra™ x=30.67 x=23.94 x=21.73 x=12.05 x=88. 38 F i r s t S e s s i o n s=6.37 S=6.37 S=5.11 s=5.12 S=18. 12 M i s s i o n : Algebra™ x=30.67 x=23.15 x=23.59 x=12.64 x=90. 05 Second S e s s i o n s=6.02 S=7.39 S=5.65 s=6.14 s=20. 23 Dependent samples t-tests, based on paired observations, were used in analyzing the data obtained in the study. For each game, testing for intra-game differences in mean motivation scores was based on the scores obtained on the same test of motivation (administered twice). Testing for inter-game differences in mean motivation 67 Results / 68 scores, on the other hand, was based on the scores obtained on the two parallel forms of the test of motivation. The degrees of freedom for each t-test were therefore reduced to accommodate only the cases where data were available on both observations. Thus the mean differences in the motivation scores reported in Tables 5, 6 and 7 do not coincide with the corresponding differences in the means of the scores reported in Table 4. 1. T o t a l M o t i v a t i o n Scores The t-values for intra-game and inter-game mean differences in total motivation scores and their significance levels are reported in Table 5. The results are illustrated in Figure 9. Table 5. Comparisons of Total Motivations Mean d f t - v a l u e s i g o f t LR X - LR 2 6.72 53 2.64 0.011* MA X - MA 2 -1.98 58 -1.05 0.298 LR X - MAj 6.39 56 3.38 0.001*** LR 2 - MA 2 -1.30 49 -0.87 0.390 LR refers to Lode Runner™, MA to Mission: Algebra' 0 30 60 90 120 Motivation Score Figure 9. Mean Total Motivation Scores — i 1 5 0 Results / 69 The data presented in Table 5 indicate the following: 1. There was a significant difference between the two mean total scores of the experimental group on Lode Runner™. The mean score of the group was higher when the subjects successfully finished the first phase of the game for the first time than at the end of the last session of playing the game. 2. There was no significant difference between the means of the two total scores of the experimental group on Mission: Algebra™. 3. There was a significant difference between the means of the two total scores of the experimental group on the two game tests, first administration. The mean score of the group, when the subjects successfully finished the first phase of each game for the first time, was higher on Lode Runner™ than on Mission: Algebra™. 4. There was no significant difference between the means of the two total scores of the experimental group on the two game tests, second administration. 2. Intra-game Subscale Motivation Scores The t-values for the intra-game mean differences in subscale motivation scores and their significance are reported in Table 6. The data presented in Table 6 indicate the following: 1. There was a significant difference between the means of the two challenge scores of the experimental group on Lode Runner™. The mean challenge score of the group was higher when the subjects successfully finished the first phase of the game for the first time than at the end of the last playing session. 2. There was a significant difference between the means of the two curiosity scores of the experimental group on Lode Runner™. The mean curiosity score of the group was higher when the subjects successfully finished the first phase of Results / '70 Table 6. Intra-game Comparison of Motivations Mean df t-value sig of t LRchi • - LRch2 2.15 53 2.57 0.013* LRcu,. -- LRcu2 2.43 53 2.76 0.008** LRcc-i • - LRco2 0.81 53 1.04 0.301 LRfnx • - LRfn 2 1.33 53 1.86 0.068 MAchi • - MAch2 -0.17 58 -0.26 0.798 MAcux • - MAcu2 0.59 58 0.96 0.339 MAcOi • - MAco2 -1.80 58 -2.43 0.018* MAfnx • - MAfn2 -0.61 58 -1.03 0.309 LR refers to Lode Runner™, MA to Mission: Algebra™, ch to challenge, cu to curiosity, co to control and fn to fantasy. the game for the first time than at the end of the last playing session. 3. There was no significant difference between the means of the two control scores of the experimental group on Lode Runner™. 4. There was no significant difference between the means of the two fantasy scores of the experimental group on Lode Runner™. 5. There was no significant difference between the means of the two challenge scores of the experimental group on Mission: Algebra™. 6. There was no significant difference between the means of the two curiosity scores of the experimental group on Mission: Algebra™. 7. There was a significant difference between the means of the two control scores of the experimental group on Mission: Algebra™. The mean control score of the group was higher at the end of the last playing session of the game than when the subjects successfully finished the first phase of the game for the first time. 8. There was no significant difference between the means of the two fantasy scores of the experimental group on Mission: Algebra™. Results / 71 3. Inter-game Subscale Motivation Scores The t-values for the inter-game mean differences in subscale motivation scores and their significance are reported in Table 7. Table 7. Inter-game Comparison of Motivations Mean d f t - v a l u e s i g o f t LR c h i • - MAchj 1.95 56 2.38 0.021* LRcUj • - MAcUi 2.07 56 3.40 0.001*** LRcOj • - MAcOj. -0.21 56 -0.31 0.761 L R f n i • - MAfn x 2.58 56. 4.25 0.000*** L R c h 2 • - MAch 2 0.34 49 0.59 0.556 L R c u 2 • - MAcu 2 0.36 49 0.59 0.556 L R c o 2 • - MAC0 2 -2.88 49 -3.86 0.000*** L R f n 2 • - MAfn 2 0.88 49 1.63 0.110 LR refers to Lode Runner™, MA to Mission: Algebra™, c h to challenge, cu to curiosity, co to control and fn to fantasy. The data presented in Table 7 indicate the following: 1. There was a significant difference between the means of the two challenge scores of the experimental group on the two game tests, first administration. The mean challenge score of the group, when the subjects successfully finished the first phase of each game for the first time, was higher on Lode Runner™ than on Mission: Algebra™. 2. There was a significant difference between the means of the two curiosity scores of the experimental group on the two game tests, first administration. The mean curiosity score of the group, when the subjects successfully finished the first phase of each game for the first time, was higher on Lode Runner™ than on Mission: Algebra™. 3. There was no significant difference between the means of the two control scores Results / 72 of the experimental group on the two game tests, first administration. 4. There was a significant difference between the means of the two fantasy scores of the experimental group on the two game tests, first administration. The mean fantasy score of the group, when the subjects successfully finished the first phase of each game for the first time, was higher on Lode Runner™ than on Mission: Algebra™. 5. There was no significant difference between the means of the two challenge scores of the experimental group on the two game tests, second administration. 6. There was no significant difference between the means of the two curiosity scores of the experimental group on the two game tests second administration. 7. There was a significant difference between the means of the two control scores of the experimental group on the two game tests, second administration. The mean control score of the group, at the end of the last playing session of each game, was higher on Mission: Algebra™than on Lode Runner™. 8. There was no significant difference between the means of the two fantasy scores of the experimental group on the two game tests, second administration. 4. M o t i v a t i o n , Gender, a n d D i v e r g e n t F e e l i n g Repeated-measures analyses of variance were performed to examine differences in motivation between boys and girls and among students with different levels of divergent feeling. In the current study, a mean of 66.69 and a standard deviation of 10.93 were obtained on the Test of Divergent Feeling. This mean was significantly higher (z=2.91) than the mean of 62.1 (standard deviation = 18), obtained by Williams (1980). Results of the analyses for Lode Runner™ and Mission: Algebra™ are reported in Table 8 and Table 9 respectively. Results / 73 Table 8. Lode Runner™ Motivations S o u r c e ms e r r o r ms df F - v a l u e s i g o f F GNDR 9.84 640.52 (1,47) 0.02 0.902 LDF 258.58 640.52 (2,47) 0.40 0.670 LDF x GNDR 894.36 640.52 (2,47) 1.40 0.258 SSN x GNDR 51.02 174.40 (1,47) 0.29 0.591 LDF x SSN 229.79 174.40 (2,47) 1.32 0.277 LDF x SSN x GNDR 15.13 174.40 (2,47) 0.09 0.917 LDF refers to level of divergent feeling, SSN to session, GNDR to gender. Table 9. Mission: Algebra™ Motivations S o u r c e ms e r r o r ms d f F - v a l u e s i g o f F GNDR 147.46 684.87 (1,52) 0.22 0.645 LDF 269.62 684.87 (2,52) 0.39 0.677 LDF x GNDR 134.92 684.87 (2,52) 0.20 0.822 SSN x GNDR 43.68 112.20 (1,52) 0.39 0.535 LDF x SSN 14.05 112.20 (2,52) 0.13 0.883 LDF x SSN x GNDR 86.05 112.20 (2,52) 0.77 0.470 LDF refers to level of divergent feeling, SSN to session, GNDR to gender. The data presented in Tables 8 and 9 indicate the following: 1. There was no significant difference between the boys' and girls' mean total motivation scores over the two playing sessions of the two games. 2. There were no significant differences among the mean total motivation scores of the divergent feeling groups over the two playing sessions of the two games. 3. There was no significant interaction between the gender of students and their level of divergent feeling on total motivation scores over the two playing sessions of the two games. 4. There was no significant interaction between session and gender or level of • divergent feeling of the students on total motivation scores of the two games. 5. There was no 3-way interaction of session, gender, and level of divergent feeling of the students on total motivation scores of the two games. Results / 74 B. C O G N I T I V E L E A R N I N G O U T C O M E S Multivariate analysis of variance was performed to examine the effects of the treatment variable, gender and level of divergent feeling, and their interactions, on cognitive learning outcomes of the experiment. Univariate analysis of variance was performed whenever a significant multivariate F-value was obtained. 1. Post-test Scores Descriptive statistics for the post-test are reported in Table 10. Results from the multivariate and univariate analyses of variance are reported in Table 11. Table 10. Post-test Means (Standard Deviations) Group C o n c e p t s R u l e s P r o c e d u r e s Treatment C o n t r o l 8.68(2.71) 8.13(3.40) 8.63(2 8.52(2 .73) .87) 4.97(2 5.12(2 .34) .46) M a l e Female 8.95(3.04) 7.92(3.05) 9.39(2.74) 7.84(2.64) 5.20(2 4.90(2 .14) .61) Low LDF Ave LDF H i LDF 8.75(3.39) 8.26(2.87) 8.83(3.29) 9.33(2 8.59(2 8.47(3 .77) .48) .27) 5.75(2 5.16(2 4.72(2 .01) .30) .60) T o t a l 8.49(3.05) 8.63(2 .76) 5.09(2 .37) LDF refers to level of divergent feeling. The data presented in Tables 10 and 11 indicate the following: 1. There was no significant difference in mean post-test scores between the experimental and control groups. 2. There was a significant difference between the boys' and girls' mean post-test scores. Boys achieved significantly higher scores than girls on the concepts and rules parts of the test. However, there was no significant difference between the mean scores of the two groups on the procedures part of the test. 3. There were no significant differences in mean post-test scores among the three divergent feeling groups. Results / 75 Table 11. Summary Table for Post-test E f f e c t ms e r r o r ms d f F - v a l u e s i ,g o f F TREATMENT ( 3 , 102) 0.33 0. 800 GNDR ( 3 , 102) 6.58 0. 000*** Concepts 115. 49 8 .53 (If 104) 13.53 0. 000*** R u l e s 112. 91 7 .20 ( 1 , 104) 15.68 0. 000*** P r o c e d u r e s 3. 29 5 .83 ( 1 , 104) 0.58 0. 448 LDF (6, 206) 0.92 0. 483 TREATMENT X GNDR (3 , 102) 0.60 0. 616 TREATMENT x LDF (6, 206) 0.97 0. 448 GNDR X LDF (6, 206) 2.30 0. 036* Concepts 43. 93 8 .53 ( 2 , 104) 5.15 0. 007** R u l e s 17. 83 7 .20 ( 2 , 104) 2.47 0. 089 P r o c e d u r e s 2. 61 5 .83 ( 2 , 104) 0.45 0. 640 TREATMENT x GNDR X LDF (6 , 206) 1.49 0. 183 LDF refers to level of divergent feeling, GNDR to gender. 4. There was no significant interaction of the treatment factor with the gender factor on the post-test scores. 5. There was no significant interaction of the treatment factor with the divergent feeling factor on the post-test scores. 6. There was a significant interaction of the factors of gender and divergent feeling on the post-test scores. The interaction was significant for the concepts subtest only. For subjects with low divergent feeling scores, boys had a higher mean on the concepts subtest than girls. This was also true for subjects with high divergent feeling scores. For subjects with average divergent scores, however, boys had only a slightly higher mean on the concepts subtest than girls. Descriptive statistics for the concepts subtest are reported in Table 12, and a graphical representation of the results is shown in Figure 10. 7. There was no significant 3-way interaction of treatment factor with the factors of gender and divergent feeling on post-test scores. Results / 76 Table 12. Interaction of Gender and LDF M a l e Mean S.D. Female Mean S .D. Low LDF Ave LDF H i LDF 10. 8. 10. 86 29 07 2.27 Low LDF 2.63 Ave LDF 3.32 H i LDF 5.80 8.24 8.05 2 3 3 .28 .13 .09 LDF refers to level of divergent feeling. C o 1 6-1 4-1 2-1 o-8 • 6 " n c e P — Male "•• Female t s 4 • 2 • n . Low Average High Divergent Feeling Figure 10. Interaction of Gender and LDF on the Post-test 2. Retention Scores Descriptive statistics for the retention test are presented in Table 13. Results from the multivariate and univariate analyses of variance are reported in Table 14. Table 13. Retention Test Means (Standard Deviations) Group C o n c e p t s R u l e s P r o c e d u r e s Treatment C o n t r o l 9.64(2.95) 9.33(3.21) 10.04(2.87) 9.78(2.83) 4.39(2.61) 4.89(2.42) Male Female 9.93(2.69) 9.16(3.31) 10.46(2.99) 9.51(2.67) 4.38(2.48) 4.82(2.54) Low LDF Ave LDF H i LDF 9.09(3.11) 9.60(3.11) 9.62(3.05) 11.55(2.84) 9.90(2.53) 9.56(3.25) 4.36(2.87) 4.82(2.27) 4.50(2.80) T o t a l 9.50(3.07) 9.92(2.84) 4.63(2.52) LDF refers to level of divergent feeling. The data presented in Tables 13 and 14 indicate the following: 1. There was no significant difference in mean retention scores between the Results / 77 Table 14. Summary Table for Retention Test E f f e c t ms e r r o r ms df F - v a l u e s i g o f F TREATMENT (3,91) 0.11 0.953 GNDR Concepts R u l e s P r o c e d u r e s 60.99 64.80 9.82 9.24 7.53 5.61 (3,91) (1,93) (1,93) (1,93) 3.12 6.60 8.61 1.75 0.030* 0.012* 0.004** 0.189 LDF (6,184) 1.40 0.216 TREATMENT X GNDR (3,91) 0.66 0.579 TREATMENT X LDF (6,184) 1.34 0.242 GNDR X LDF Concepts R u l e s P r o c e d u r e s 29.00 29.37 39.84 9.24 7.53 5.61 (6,184) (2,93) (2,93) (2,93) 3.38 3.14 3.90 7.10 0.003** 0.048* 0.024* 0.001*** TREATMENT X GNDR x LDF (6,184) 0.71 0.641 LDF refers to level of divergent feeling , GNDR to gender. experimental and control groups. 2. There was a significant difference between the boys' and girls' mean retention scores. Boys achieved significantly higher scores than girls on the concepts and rules parts of the test. However, there was no significant difference between the mean scores of the two groups on the procedures part of the test. 3. There were no significant differences in mean retention scores among the three divergent feeling groups. 4. There was no significant interaction of the treatment factor with the gender factor on the retention scores. 5. There was no significant interaction of the treatment factor with the divergent feeling factor on the retention scores. 6. There was a significant interaction of the factors of gender and divergent feeling on the retention scores. The interaction was significant for the all three subtests. For subjects with low divergent feeling scores, boys had a higher mean on the concepts and rules subtests than girls. This was also true for Results / 78 subjects with high divergent feeling scores. For subjects with average divergent feeling scores, however, girls had a slightly higher mean on the concepts subtest than boys. For subjects with low divergent feeling scores, boys had a higher mean than girls on the procedures subtest. For subjects with high divergent feeling scores, however, girls had a higher mean than boys. Girls with average divergent feeling scores also had a slightly higher mean than boys in the same category. Descriptive statistics are reported in Table 15, and a graphical representation of the results is shown in Figure 11. 7. There was no significant 3-way interaction of treatment factor with the factors of gender and divergent feeling on retention scores. Table 15. Interaction of Gender and LDF Male Mean S.D. Female Mean S.D. Concepts Low LDF Ave LDF H i LDF 10.57 9.37 10.83 2.37 2.68 2.76 Low LDF Ave LDF H i LDF 6.50 9.79 8.95 2.65 3.46 3.05 R u l e s Low LDF Ave LDF H i LDF 13.29 9.74 10.42 2.99 2.51 3.73 Low LDF Ave LDF H i LDF 8.50 10.03 9.09 1.29 2.58 2.94 P r o c e d u r e s Low LDF Ave LDF H i LDF 6.14 4.48 3.25 0.69 2.19 3.25 Low LDF Ave LDF H i LDF 1.25 5.09 5.18 2.50 2.33 2.32 LDF refers to level of divergent feeling. C. M O T I V A T I O N A N D L E A R N I N G Pearson correlation coefficients were computed and tested for significance in order to examine the relationship between each of the three post-test scores of the experimental group and their subscale scores on the Mission: Algebra™ test of motivation. The second administration of the motivation test was separated from the post-test by a one-day interval of time, and only the scores obtained in the last Results / 79 C o n c e P t s 1 6-1 4" 1 2 1 0 8 6 4 • 2 • Male Female Low Average Divergent Feeling High R u I e s 1 6 1 4 1 2 1 0 8 6 4 2 0 L Male Female ow Average Divergent Feeling High Male Female P r o c e d u r e s Low Average High Divergent Feeling Figure 11. Interaction of Gender and L D F on Retention session of playing the game were used. The resulting correlation coefficients and their significance levels are reported in Table 16. The data presented in Table 16 indicate the following: 1. There were significant positive correlations between the post-test concepts scores Results / 80 Table 16. Correlation of Motivation and Learning MAch 2 MAcu 2 MAco 2 MAfn 2 C o n c e p t s r=0.23 p=0.049 r=-0.13 * p=0.176 r=0.27 p=0.029* r=-0.12 p=0.211 R u l e s r=0.20 p=0.081 r=0.08 p=0.291 r=0.25 p=0.041* r=0.24 p=0.045* P r o c e d u r e s r=0.02 p=0.441 r=-0.02 p=0.440 r=0.19 p=0.096 r=-0.09 p=0.266 MA refers to co to control Mission: and f n Algebra™, ch to fantasy; n: to challenge, cu = 51. to curiosity, of the experimental group and their scores on each of the challenge and control subscales of the Mission: Algebra™ test, second administration. 2. There were no significant correlations between the post-test concepts scores of the experimental group and their scores on either of the curiosity or fantasy subscales of the Mission: Algebra™ test, second administration. 3. There were significant positive correlations between the post-test rules scores of the experimental group and their scores on each of the control and fantasy subscales of the Mission: Algebra™ test, second administration. 4. There were no significant correlations between the post-test rules scores of the experimental group and their scores on either of the challenge or curiosity subscales of the Mission: Algebra™ test, second administration. 5. There were no significant correlations between the post-test procedures scores of the experimental group and their scores on any of the challenge, curiosity, control or fantasy subscales of the Mission: Algebra™ test, second administration. V . I M P L I C A T I O N S O F T H E S T U D Y A. I N T E R P R E T A T I O N S An interpretation of the results will be presented next. Intrinsic motivational effects of playing the two games will be discussed first, followed by a discussion of ensuing cognitive learning outcomes and their relationship with the motivational effects of playing Mission: Algebra™. 1. I n t r i n s i c M o t i v a t i o n a l E f f e c t s The foregoing analyses of data revealed a significant decrease in the mean of total motivation scores of the experimental group on Lode Runner™. The corresponding null hypothesis of no intra-game difference in total mean scores of the experimental group on the test was thus rejected. The results suggest that the level of motivation at playing the game decreases after a player spends some time playing the game. Additional analyses were performed to examine those factors of intrinsic motivation that contributed to the decrease. Further analyses revealed a significant decrease in the means of the challenge and curiosity subscale scores of the experimental group. The corresponding null hypotheses of no intra-game differences in the mean score of the group on each of these two factors were rejected. The means of the factors of control and fantasy decreased only marginally. The corresponding null hypotheses of no intra-game differences in the mean scores of the group on the two factors were not rejected. The results suggest that Lode Runner™ is more challenging and curiosity evoking for a beginner than an experienced player. A closer examination of the game-and the way that the study was conducted would support the foregoing argument. 81 Implications of the Study / 82 For example, a beginner would find it challenging to finish the first frame and curiosity evoking to see what the next frame looks like. However, as the player becomes more skilled at playing the game, he/she will be able to manipulate frames more easily so that the game becomes less challenging and less curiosity evoking. For the Mission: Algebra™ test, a marginal increase in the means of total motivation scores of the experimental group was detected. The corresponding null hypothesis of no intra-game difference in total mean scores of the group on the test was not rejected. Further analysis revealed that there was a significant increase in the mean of the control factor, leading to rejecting the null hypothesis of no difference in the mean scores of this factor. No significant differences were detected for the factors of challenge, curiosity or fantasy. The corresponding null hypotheses of no differences in the mean scores of these factors were not rejected. The results suggest that the level of motivation at playing Mission: Algebra™ does not change by more playing. However, an increase in the control factor was detected. The nature of the game and the way the experiment was conducted would explain this increase. At the beginning of playing the game, the students were asked to play a mission called Diamond. After finishing the Diamond mission, the students discovered that there are other paths to choose from, that they could create and edit such paths, and that they could choose the difficulty level of the problems. Motivation levels of the two games were compared with each other twice: after finishing the first phase of each game, and at the end of the last session of playing each game. Analysis of the data revealed that, for the first motivation testing session, the mean total motivation score of the experimental group on Lode Runner™ was significantly higher than the respective mean on Mission: Algebra™. The Implications of the Study / 83 corresponding null hypothesis of no significant difference in the means of total motivation scores of the group on the two games was thus rejected. The results suggest that the motivational effects of Lode Runner™ is higher for a beginner than that of Mission: Algebra™. Further analysis revealed that, for the same testing session, the mean of each of the challenge, curiosity and fantasy subscores of the experimental group on Lode Runner™ was higher than their respective mean on Mission: Algebra™. The corresponding null hypotheses of no significant differences in the means of each of the challenge, curiosity and fantasy subscores of the group on the two games was rejected. The corresponding null hypothesis for the control subscore of the group was not rejected. The results suggest that, for a beginner, Lode Runner™ is more challenging, and curiosity and fantasy evoking than Mission: Algebra™. For the second motivation testing session, the mean total motivation score of the experimental group on Lode Runner™ was not significantly different from their mean on Mission: Algebra™. The corresponding null hypothesis of no significant difference in the means of total motivation scores of the group on the two games was not rejected. The results suggest that the motivational level of the students at playing both games is the same after the students gain some playing experience. Further analysis revealed that, for the same testing session, the mean of the control subscore of the experimental group on Mission: Algebra™ was significantly higher than the respective mean on Lode Runner™. The corresponding null hypothesis of no significant difference in the means of the control subscores of the group on the two games was rejected. The corresponding null hypotheses for the challenge, curiosity and fantasy subscores of the group was not rejected. The results suggest that, as Implications of the Study / 84 the students become more experienced at playing the games, they want to have more control over Mission: Algebra™ than over Lode Runner™. Results of the analyses further indicated no significant differences in total motivation scores according to gender or level of divergent feeling over the two playing sessions of the two games. Therefore, the corresponding null hypotheses of no difference in mean total motivation scores between boys and girls or among the three divergent feeling groups over the two playing sessions of the two game was not rejected. The results indicate that none of the two games differentiates in motivation between boys and girls or among low, average and high divergent feeling groups. Further analysis revealed no significant interactions between session and gender, session and divergent feeling, gender and divergent feeling, or session, gender, and divergent feeling on total motivation scores of the two games. The corresponding null hypotheses of no interactions between session and gender, session and divergent feeling, gender and divergent feeling or session, gender and divergent feeling on total motivation scores of the two games was not rejected. Again, the results suggest that neither of the two games differentiates in motivation among subgroups determined by the factors of gender and divergent feeling over the two playing sessions. In sum, the results suggest that there are some factors that make Lode Runner™ more motivational for beginners than Mission: Algebra™. These factors include challenge, curiosity and fantasy. The motivational appeal of Lode Runner™ fades away as the players become more experienced. The players, however, seem to be motivated to control certain features of Mission: Algebra™ after spending some time playing the game. It is believed that, as the players become more familiar with Mission: Algebra™, they want to manipulate such features as paths of missions or Implications of the Study / 85 the difficulty level of the problems. The motivational effects of both games are not differentiated by gender, divergent feeling, or both, over the two playing sessions. Thus it seems- safe to say that Lode Runner™ has certain characteristics that make it more motivating than Mission: Algebra™ for the novice player. 2. Cognitive Learning Outcomes The cognitive learning outcomes of the study were examined by an achievement test given on two occasions. The test was administered one day following the last playing session of Mission: Algebra™, then, as a retention test in the fifth week after that. It includes the subtests of concepts, rules, and procedures. Results of the analysis dealing with cognitive learning outcomes revealed no significant differences in mean achievement scores between the experimental and control groups on both the post-test and retention test. The hypotheses of no differences in mean achievement scores between the experimental and the control groups on the post-test and the retention test were not rejected. The results suggest that the game is as effective in achieving cognitive learning as the more traditional worksheets, both one day following the last learning episode and 4-J weeks later. The results revealed a significant difference in mean achievement scores between boys and girls on both the post-test and the retention test. The hypotheses of no differences in mean achievement scores between boys and girls on the post-test and the retention test were rejected. Boys achieved higher scores on the post-test and the retention test than girls. Further analysis revealed that, on both the post-test and the retention test, boys achieved higher scores than girls on each of the concepts and rules parts of Implications of the Study / 86 the test. The corresponding null hypotheses of no differences in mean scores of the concepts and rules subtests between boys and girls were rejected. On both tests, there was no significant difference between the boys' and girls' mean achievement score on the procedures part. Therefore, the null hypotheses of no differences in mean scores of the procedures subtests of both tests between boys and girls were not rejected. The results further revealed no significant difference in mean achievement scores among the three divergent feeling level groups. The corresponding null hypothesis of no difference in the mean achievement scores among the three groups was not rejected. Low, average and high level divergent feeling students did not differ in their achievement on both the post-test and the retention test. On both the post-test and the retention test, no interactions between treatment and gender, between treatment and level of divergent feeling, or among all three factors, were detected. The corresponding null hypotheses of no significant interactions between treatment and gender, between treatment and level of divergent feeling, or among all three factors, on the post-test and the retention test were not rejected. There was, however, a significant interaction of gender and divergent feeling on both the post-test and the retention test. The null hypotheses of no interactions between gender and divergent feeling on both the post-test and the retention test were thus rejected. On the post-test, further analysis revealed significant interaction of gender and level of divergent feeling on the concepts subtest only. The null hypothesis of no interaction between gender and divergent feeling on the concepts part of the post-test-was thus rejected. The corresponding null hypotheses of no interactions between Implications of the Study / 87 gender and divergent feeling on the rules and procedures parts of the post-test were not rejected. On the concepts part of the post-test, boys with either low or high level of divergent feeling achieved better on the concepts subtest than girls with corresponding levels. There was marginal difference between the achievement of boys and girls with average level of divergent feeling. On the retention test, however, further analysis revealed significant interaction of gender and level of divergent feeling on all three subtests of concepts, rules and procedures. The corresponding null hypotheses of no interaction between gender and divergent feeling on the three part of the retention test were thus rejected. On the concepts and rules parts, the interaction was similar to that obtained on the post-test. Boys with low or high levels of divergent feeling achieved higher scores than respective girls. Average divergent feeling level girls achieved slightly higher scores than boys. On the procedures part of the test, the pattern was similar except that girls with high levels of divergent feeling achieved higher scores than boys. The results suggest that, except for the procedures part of the retention test, boys with a high level of divergent feeling achieve higher scores than girls with the same level of divergent feeling. A possible explanation for the results is that boys and girls with high level of divergent feeling may differ in the way they perceive creativity aspects of themselves. For example, boys and girls may have different fantasies, yet both obtain high scores on the imagination part of the test. The same argument may be true for subjects with low levels of divergent feeling, an argument which may be delimited by the small size of the group (7 boys and 5 girls took the post-test, 7 boys and 4 girls took the retention test). In sum, the results indicate that the instructional microcomputer game is as Implications of the Study / 88 effective in achieving cognitive learning scores as the more traditional worksheets. Boys achieve higher scores on the test of algebra than girls. The higher achievement of boys on the test of algebra is more observed for boys of high or low divergent feeling level than for boys of average level. 3. Motivation and Learning Analysis of the data revealed a positive significant correlation between the motivational factor of challenge and the concepts part of the test. The null hypothesis of no correlation between challenge and concepts learning was thus rejected. There were no significant correlations between challenge and rules learning or between challenge and procedures learning. The corresponding null hypotheses were therefore not rejected. The analysis further revealed no significant correlations between the motivational factor of curiosity and any of the concepts, rule or procedures part of the post-test. The corresponding null hypotheses of no correlations between curiosity and any of the concepts, rules or procedures subtest scores were not rejected. The analysis, however, revealed significant positive correlations between control and each of the concepts and rules parts of the post-test. The corresponding null hypotheses of no correlation between control and concepts learning and between control and rules learning were rejected. The correlation between the control factor and the procedures part of the test was not significant. Therefore the null hypothesis of no correlation between control and procedures learning was not rejected. There was a significant positive correlation between the factor of fantasy and the rules part of the post-test. The null hypothesis of no correlation between fantasy Implications of the Study / 89 and rules learning was thus rejected. On the other hand, there were no significant correlations between fantasy and concepts or between fantasy and procedures. The null hypothesis of no correlation between fantasy and each of concepts and procedures learning was not rejected. The results indicate that the control factor correlates positively with both concepts and rules learning. There is also a positive correlation, significant at the 0.096 level, between control and procedures learning. It is however the control factor which experienced a significant intra-game increase for Mission: Algebra™. The positive correlations so obtained may form an important finding, an indication that learning does correlate positively with motivation, thereby supporting the popular belief that motivation affects learning. Other significant positive correlations between factors which experienced marginal increase in motivational effects and parts of the post-test support the foregoing finding. Challenge correlated with concept learning, and fantasy with rules learning. Curiosity, on the other hand, experienced a marginal decrease in motivational effect and did not correlate significantly with any part of the test. 4. L i m i t a t i o n s A number of factors encountered in the research design and procedures might have affected the results obtained in the study. Firstly, a total of seventy statistics were computed and tested for significance. A significance level of 0.05 was set as an acceptable risk of Type I error. Accordingly, a number of the significant results obtained might have occured by chance. Implications of the Study / 90 Secondly, the experiment was not conducted in a strict laboratory setting. Participation of the teachers, and their request that the experiment form part of the unit of instruction the students were studying, helped create an almost natural relaxed classroom ambiance. The setting of the experiment might have therefore increased the generalizability power of the results for classroom practice. Thirdly, a variety of different monitors, single-colour or coloured, were attached to the Apple II® microcomputers. The single-colour monitors, furthermore, had either green, grey or amber screens. Colour, it was argued, could act as a distractor from other more vital cues in a learning situation (Nawrocki and Winner, 1983). Monitor type was not a studied factor in the current study, and its effects on the results thus remain unknown. Finally, while the experimental group was supervised by the investigator throughout the experiment, the control group was supervised by three different teachers. No data were collected to investigate the teacher effect. Therefore the way the teacher effect might have affected the results also remains unknown. 5. Summary The available evidence suggests that Mission: Algebra™ was less motivating than Lode Runner™ for the group of students involved. The motivational appeal of Lode Runner™ decreases as the players gain some experience at playing the game, while the motivational appeal of Mission: Algebra™ remains at a relatively lower level. Intra-game differences in total scores are attributed to the factors of challenge and curiosity, while inter-game differences are attributed to the factors of challenge, curiosity and fantasy. However, there were intra-game and inter-game differences in the scores of the control factor in the cases where no significant differences in the Implications of the Study / 91 corresponding total motivation scores were obtained. The instructional game was as effective in achieving cognitive learning scores as the worksheets, both one day following the learning episode and 4\ weeks later. Boys, however, achieved higher scores than girls on both tests. The differences in achievement scores were mainly obtained for subjects with either low or high level of divergent feeling. The corresponding differences for subjects with an average level of divergent feeling were marginal. Cognitive learning outcomes correlated positively with those motivational factors which experienced some increase over the two playing sessions. The control factor, which increased significantly over the two playing sessions, correlated positively with both concept and rule learning. Challenge correlated with concept learning, and fantasy correlated with rule learning. B. DISCUSSION A discussion of the results as they relate to theory and results from ealier research will be presented in the current section. 1. Motivation If Bruner's systems of representation (the enactive, iconic and symbolic modes of thinking) are applied to the two games used in the study, one can recognize that symbolic representations are heavily employed in Mission: Algebra™ (solving equations, etc.), but not as much in Lode Runner™. Enactive (motor) representations, on the other hand, are prominent in Lode Runner™ (manipulating the movement of the hero around treasure halls), not as much in Mission: Algebra™. Iconic modes of thinking-are not well represented in either game. Implications of the Study / 92 Kay (1984) observed that the "world of the symbolic can be dealt with effectively only when the repetitious aggregation of concrete instances becomes boring enough to motivate exchanging them for a single abstract insight." Kay cited a finding by the French mathematician Hadamard that, among 100 leading mathematicians he interviewed, the majority claimed to make no use of symbols in their thinking but were instead primarily visual in their approach: The foregoing observations may form a suitable basis for explaining the existence of various intra-game and inter-game differences in the motivational effects of the two games. On one hand, the enactive representations prominent in Lode Runner™ seem to make the game more appealing for the beginner than the experienced player. Motor representations require motor responses (thinking takes place through action and in terms of motor responses), probably making playing a game relatively easy to handle and enjoyable for the beginner. But, as a subject gains some playing experience, the game becomes easier to play and probably less enjoyable. There are no factors inherent in the game that help maintain its difficulty level and the ensuing enjoyment. The motivational appeal of Lode Runner™ is attributed more to the factors of challenge and curiosity than to control and fantasy. Challenge and curiosity aspects of the game mostly deal with "physically" manipulating the hero to achieve certain goals and thus relate to enactive representations. Control deals with symbolic representations. To control various elements of the game, a player must go through a "verbal" dialogue with the computer, hitting control keys and answering a series of questions. Fantasy, involving imagination, is iconic in nature. Although there was some initial fantasy effect, this factor is not sufficiently represented in the game. Implications of the Study / 93 Symbolic representations, on the other hand, form a salient feature of Mission: Algebra™. The involvement of the students in such more demanding modes of thinking thus seems to reduce the motivational appeal of the game. However, a marginal increase in the motivational level of the game, significant for the control factor, has been detected. It is believed that this increase in control is partly due to an activity enactive in nature. The subjects, after finishing the initial (Diamond) mission, were able to design their own missions by calling the graphs screen and moving a cursor to positions of their own choice. The foregoing argument extends to inter-game differences in motivation. In general, while the motivation test items deal with mainly enactive representations in Lode Runner™, they deal with mainly symbolic representations in Mission: Algebra™. The beginning players seem to favour the less demanding game which involves more enactive modes of thinking over the more demanding game which involves more symbolic modes of thinking. With more playing experience, however, the motivational appeal of both games is reduced. The finding of Malone (1981a) that computer games which incorporated mainly graphic and sound effects were preferred to word games, can also be interpreted within the context of modes of thinking. The preferred games in Malone's study employed mostly enactive representations, while the disliked ones employed mostly symbolic representations (verbal communication between the player and the computer). Thus, the current study adds to Malone's study that games which employ mostly enactive representations are in general more motivational (not just preferred) than those which employ mostly symbolic representations. Implications of the Study / 94 2. L e a r n i n g The treatment group achieved higher, albeit nonsignificantly, on two parts of the post-test and the retention test than the control group. The fact that the motivational appeal of the instructional game had been at a relatively low level throughout the experiment may account for the nonsignificance of the difference. The significant correlations between some learning factors and some factors of motivation would support this argument. Significant correlations occured only where there were marginal or significant increases in motivational levels, and disappeared otherwise. The involvement of the control group in the learning activity may also account for the nonsignificance of the differences in cognitive learning between the experimental and control group. Subjects in the control group were reported to be deeply involved in the activity of practicing the worksheets, indicating, according to the supervisors of this group, a high level of motivation in the activity. However, there were significant gender differences in achievement, favouring boys over girls, on both tests. Gender differences in mathematics achievement at the secondary school level have been a controversial issue. Fennema (1978) discussed various cognitive, affective and educational factors that may account for such differences. She argued that gender differences in mathematics achievement are attributed to the fact that fewer females elect to study mathematics at the secondary school level, which in turn is attributed to the females' lesser confidence in learning mathematics and to a belief that mathematics is not useful to them. An earlier study (Helson, 1971) found out that the professional achievement (measured by the number of published papers and university position) of highly creative men who obtained a Ph.D. degree in mathematics was significantly higher Implications of the Study / 95 than that of comparable women. The author suggested that such a difference reflects "social roles and institutional arrangement more than fundamental creative traits." Helson also refered to similar results involving adolescents and college students. The current study found that the mathematics achievement of lOth-grade boys with high perceived level of creativity was higher than that of comparable girls. The study thus extends Fennema's disposition that gender differences in mathematics achievement is related to the way females feel about the learning and usefulness of mathematics, to include their level of divergent feeling (how creative they feel about themselves). Helson's argument about social roles and institutional arrangement may, however, still hold as the two factors may relate to the perceptions of the students about themselves on such factors as risk-taking, imagination, curiosity and complexity. 3. Motivation and Learning A primary purpose of the current study has been to investigate the relationship between motivation and learning in a microcomputer game environment. According to Lepper and Malone (1987), the question of whether the devices used to enhance intrinsic motivation also further the learning of relevant educational content contained in a certain program is a fundamental question about which virtually nothing is known. In the current study, significant positive correlations between challenge and concept learning, between control and each of concept and rule learning, and between fantasy and rule learning were obtained. No significant correlations with the factor of curiosity were obtained. A number of computer-based learning systems have been developed based on the principle of learner control. The principle, in general, suggests that "the size of the effects on learning of the use of choice and other forms of learner control to Implications of the Study / 96 increase student motivation will vary with the extent to which control is provided over instructionally-critical aspects of the activity ... versus instructionally-incidental aspects of the program" (Lepper & Malone, 1987). Control may be instructionally-relevant, instructionally-irrelevant, or in between allowing for limited choice. In the current study, the subjects had limited control over the instructional game. Initially, the subjects played the Diamond mission while, later on, they could choose or even design their own missions, a process that resulted in a significant increase in control scores. The findings of the current study thus tend to support the principle of learner control in a microcomputer game environment. The existence of positive correlations between cognitive learning and other motivational factors, however, indicate that control is not the only factor that may affect learning. The other two factors (challenge and fantasy) which correlated with some types . of cognitive learning outcomes experienced a marginal increase only over the two playing sessions. The issue of correlations between motivation and learning thus remains open to further investigation. The indication nevertheless is that cognitive learning correlates with motivational factors which have some relatively high value. Yet, the issue of whether a motivational environment would produce better learning than a non-motivational one also remains open to further investigation. C. CONCLUSIONS AND RECOMMENDATIONS In the current section, conclusions drawn from the results of the analyses will be presented. The conclusions will mainly relate to the problem, stated in general terms in the first chapter, and to some issues raised in the review of the literature. Recommendations for software design and for further research will also be presented. Implications of the Study / 97 1. C o n c l u s i o n s The study dealt with two related issues. The first issue involved the factors which determine intrinsic motivation, and the way these factors work. The second issue involved the types of cognitive learning outcomes that can be achieved and maintained in an intrinsically motivating environment, and the way these outcomes correlate with factors of intrinsic motivation. Conducted in a microcomputer game environment, the study investigated challenge, curiosity, control and fantasy as factors of (individual) intrinsic motivation, and the acquisition and retention of concepts, rules and procedures as types of cognitive learning outcomes. a. Intrinsic Motivation Significant intra-game differences were obtained for the challenge and curiosity factors on Lode Rummer, and for the control factor on Mission: Algebra™. Significant inter-game differences were obtained for the challenge, curiosity, and fantasy factors favoring Lode Runner™, and for the control factor favoring Mission: Algebra™. The following conclusions, related to the first issue presented above, are drawn: 1. Each one of the factors of challenge, curiosity, control and fantasy played a role in determining individual intrinsic motivation. Thus the four factors are suitable for the study of intrinsic motivation. 2. In a game with a relatively high level of motivational appeal for the novice player, the four factors of intrinsic motivation may decrease as the player gains some playing experience. However, the significance of the decrease depends on the initial value of the factor. For example, challenge and curiosity experienced significant decrease for Lode Runner™, while the decrease in control and fantasy was not significant. Implications of the Study / 98 3. For a game with a relatively low level of motivational appeal, significant change in one motivational factor may be obtained, yet no such change can be expected for the other factors. In fact, to make a game motivating, software designers should deliberately manipulate the motivational factors of the game in such a way that a change in some factors would occur from time to time as long as the game is played. No significant differences in motivation according to gender or level of divergent feeling were obtained. Also no significant interactions on motivation between session and gender, between session and divergent feeling, or between gender and divergent feeling were obtained. It is thus concluded that the games do not differentiate in motivation between boys and girls, among various divergent feeling groups, or between boys or girls of different levels of divergent feeling. b. Learning and its Correlation with Motivation Relative to the second issue, no significant differences in cognitive learning outcomes between the experimental and control groups or among the three divergent feeling groups were obtained. Significant gender differences and interactions between gender and level of divergent feeling were however obtained. No other significant interactions were obtained. These results lead to the following conclusions: 1. The three types of cognitive learning outcomes (concepts, rules and procedures) can be equally well achieved and maintained in a microcomputer game environment as opposed to using the more conventional worksheets to practice the same instructional contents of the game. 2. Perceived creativity, which would be influenced by social factors, seem to be different in boys than girls. The difference in perceived creativity would thus Implications of the Study / 99 account for gender differences in mathematics achievement. Relative to the second issue also, the results indicated significant correlations between challenge and concept learning, between control and concept and rule learning, and between fantasy and rule learning. The challenge and fantasy factors, it is recalled, experienced marginal intra-game increase for Mission: Algebra™. Control experienced a significant intra-game increase, while curiosity experienced a marginal intra-game decrease and correlated significantly with none of the three types of learning. It can thus be concluded that cognitive learning correlates positively with those individual intrinsic motivations that have some relatively high value at the time when learning occurs. 2. Recommendations Based on the results of the current study, the following recommendations are presented: 1. It is recommended that software designers of instructional computer games employ the factors of individual as well as interpersonal motivation variably at different phases of a game to ensure an optimal level of motivation. 2. It is recommended that further research be carried out that investigates individual and interpersonal motivations of other types of computer games. 3. It is recommended that further research be carried out that investigates cognitive as well as affective learning outcomes of playing instructional computer games at different school levels. R E F E R E N C E S Ausubel, D. P. Education for Rational Thinking: A Critique. In Lawson, A. E. (Ed.): 1980 AETS Yearbook. The Psychology of Teaching for Thinking and Creativity. Columbus, Ohio; ERIC/SMEAC; 1979; 174-190. Ball, S. Student Motivation: Some Reflections and Projections. In Ames, R. and Ames, C. (Eds.): Research on Motivation in Education. Orlando; Academic Press, Inc.; 1984; 313-327. Berg, R. Analytical Criteria for Microcomputer-Based Simulation/Games. ERIC ED231741. June, 1983. Bitzer, D. L. - Lyman, E. R. - Easley, J. A. Jr. The Uses of PLATO. A Computer Controlled Teaching System. Audiovisual Instruction, 1966, 11:1, 16-21. Bredemeier, M. - Greenblat, C. S. The Educational Effectiveness of Simulation Games: A Synthesis of Findings. Simulation and Games, 1981, 12:3, 307-332. Bright, G. W. What Research Says: Teaching Probability and Estimation of Length and Angle Measurement Through Microcomputer Instructional Games. School Science and Mathematics, 1985, 85:6, 513-522. Bright, G. W. - Harvey, J. G. Computer Games as Instructional Tools. Computers in the School, 1984, 1:3, 73-79. Bright, G. W. - Harvey, J. G. - Wheeler, M. M. Cognitive Effects of Games on Mathematics Learning. ERIC ED 166007; 1978. Bright, G. W. - Harvey, J. G. - Wheeler, M. M. Using Games to Retrain Skills with Basic Multiplication Facts. Journal for Research in Mathematics Education, 1979, 10:2, 103-110. Bright, G. W. - Harvey, J. G. - Wheeler, M. M. Using Games to Maintain Multiplication Basic Facts. Journal for Research in Mathematics Education, 1980a, 11:5, 379-385. Bright, G. W. - Harvey, J. G. - Wheeler, M. M. Achievement Grouping with Mathematics Concepts and Skill Games. Journal of Educational Research, 1980b, 37:5, 265-369. Bruner, J. On Learning Mathematics. The Mathematics Teacher, December, 1960a, 610-19. Bruner, J. The Process of Education. Cambridge; Harvard University Press; 1960b. Bruner, J. The Act of Discovery. Harvard Educational Review, 1961, 31:1, 21-32. Bruner, J. On Knowing: Essays for the Left Hand. Cambridge; Harvard University Press; 1962. 100 / 101 Bruner, J. The Growth of Mind. American Psychologist, 1965, 20:12, 1007-17. Bruner, J. Toward a Theory of Instruction. Cambridge; Harvard University Press; 1966. Chaplin, J. P. Dictionary of Psychology. New York; Dell Publishing Co., Inc.; 1979. Cofer, C. N. - Appley, M. H. Motivation: Theory and Research. New York; John Wiley & Sons, Inc.; 1964. Coleman, J. S. Social Processes and Social Simulation Games. In Boocock, S. S. and Schild E. (Eds.): Simulation Games in Learning. Beverly Hills; Sage Publications, Inc.; 1968. Condry, J. - Chambers, J. Intrinsic Motivation and the Process of Learning. In Lepper, M. and Greene, D. (Eds.): The Hidden Costs of Reward. Hillsdale; Lawrence Erlbaum Ass.; 1978; 61-84. DeCharms, R. Motivation Enhancement in Educational Settings. In Ames, R. and Ames, C. (Eds.): Research on Motivation in Education. Orlando; Academic Press, Inc.; 1984; 275-310. Deci, E. Intrinsic Motivation. New York; Plenum Press; 1975. Dewey, J. The School and Society. (Revised Edition). Chicago; The University of Chicago Press; 1915. Dubin, S. - Okun, M. Implications of Learning Theories for Adult Education. Adult Education, 1973, 24:1, 3-19. Eysenck, M. W. A Handbook of Cognitive Psychology. Hillsdale; Erlbaum Ass.; 1984. Fennema, E. Sex-Related Differences in Mathematics Achievement: Where and Why? In Jacobs J. E. (Ed.): Perspectives on Women and Mathematics. Columbus; ERIC/SMEAC; 1978. Fletcher, J. L. The Effectiveness of Simulation Games as Learning Environment: A Proposed Program of Research. Simulation and Games, 1971, 2, 425-454. Gagne, R. M. The Conditions of Learning. New York; Holt, Rinehart and Winston; 1965. Gagne, R. M. Learning Outcomes and Their Effects: Useful Categories of Human Performance. American Psychologist, 1984, 39:4, 377-385. Greenblat, C. S. Gaming-Simulations for Teaching and Training: An Overview. In Greenblat, C. S. - Duke, R. D. (Eds.): Gaming-Simulation: Rationale, Design, and Applications. New York; John Wiley & Sons, Inc.; 1975. / 102 Hall, K. A. Computer-Based Education. In Mitzel H. E. (Ed.) Encyclopedia of Educational Research. Volume 1 (5th. Edition). New York; The Free Press; 353-367; 1982. Helson, R. Women Mathematicians and the Creative Personality. Journal of Consulting and Clinical Psychology, 1971, 36:2, 210-220. Holt, J. How Children Fail. New York; Pitman Publishing Corp.; 1964. Kay, A. Computer Software. Scientific American, 1984, 251:3, 52-59. Laurillard, D. Evaluation of Student Learning in CAL. Computers and Education, 1978, 2, 259-265. Lepper, M. R. Microcomputers in Education. Motivational and Social Issues. American Psychologist, 1985, 40:1, 1-18. Lepper, M. R. - Chabay, R. W. Intrinsic Motivation and Instruction: Conflicting Views on the Role of Motivational Processes in Computer-Based Education. Educational Psychologist, 1985, 20, 217-231. Lepper, M. R. - Greene, D. - Nisbett, R. E. Understanding Children's Intrinsic Interest with Extrinsic Rewards. Journal of Personality and Social Psychology, 1973, 28, 129-137. Lepper, M. R. - Malone, T. W. Intrinsic Motivation and Instructional Effectiveness in Computer-Based Education. In Snow, R. and Farr, M. (Eds.): Aptitude, Learning, and Instruction. Volume 3: Conative and Affective Process Analysis. Hillsdale; Lawrence Erlbaum Ass.; 1987; 255-286. Levine, F. M. - Fasnacht, G. Token Rewards May Lead to Token Learning. American Psychologist, 1974, 29, 816-820. Loftus, G. R. - Loftus, E. F. Mind at Play. The Psychology of Video Games. New York; Basic Books, Inc.; 1983. Malone, T. W. Toward a Theory of Intrinsically Motivating Instruction. Cognitive Science, 1981a, 4, 333-369. Malone, T. W. What Makes Computer Games Fun? Byte, Dec. 1981b, 258-274. Malone, T. W. - Lepper, M. R. Making Learning Fun: A Taxonomy of Intrinsic Motivations for Learning. In Snow, R. and Farr, M. (Eds.): Aptitude, Learning, and Instruction. Volume 3: Conative and Affective Process Analysis. Hillsdale; Lawrence Erlbaum Ass.; 1987; 223-253 McGraw, K. The Detrimental Effects of Reward on Performance: A Literature Review and a Prediction Model. In Lepper, M. and Greene, D. (Eds.): The Hidden Costs of Reward. Hillsdale; Lawrence Erlbaum Ass.; 1978; 33-60. / 103 Merrill, M. D. Learner Control in Computer Based Learning. Computers and Education, 1980, 4, 77-95. Nawrocki, L. H. - Winner, J. L. Video Games: Instructional Potential and Classification. Journal of Computer-Based Instruction, 1983, 10:3 & 4, 80-82. Orbach, E. Simulation Games and Motivation for Learning. Simulation and Games, 1979, 10:1, 3-40. Papert, S. Mindstorms. Children, Computers, and Powerful Ideas. New York; Basic Books, Inc.; 1980. Piaget, J. The Origins of Intelligence in Children. New York; International Universities Press, Inc.; 1952. Piaget, J. Play, Dreams and Imitation in Childhood. New York; Norton & Co., Inc.; 1962. Pierfy, D. A. Comparative Simulation Game Research: Stumbling Blocks and Steppingstones. Simulation and Games, 1977, 8:2, 255-268. Richmond, P. G. An Introduction to Piaget. London; Routledge & Kegan Paul Ltd.; 1970. Scandura, J. M. Role of Rules in Behavior: Toward an Operational Definition of What (Rule) is Learned. Psychological Review, 1970, 77:6, 516-33. Skinner, B. F. Why We Need Teaching Machines. Harvard Educational Review, 1961, 31:4, 50-71. Spraggins, C. C. - Rowsey, R. E. The Effect of Simulation Games and Worksheets on Learning of Varying Ability Groups in a High School Biology Classroom. Journal of Research in Science Teaching, 1986, 23:3, 219-229. Stipek, D. J. The Development of Achievement Motivation. In Ames, R. and Ames, C. (Eds.): Research on Motivation in Education. Orlando; Academic Press, Inc.; 1984; 145-174. Weiner, B. Principles for a Theory of Student Motivation and Their Application within an Attribution Framework. In Ames, R. and Ames, C. (Eds.): Research on Motivation in Education. Orlando; Academic Press, Inc.; 1984; 15-38. Williams, F. E. Creativity Assessment Packet (CAP). New York; D.O.K. Publishers, Inc.; 1980. APPENDIX A LODE RUNNER QUESTIONNAIRE Student Name This questionnaire is to find out how you felt about the last frame you finished, that is when the hero started moving to collect chests of gold until he reached the next frame. Your answers will help us understand how games motivate people, and how motivation can be used for better learning in the school. The questionnaire is not a test There are no right or wrong answers What vou have to do 1. Read the first statement carefully then cross (X) the one box that best indicates your feeling about the statement. Remember, all items concern your feelings while you were playing the last frame. Cross under SA if you strongly agree with the statement A if you Just agree with the statement D if you just disagree with the statement SD if you strongly disagree with the statement Here is an example SA A D SD 1. I liked playing this game X If you strongly agree with this statement then you should cross SA. 2. Go on to each of the other statements in the order set. 3. Answer all statements. 4. Do not go back to previous answers. 5. If you make a mistake, cross out the wrong answer and put in the new mark clearly. 6. Remember this is not a test. Give your own honest opinion for each statement (NOT what you think we might like you to say). 7. Ask the teacher in charge if there is any statement you do not understand. 8. The questionnaire will take 7 minutes to complete. Do not spend too much time on any one item. 104 / 105 SA A D SD 1. I related guards to people who work for a wrong cause 2. I felt better about myself everytime I retrieved a chest of gold 3. I wanted to guess which guards carried concealed chests of gold 4. I wanted to know how this game can be played better 5. I wanted to have the power to control the number of guards 6. I wanted to know how well I was playing 7. I wanted to know how the game ends with the last frame 8. I wanted to have the power to design my own frames 9. I wanted to know what graphic and sound effects occur when I finish a frame 10. I imagined I was the hero everytime I trapped a guard 11. I wanted to have the power to control the number of chests of gold 12. I wanted to know what graphic and sound effects occur when a guard is trapped in a ditch I dig 13. I compared retrieving a chest of gold with doing good things for society 14. I wanted to know where a new guard will appear when a guard disappears In a ditch I dig 15. I imagined I was the hero everytime I retrieved a chest of gold 16. I wanted to trap more guards 17. I felt better about myself by finishing the last frame 18. I wondered if the score I get really depends on how many chests of gold I retrieve 19. I wanted to control the movement of guards 20. I wanted to get a high score 21. I wondered if the score I get really depends on how many guards I trap 22. I wanted to finish this frame 23. I wanted to know what graphic and sound effects occur when I make a mistake 24. I wanted to get more chests of gold 25. I thought of good people in society similar to the hero in the game, all doing right things 26. I wanted to play with frames of my own choice 27. I wanted to know what the next frame looks like 28. I wanted to know what graphic and sound effects occur when I retrieve a chest of gold 29. I felt better about myself everytime I trapped a guard 30. I imagined I was the hero when I finished the last frame APPENDIX B MISSION: ALGEBRA QUESTIONNAIRE Student Name This questionnaire Is to find out how you felt about the last mission you finished, that is when the hero started solving the first linear equation until he saved the lost ship Your answers will help us understand how games motivate people, and how motivation can be used for better learning in the school. The questionnaire Is not a test What Y O U have to do There are no right or wrong answers Read the first statement carefully then cross fX) the one box that best Indicates your feeling about the statement. Remember, all items concern your feelings while you were playing the last mission. Cross under SA if you strongly agree with the statement A if you just agree with the statement D if you just disagree with the statement SD if you strongly disagree with the statement Here is an example SA A D SD 1. I liked playing this game X If you strongly agree with this statement then you should cross SA. 2. Go on to each of the other statements In the order set. 3. Answer all statements. 4. Do not go back to previous answers. 5. If you make a mistake, cross out the wrong answer and put in the new mark clearly. 6. Remember this is not a test. Give your own honest opinion for each statement (NOT what you think we might like you to say). 7. Ask the teacher in charge if there is any statement you do not understand. 8. The questionnaire will take 7 minutes to complete. Do not spend too much time on any one item. 106 / 107 SA A D SD 1. I wondered if the score I get really depends on how many points I plot on the graph 2. I wanted to know what the path of the next mission looks like 3. I wanted to plot more points on the graph 4. I wanted to have the power to control the number of points to be plotted on the graph 5. I wanted to have the power to control the number of equations to be solved 6. I imagined I was the hero everytime I solved an equation that helps save a lost ship 7. I wanted to get a high score 8. I related solving equations to solving actual daily-life problems 9. I wanted to know what graphic and sound effects occur when I plot a point on the graph 10. I wanted to know how well I was playing 11. I imagined I was the hero when I finished the last mission 12. I wanted to make guesses about the position of the lost ship 13. I wanted to know what graphic and sound effects occur when I finish a mission 14. I wanted to solve more equations 15. I felt better about myself by finishing a mission of saving a lost ship 16. I wanted to know what graphic and sound effects occur when I make a mistake 17. I thought of good people in society similar to the hero in the game, all doing right things 18. I wanted to control the difficulty level of the equations that have to be solved 19. I felt better about myself everytime I solved an equation that helps locate the lost ship 20. I imagined I was the hero everytime I plotted a point on the graph to help save a lost ship 21. I wanted to play with missions of my own choice 22. I wanted to finish this mission 23. I wanted to know what sort of equation comes after the one I have been solving 24. I wanted to have the power to design my own missions 25. I compared plotting points to trace a lost ship with doing good things for society 26. I wanted to know what graphic and sound effects occur when I solve an equation 27. I felt better about myself everytime I plotted a point to help locate the lost ship 28. I wanted to know how the game ends with the last mission 29. I wondered if the score I get really depends on how many equations I solve 30. I wanted to know how this game can be played better APPENDIX C TEST IN ALGEBRA Part I Name: Purpose: The following items measure how well you understand some algebraic concepts and rules. Instructions: This section contains 32 items. Read each item and choose the answer you think is correct. If the correct answer is not given, choose "none of these." To answer an item, just circle the letter that precedes the correct answer. Answer all items. If you make a mistake, cross out the wrong answer and put in the new mark clearly. Sample items: A 2 + 3 = a 4 (j>)5 c 6 d 23 e none of these. the letter "b" is circled because 2 + 3 = 5 R Which of the following numerals means the same as six? a 4 b 5 c 6 d 60 e none of these. Which letter should be circled? Time: This section takes 13 minutes to complete. Do not spend too much time on any one item. 108 / 109 1. If n = 3, then -(2n - l)/2 = a -11 /2 b -7 /2 c - 5 /2 d -13 /2 < e none of these. 2. If 3x - 4y - 7 = 9, then 3x - 4y = a 16 b 2 c -2 d -16 e none of these. Questions 3-4 refer to figure 1. i±:|:±t±i:dj .;..;...;...;..-U^...;.. | rj. Figure 1 3. The equation of TU is y - x = 3. If D then the coordinates of D are a 0 b (0,3) c (3,0) Is a point on TU wi th y = 3, d 3 e none of these. 4. The equation of VW is y = x + 9. If B then the coordinates of B are a -8 b 1 c (1,-8) is a point on d (-8,1) VW wi th x = -8, e none of these. 5. If x = 0, then 3{2x - 4)/(x - 2) = a 3 b 6 c 1/2 d -1 /2 e none of these. 6. -3(2x - 5y + 4x) = a 15y - 18x b 15y + 18x C -3xy d 15y - 24x B none of these. 7. 3 x 4 - 6 + 2 = a -3 b 3 c -12 d 9 B none of these. 8. -4(-5x - l ) / (5x + 1) = a - 4 b 4 c 0 d 4(5x + l ) 2 e none of these. 9. If x + y - 3 = 0, then x + y + 1 = a -4 b 4 c -2 d 2 e none of these. 10. 19 - (-x)(2y) = a 38xy b -38xy c 19 - 2xy d 19 + 2xy e none of these. / 110 11. If x = 2 and y = -1, then 3 - 4xy = a 2 b 24 c -5 d 11 e none of these. Questions 1 2 - 1 4 refer to figure 2. 9 , f 4 " E 12. W h i c h of the following cannot be represented on the number line? a 7 b 0 c -10 d (2,3) e none of these. 13. Point G on the number line represents: a (0,-7) b (-7,0) c (-8,0) d (0,-8) e none of these. 14. Point E on the number line represents: a (0,0) b (0,-1) c -1 d (-1,0) e none of these. 15. 5x(-2 + 3x) = a 15x2 - 10x b 15x2 - 10 C 8x2 - 10x d 8x2 - 7x e none of these. 16. If x = -1, then 4(x - 2) + 3(4 - 3x) = a 3 b -33 c 9 d -63 e none of these. 17. (-3x)(4 - x) = a 3x - 12x b - 1 2 x - 3 x 2 c -12x + 3x d 3 x 2 - 1 2 x e none of these. 18. If 7x + 13 = 8, then 7x = a -21 b 21 c -5 d 5 e none of these. 19. W h i c h of the following is not a right operation i n simplifying the expression 2(x + 3) - 3(x - 4)? a multiply 3(x-4) b multiply -3(x-4) C subtract (x-4) from (x+3) d multiply 2(x+3) e none of these. 20. If x + 4 = 3, then x + 1 = a -1 b 0 c 1 d 8 e none of these. 21. -(9y - 15) = a - 9 y - 15 b 9y - 15 c 15 - 9y d -15 - 9y e none of these. 22. If x - 2y = 7, wh ich of the following is true? a x-2y + 7 = 0 b 2y-x-7 = 0 C 2y + x + 7 = 0 d x-2y-7 = 0 e none of these. 23. -(9x - 3)/(-3x + 1) = a -3x + 1 b -3 c 0 d 3 e none of these. 24. If - x - 7y + 9 = 1, then -7y = a x + 1 0 b x + 8 c x - 8 d x - 10 e none of these. / 111 25. 6 - (4x - 2) = a 6 - 2x b 4 - 4x c -4 - 4x d -8 - 4x e none of these. 26. If x - y - 5 = 6, then y - x = a -11 b 11 c -1 d 1 e none of these. Questions 27 - 29 refer to figure 3. ...j...j..-j..r9 .^.;...;...j...j...; j.-.j. "f|+nTffli::4::!: ...:...:...:...:...:...:...:...:...:..4. :f::|5::Ei:Hl:I± ± m m ± ::i:±:i:±:j::tM:: 27. W h i c h of the following is not represented on the XY-plane? a (-5,-2) b (8,-2) c (-6,2) d (9,-3) e none of these. 28. The coordinates of point H are a (8,2) b (-8,2) c (-8,-2) d (7,-2) e none of these. 29. The coordinates of point K are a (8,-1) b (-1,8) c (8,1) d (1,8) e none of these. 30. W h i c h of the following is a right operation i n finding the value of the expression - 4 - 16 + (-8)? a divide -20 by -8 b divide 16 by 8 C divide 20 by 8 d subtract 16 from -4 e none of these.  31. Assume that a = b and c.O. Which of the following sentences is not true. aa+c = b + c ba-c = b-c Cac = bc d a/c = b/c e none of these. 32. If a = 2, then 3 a 2 - 4 = a 0 b 2 c_8 d -48 e none of these. APPENDIX D TEST m ALGEBRA Part II Name:  Purpose: The following items measure how well you can solve linear algebraic equations in two variables. Instructions: This section contains 8 questions. Solve for y each of the eight questions; that is, reduce each equation into the form y = mx + b. In solving an equation, all steps must be written down. Try not to miss any single step that you find adequate for solving the equation even if you think it is obvious. Here, we are interested in procedures, that is in the steps needed to solve linear equations in two variables. Answer all questions in the space provided. Time: This section takes 12 minutes to complete. Do not spend too much time on any question. 112 / 113 1. 7(y + 4) - 6 = 7x - 41 2. -4(y - 2) - 4 = 4(2x - 17) 3. 3(3y - 17) = 9(x - 3) + 3 4. 5(y - 4) + 3 = lOx + 73 5. -12(y - 7) + 27 = 3(4x + 73) 6. -6(y - 3) - 4 = -18x - 118 7. -8(y + 7) - 8 = 24x + 96 8. 2y + 20 = 2(-x + 6) + 8 APPENDIX E 1. 2y + 25 = 2(2x + 1) + 7 y = Given x = 8 y = Plot poin t ( , ) 2. y + 4 = 3x + 14 y = Given x = -2 y = Plot point ( , ) 3. 9y - 40 = 9(-x + 11) + 5 y = Given x = 6 y = Plot point ( , ) 4. y + 7 = -2x + 7 y = Given x = -4 y = Plot point ( , ) y X 114 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0098171/manifest

Comment

Related Items