Ilium/llIllll!!!1W!“I”!!!”WNWllllllflllllllll THESE ( ., ,.,«.a.~ 1aLjd-u! _. ...:L.. .. a. 533 0439 yak L LEAR. Yk ‘ A WWII” “IMJIR 3f? W573 - .- 4:}Au'.‘ vkuh ‘ '1“; MVBW ‘Mfl?q"-- This is to certify that the thesis entitled A COMPARISON OF THE USE OF MICROCOMPUTER GRAPHICS SYSTEMS VERSUS A CENTRAL BATCH FACILITY IN AN INTRODUCTORY UNDER- GRADUATE COURSE ON COMPUTER GRAPHICS presented by William John Kolomyjec has been accepted towards fulfillment of the requirements for Ph.D. degree inAdmin. G Higher Ed. Major professor Date M f/ 0-7 639 RETURNING MATERIALS: Place in booE drop to remove this checkout from i your record. FINES will be charged if book is MSU LIBRARIES y l returned after the date “ stamped below. Copyright by WILLIAM JOHN KOLOMYJEC 1981 ABSTRACT A COMPARISON OF THE USE OF MICROCOMPUTER GRAPHICS SYSTEMS VERSUS A CENTRAL BATCH FACILITY IN AN INTRODUCTORY UNDER— GRADUATE COURSE ON COMPUTER GRAPHICS BY William John Kolomyjec Computer graphics instruction to undergraduate students beckoned to be reevaluated. The microcomputer had inter— active graphics capabilities whereas batch mode processing of computer graphics was passive. The study was an objective comparison relative to student achievement and attitude, as well as cost. This kind of study can provide the framework for decision making relative to changing computer technology which should be performed routinely and should be based on evidence. A review of scholarly literature for the study centered on three topical areas: 1) teaching with computers, 2) com- puter graphics instructional materials, and 3) the use of microcomputers in education. There were two independent or design variables: 1) processing mode, and 2) students' major types. The dependent variables were achievement scores and attitude measures. Treatment groups were assigned using proportionate strati- fied sampling. The experiment was conducted over two terms William John Kolomyjec and the data were combined. The achievement measure was the students' final grades in the course. The attitude measures were taken from a survey instrument. Post test attitude was adjusted by pretest attitude. The design matrix was com- pletely crossed and balanced. The two-way design allowed three separate investigations on the same subjects. Cost observations were made via cost models developed for each processing mode. Significant differences were found in achievement be- tween major types. Also, there was a significant interaction effect between major type and processing mode on the attitude toward computer systems scale. The cost models showed that costs in the batch mode increased directly proportionate to use, while in the microcomputer mode, costs decreased with increased total enrollment, increased numbers of students per microcomputer graphics system, and prolonged equipment life. The microcomputer compared favorably to batch and was, therefore, the preferred processing mode. The microcomputer was recommended for introductory computer graphics instruc- tion, particularly for engineering and computer science majors. DEDICATION To my wife Joyce and our son Tristan ACKNOWLEDGMENTS There are several noteworthy individuals who were directly responsible for providing me with the motivation to pursue a Doctor of Philosophy degree and others who are directly responsible for facilitating my achievement. I thank Professor Henry Krause for my first opportunity to teach at the post secondary level in the Department of Engineering Instructional Services at Michigan State Univer- sity. Professor Krause urged me to obtain the Ph.D. and to continue teaching. I owe the most thanks to Professor James R. Burnett. Professor Burnett was my mentor as a teacher and as a student of engineering graphics and computer graphics. He distin- guished both my M.F.A. and Ph.D. committees with his service. No other person has had a more profound influence on my adult and professional life. I am honored and proud to have him as a friend. Others to whom I am deeply indebted are Dr. Richard L. Featherstone, Dr. Norman T. Bell, and Dr. Walter F. Johnson. These competent professionals comprised the remainder of my committee. Dr. Featherstone directed my study and provided fl me with self-confidence. Dr. Bell provided expert advice in many aspects of research and teaching using microcomputers. Dr. Johnson often encouraged me and offered excellent sug- gestions to assist me in writing this dissertation. I thank Two other individuals deserve special mention. I thank Dr. Howard Teitelbaum for the many hours we spent together designing the study and the time he spent explaining statis- tical procedures to me. Also, I thank Dr. Thomas Freeman / who, as Director of the Office of Institutional Research at J Michigan State University, found time to help me formulate a strategy for the evaluation of costs. His input gave me an administrative point of View associated with the problems of determining costs in higher education. Finally, I thank my wife Joyce who supported me during this endeavor. I could not have completed either the course- work or the dissertation without her excellent editorial and Clerical skills and, most important, her encouragement. I iv TABLE OF CONTENTS CHAPTER I. INTRODUCTION Background . . . . Identification of the Problem Statement of Research Questions Need for the S udy . Purpose . . . Importance . Generalizability . . . Limitations of the Study . . Overview of Subsequent Chapters II. REVIEW OF LITERATURE Overview . . . . Computers in Education . . . Computers in Education and Training. Computer Assisted Learning . . Instructional CAL Revealatory CAL Conjectural CAL Emancipatory CAL Computer Managed Learning Informatics and Education Computer Graphics Instruction . General and Introductory Materials Engineering and Technical Materials The Use of Microcomputers in Education Teachers' Attitudes Toward Micro- computing . . . . . Microcomputers in Schools Microcomputers and Learning _ Microcomputers and Time Sharing Summary . . Page 1 3 3 5 5 7 8 9 14 16 l6 l6 I7 19 20 20 20 23 26 30 32 35 39 4O 49 52 56 CHAPTER Page III. RESEARCH QUESTIONS AND HYPOTHESES, DEFINITIONS OF IMPORTANT TERMS . . . . 67 Introduction . . . . . . . . . 67 Research Questions . . . . . . . 69 Research Hypotheses . . . . . . . 69 Achievement Hypotheses . . . . . . 71 Attitude Hypotheses, Attitude Toward Computer Graphics . . . . . . . 71 Attitude Hypotheses, Attitude Toward Computer Systems . . . . . . . 72 Significance Level . . . . . . . 72 Observations Relative to Cost . . . . 73 Definition of Important Terms . . . . 74 IV. DESIGN AND PROCEDURES . . . . . . . 86 Introduction . . . . . . . . . 86 Design . . . . . . . . . . 88 Design Over Time . . . . . . . 88 Design Over Time - Achievement . . . 89 Design Over Time - Attitude . . . . 90. Design Over Variables . . . . 91 Two-Way Analysis of Variance . . . . 91 Two-Way Analysis of Covariance . . . 92 Design Matrix . . . . . . 92 Summary Data Format . . . . 94 Statistical Models and Techniques of Analysis . . . . . . . . . 94 Reliability and Validity Concerns . . . 95 Procedures . . . . . . . . . 97 Population and Sample . . - - 97 Treatment . . . . ._ . 101 Instrumentation and Data Collection 103 Achievement Measure . . . - . - 103 Attitude Measures . . . . . . . l8? Considerations . . . . . . . 108 Procedures . . . . . . . . . 110 Implementation . . . . . . - . Preliminary Study - Form 1 and Form 2 . . iii Preliminary Study - Combined Form . . . 119 Cost Model . . . . . - . ~ - 120 Batch Processing.Mode Cost Model . . . Microcomputer Processing.Mode Cost 120 Model . . . . - - ° ° ° vi CHAPTER Pa e V. RESULTS . . . . . . . . . l 4 Introduction . . . . . . . 124 Sampling Results . . . . . . 124 Combining Data . . . . . . . 125 Statement of Research Hypotheses in Statistical Terms . . . . . . 134 Achievement . . . . . . . . 134 Graphical Results - Achievement . . . 138 Attitude . . . . . . . . 140 Attitude Towards Computer Graphics . . 141 Graphical Results - Attitude Towards Computer Graphics . . . . . . 145 Attitude Toward Computer Systems . . 147 Graphical Results - Attitude Toward Computer Systems . . . . . 149 Cost Model . . . . . . . . 149 Cost Strategy . . . . . . . 151 Cost Graphs . . . . . . 153 VI. SUMMARY AND CONCLUSIONS . . . . . 161 Introduction . . . . . . 161 Discussion of the Research Questions 164 The Study in Perspective . . 173 Implications for Further Research 178 APPENDIX A. Curriculum Codes, College of Engineering 181 B. Course Outline . . . . - - .’ IE4 Syllabus . . . . . - - - 185 Assignments One Through Eight _ . Achievement Score Summary Data Winter 196 Term 1981 . . . . - : - - Achievement Score Summary Data Spring 197 Term 1981 . . . . . - - - . . . 198 Distribution of Scores - - ~ ° C. Request for Confidential Information . . $38 Reply from Office of the Provost . . Memorandum to University Committee on Research 201 Involving Human Subjects (UCRIHS) . . 204 Reply from UCRIHS . . . . . . . vii APPENDIX D Questionnaire Statement Questionnaire Preliminary Versions, Form 1 (Red Version) Questionnaire Preliminary Versions, Form 2 (Blue Version) . Response Data, Form 1 (Red Version) Response Data, Form 2 (Blue Version) Combined Preliminary Questionnaire (Pretest Version) . Response Data, Combined Form Improving Reliability of Scales Factor Analysis of Scales Combined Questionnaire (Post Test Version). Experimental Data E. Shuffle Algorithm . F. Historical Cost Data EGR 270 G. Analysis of Variance by Term - Achievement . IL Analysis of Covariance by Term - Attitude . . . BIBLIOGRAPHY viii Page 205 206 207 208 209 210 211 212 214 215 216 217 219 220 222 226 LIST OF TABLES Table Page 1. Reporting Format — Two-Way Design . . . . 93 2. Enrollment Figures for EGR 270 1976-1980 . . 98 3. Pretest and Post Test Item Arrangement Generated by Shuffle Algorithm . . . . . 102 4. Question Areas . . . . . . . . . 112 5. Question Senses . . . . . . . . . 114 6. Combined Questionnaire Source and Sense . . . 116 7. Sampling Results, Winter Term 1981 . . . . 126 8. Sampling Results, Spring Term 1981 . . . . 127 9. Sampling Results, Combined . . . . . . 128 10. T-Test Summary Statistics, Achievement Scores . . . . . . . . 130 ll. T-Test Summary Statistics, Attitude Factor - Graphics . . 131 12. T-Test Summary Statistics, Attitude Factor - Systems . . 132 13. Design Matrix, Achievement - 135 Combined Terms . . - 14- ANOVA Table, Achievement — 135 Combined Terms . - - ' ' ' ' ix Table 15. 16. 17. Design Matrix, Attitude Toward Computer Graphics - Combined Terms . . ANCOVA Table, Attitude Toward Computer Graphics - Combined Terms Design Matrix, Attitude Toward Computer Systems - Combined Terms ANCOVA Table, Attitude Toward Computer Systems - Combined Terms 143 143 Figure l. 10. 11. LIST OF FIGURES Page Flow Chart Example. Rectangle Subroutine . . . .. 12 EGR 270 Enrollment Profile 1976-1981 . . . 100 Achievement Means. Winter 1981, Spring 1981 and Combined Terms . . . . . . . 139 Attitude Means, Computer Graphics Measure. Winter 1981, Spring 1981 and Combined Terms . . . . . . . 146 Attitude Means, Computer Systems Measure. Winter 1981, Spring 1981 and Combined Terms . 150 Comparison of Cost Per Student Per Term by Number of Students in Each Mode . . . . 155 Number of Students Per Station By Cost Per Student Per Term . . . . . . 156 Cost Per Station (Hardware) By Cost Per Student Per Term . 157 Yearly Maintenance Cost by Cost Per 158 Student Per Term . . - - Years of Usage By Cost Per Student p . . 159 er Term . . - - ° Labor Cost Per Term By Cost Per Student . . 160 Per Term . . . ' xi 31': '1 p1) CHAPTER I INTRODUCTION Background In higher education and administration, decision makers often are faced with difficult choices concerning all aspects of the University. None may be more difficult than those which involve the utilization of technology. The acquisition and use of computer hardware affects many depart— ments and colleges. Institutions are becoming more, and not less, dependent upon computers and computer technology. Pro- viding adequate computer services can be a frustrating ex- perience for academic administrators. Not only is there concern for constant updating of computer facilities but also there is the problem of equipment becoming obsolete. Major financial commitments to equipment configurations that soon prove unusable can be disastrous. Compounding the issue is the rate at which this hard- ware continues to evolve. Since the invention of computers thirty years ago, there is already a fourth generation of this equipment. Roughly described, these four generations have been identified as follows: First generation 1951-1959 Computers with vacuum tubes Second generation 1959-1965 Vacuum tubes replaced by transistors Third generation 1965—1971 Solid state integrated circuits replaced larger electronic components Fourth generation 1971- Minicomputers and micro- Present computers (Demel, et a1, 1979) Directly related to the evolution of computers is graphical output or, as it is more commonly called, computer graphics. Originally this type of output was produced by a line printer. The first graphic device can be traced back to 1957 and was simply a line drawing, numerical controlmachine. Computer graphic systems did not evolve until 1963 and credit for this pioneering work can rightfully be given to Sutherland at the Massachusetts Institute of Technology Lincoln Labora- tory for a system he developed called SKETCHPAD (Coons, 1967). Soon after, IBM developed the DAC—l graphic system for General Motors (Jacks, 1967). Computer graphic equipment developed rapidly along with the third generation computers as peripheral devices. How- ever, it was not until the fourth generation devices that graphics capabilities were integrated into the design of the total system. Since this equipment is highly accessible and imagery is easy to produce, computer graphics can be con- sidered as a contributing factor to the overall popularity of small or "personal” size computers. Along with this popu- larity, significant demand for graphics software and instruc- tion has arisen. Identification of the Problem Computer graphics, in a general sense, is the formula— tion, display or processing of visual imagery by way of digi- tal computers and peripheral devices having graphics capabil- ities. The significance of computer graphics lies in the fact that it can be considered to be an instrument which can be applied within.many discipline areas. The instruction of computer graphics to undergraduate students beckons to be reevaluated. Presently, introductory computer graphics at Ifichigan State University, and perhaps other institutions ifith centralized computer facilities (third generation com— puters), is accomplished in a passive sense by the batch nmde. In the batch mode, submitted program decks are grouped for processing at the convenience of the machine. The tech— nique is similar to playing a piano and having to waitseveral hours to hear the resultant sounds. The procedure of sub- nutting graphics programs punched into Hollerith cards, deck fashion, for batch mode processing is seemingly antiquated, cumbersome and needs to be compared to alternatives made available by today's technology. §Iatement of Research Questions The intent of this study was to make a comparison of the present practice of batch mode graphics with a popular nucrocomputer system with graphics capabilities. Both modes lmve their advantages and disadvantages. What are the per- tinent issues both to higher education and administration? The higher education issues are qualitative and have to do imth improving curriculum - concerns that every academic lnfit must face if it is to provide relevant instruction. The administrative issues have to do with accountability and quantitative concerns: justification of actions relative to students, faculty, and equipment vis-a-vis cost assessment. The independent variables in the study were: 1) pro- cessing mode, central batch facility with graphics periph- erals, and microcomputer systems with graphics capabilities; and 2) student type by major, two distinct subpopulations of students enrolled in the course; engineering and computer science majors, and engineering arts and other major types. General research questions relevant to these variables were answered. First, how did processing mode or student major type affect performance in an undergraduate introductory course in computer graphics? Secondly, how were student attitudes affected by processing mode or student major type? Lastly, could cost models be formulated which reasonably replicated processing modes and what observations could be made with them to aid decision making? Based on these gen- eral research questions, it became clear that the dependent variables in this study were to be student achievement and student attitude. Need for the Study "The era of microelectronics, microprocessor and per- sonal computers has made significant inroads into higher education and this technology lends particular aid to the training of graphics in engineering” (Zinn, 1978). Very little formal research has been done comparing batch mode graphics to microcomputers with graphics capabilities. With the appropriate configuration of popular hardware it was feasible to assemble a system for computer graphics instruc— tion and problem solving. Also, it was possible to restruc- ture a computer graphics curriculum around a relatively inex- pensive microcomputer having graphics display capabilities. A study such as this was needed to provide a means for com- parison - evidence to those in the business of computer graphics instruction. Purpose With the new technology it seemed that the present situation beckoned to change. How would a perceptive deci— sion maker respond to what appeared to be a challenge to improve and update present practice? The inconvenience ex— perienced by undergraduates was excessive. Although the central batch facility has certain advantages for non-graphics applications, it is less than ideal for instructing under— graduates in contemporary computer graphics concepts. An alternative needed to be considered if there was any hope of keePing curriculum up to date. The primary purpose of the study was to evaluate the use of a microcomputer with graphics capabilities in an existing undergraduate introductory course in terms of stu- dent achievement. Two design variables were identified to address this issue: processing mode (microcomputer and batch), and two distinct student major types. Thus, with respect to these variables, it was intended to focus upon performance in the course. The second major purpose of this study was to examine student attitudes. An attitudinal survey identified two com- ponents to student attitude: 1) attitude toward computer graphics, and 2) attitude toward computer systems. Once these attitude measures were established, the intent was to use the design variables to examine these attitude factors. Lastly, the administrative considerations of using one methodology over the other needed to be investigated. Ad- ministrators often are faced with making short-term and long- term policy decisions. On one hand was a need to make com- parisons between processing mode; on the other hand was the necessity for hard data in order to make the right decision based on evidence. How would one justify the need for addi— tional resources if the alternative was thought to be the most appr0priate course of action? What factors would be involved relative to cost? This study attempted to identify and address administrative concerns by making a cost model of each processing mode. Importance This study was important for the following four reasons. First, the present delivery system is outdated. Graphics processes and algorithms can, and should be executed on sys- tems with real time capabilities. The effectiveness of teaching in these modes needed.to be determined. Secondly, most of the literature indicated that student motivation would be increased by direct interaction with microcomputers. In a typical article, it has been stated that this "teaching approach at the (Lawrence) Hall (of Science) provides hands on participatory exercises that will promote discovery learn- ing . . . for computer education this means that students have a chance to create, write and debug their own programs while working at a computer terminal" (Hakensson and Roach, 1979). A third reason for this study's importance was that student productivity, i.e., that which can be achieved in a given time frame, should be increased. By removing the in- convenience of accessing a central batch facility with its slow turnaround time, students should be able to Spend more time, not only solving course assignments, but also inves- tigating variations thereof. Lastly, students' attitudes about computing in general should be affected. "Now personal computers with their low costs, easy accessibility, total dedication to the user, and person-on-the-street popularity may provide the long awaited catalyst that is needed to make some dramatic change in how computers are used in schools" (Bell, 1979). Generalizability It is worthwhile indicating that the findings of this study may have an impact far beyond the limits of the study itself. First, the relative ease of learning computer image— generating techniques on a microcomputer graphics system can increase the student's eagerness to learn more about other aspects of computing and apply this knowledge to other disci- pline areas outside engineering and the University setting. These devices are affordable and portable and can be used in many academic applications, as well as the home or office. Second, "computer graphics is a powerful problem-solving tool for the engineer and scientist; students who enter these fields must be familiar with the use of computer graphics" (Demel, et a1, 1979). This is the opening statement from a text entitled Computer Graphics. The greatest application of computer graphics lies in the areas of Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM). To quote the textbook again: CAD and CAM had their beginnings in the first use of com- puter graphics systems in the design process: automated drafting, an outgrowth of numerical control. The aero- space and automotive industries were leaders in the CAM application. The electronic industry quickly adapted computer graphics, in particular interactive computer graphics, to the making of printed circuit and integrated circuit masks so that by 1976 75% of all Computer Aided Design work was electronic applications. The remaining 25% was devoted to mechanical, electrical, cartographical and architectural work. (Demel, et al, 1979) ’ Teaching introductory computer graphics to students preparing themselves fer careers in engineering and science nmy provide them with the key to future employment. The need to explore contemporary delivery methods is critical to the preparation of these students. Third, a fairly substantial number of educational insti- tutions still use batch mode computer image-generation. The introduction of microcomputer graphics systems to the instruc- tion of computer graphics at these institutions might prove more effective. The results of this study may affect deci- sions to update course offerings at other institutions. Limitations of the Study An obvious limitation that needs to be addressed is that a comparison was made between two generations of computer hardware. One might argue that the microcomputers have the advantage of several years of technological improvements and, therefore, a comparison between these systems severely limits the study. Actually, the study does not compare the hard- ware itself but rather the utilization of these equipment configurations to instruct computer graphics. Both modes are capable of being used for this type of instruction. The salient issue was how the batch mode (which represents present practice) compared to the microcomputer mode in terms of student achievement, student attitude, and cost. A comparison 0f computer graphics instruction in these terms will illus- trate the viability of these respective modes.In particular, 10 the aspect of cost assessment challenges a major justifica- tion for utilization of large main frame systems for instruc- tion. In the experiment, the population consisted of college students interested in computer graphics and was limited to those who chose to enroll in EGR 270 during Winter Term and Spring Term 1981. The course was only required for a few Engineering Arts majors (a degree program within the College of Engineering) in one of the four Option areas. The majority of the students who enrolled in the class chose it as an elective course. These students, therefore, had an interest in, or a personal need for, computer graphics. Thus, the two section sample of a pOpulation of students interested in com- puter graphics was biased toward the material and should not be considered average or typical University students. The number of students who could receive treatment in the microcomputer mode was limited by the fact that only one microcomputer graphics system was available. Thus it had to be determined how many students could be in the sample group. Sample size was determined based on personal experience of the researcher. From previous experience with other groups who had used this kind of equipment configuration, it was decided that up to a maximum of eight students were to inter- act with a single microcomputer each term. The second term the experiment was conducted, several students expected to be using the microcomputer. Students 1"] “L u: but I On: ’1‘. F.7- ll mutring the course had received information via word of mouflithat a microcomputer was being used in the course. Several students consequently dropped.the course because they vwre not able to be in the microcomputer group. It was felt that this was more of a limiting factor in the second term than in the first term. The use of different programming languages might appear to be a limitation of the study, however, it was not. BASIC, also called Dartmouth-Basic, was derived from FORTRAN (Kemeny and Kurtz, 1967). Those who used the microcomputer could only use the BASIC programming language. Since BASIC is essentially a subset of FORTRAN and since all students who took the course had knowledge of FORTRAN, then the programming language difference was not considered to be a factor in the study. Those who learned to program first in FORTRAN had little or no problem using BASIC. The majority of-the class who utilized the central computer facility programmed in FORTRAN and, in fact, were allowed to use any dialect of that language with which they felt comfortable. Hence, FORTRAN IV, Minnesota FORTRAN, and FORTRAN V were utilized. Again, these language differences were assumed not to have an ef- fect on the experiment. All students were present at class sessions where new material was presented. Graphics algorithms were presented graphically, that is to say, in flow chart form. Figure One is an example of a flow-charted graphics algorithm. Although it represents the logic for plotting a rectangle, 12 RECTANGLE SUBROUTINE x1, Y1, H, w COMMENT : This subroutine draws a rectangle whose height is H, and whose width is W. X1, Y1 will be the lower left corner. Figure l. 1 MOVE TO X1, Y1 l PLOT TO x1, Y1+H l [ PLOT TO x1+w, Y1+H l PLOT TO x1+w, Y1 I PLOT TO X1, Y1 RETURN Flow Chart Example. Rectangle Subroutine 13 understanding specifically what it represents is irrelevant to this study. Hence, the course was designed to present material independent of language. Again, since all students knew FORTRAN, when specific operations called for coded examples, FORTRAN was the default programming language. Also, the text used gave examples in both programming lan— guages and, for all practical purposes, was bi-lingual. The point to be reiterated here is that programming language was not a limiting factor. In fact, the use of graphical repre- sentations of computer logic to teach graphics seemed to be the most appropriate technique. A final word should be said about the instructor’s bias. After the initial assessments, or pretesting, students were divided into groups. Students quickly became aware of the experiment. When questions were asked about the experi- ment, they were answered honestly and openly. For the most part the students were enthusiastic and supportive. Indeed, it would be impossible to'deny that a sort of Hawthorne ef- fect was present (Campbell and Stanley, 1963). Also, the instructor's bias could have had a significant effect on the results, particularly in the achievement portion of the study. It should be mentioned that the grades students re— ceived on homework and programming assignments were based on objective criteria. The nature of the material sought from the students in the course did not allow for non—specific or subjective interpretation. Therefore, biased grading could be reduced or eliminated. Although grades were assigned to 14 all students by the researcher, a strict format was utilized to minimize inequities in the achievement measure. It will be up to the reader of this study to determine the degree to idfich this limits the credibility of this research. Overview of Subsequent Chapters Chapter 11 contains a review of literature intended to provide a background for those interested in the use of computers in education, computer graphics, and the more re- cent developments and applications of microcomputers in edu— cation. Chapter III consists of a formal statement of re- search questions which provided the primary thrust of this research project. Derived from these questions are the re- search hypotheses. Finding the answers to these hypothetical questions, in effect, provided the means to address or answer the research questions. Also the definitions of important terms used in this study may be found in this chapter. Chapter IV is divided into two parts: the design of the study and the procedures used in the study. The experi— mental approach was explained under Design Over Time and Design Over Variables. The design matrix was explained. Statistical models and techniques of analysis and validity and reliability concerns were also addressed. The procedures section elaborates upon the details of the study. The en- rollment history and rationale for sampling were discussed under POpulation and Sample. A unique approach used to as- sign subjects to the treatment groups is discussed in the 15 Treatment Subsection. Instrumentation and data collection was divided into three parts: achievement measure, atti- tude survey, and formation of the cost models. Chapter V contains the results obtained by the study. The results of sampling are provided. The issue of com— bining data from the two terms during which the experiment was conducted was addressed. For each dependent variable the research hypothesis was restated and tranSlated into statis- tical form, statistical analyses and decisions were given. The cost models were exercised by manipulating various para- meters and results were presented in graphical form. Chap— ter VI presents a summary review of the study with a discus- sion of the conclusions drawn and implications for future research. CHAPTER II REVIEW OF LITERATURE Overview In reviewing the scholarly literature for the present study, three topical areas related most meaningfully to the research variables of interest. Therefore, in discussing the literature, the review will be arranged in three sec- tions: 1) teaching with computers, 2) computer graphics in- structional materials, and 3) the use of microcomputers in education. Section one addresses the issues of teaching with com- puters-rather than teaching about computers. Most of the computing activities in higher education have been involved either with teaching computing or using computers as an aid to research. It is necessary to examine these traditional roles. Section two focuses on computer graphics in an attempt to clarify what it is, its importance, and instructional as- pects in terms of the literature available as general texts and references. Section three reviews the most current literature which is almost exclusively taken from microcom— PUting periodicals concerning the use of microcomputers or Personal-sized computers in education. 16 L_ 11 17 Section One Computers in Education There are several good reviews of literature in the area of computers in education. Three recommended by leeachie in Teaching Ti s are: Hunter, B., et al, 1975; Levien, R. E., 1972; and Rockart, J. F., and Morton, M.S.S., 1975 (McKeachie, 1978). A fourth, and by far the most con— temporary review of material in this area, is a work entitled Computers in the Teaching Process (Rushby, 1979). Following Rushby's outline, the aspects of teaching with computers were examined more logically. If one begins with the assumption that those who teach and use computers should not need to know more about them than any other machine that they use to facilitate education, viz., an overhead projector, then one begins with the per- spective that most educators have who use the computer. Computers in Education and Training Research and teaching together account for most of the computing in higher education. Computers have been used both for education and training and these two applications are different. In education the student benefits, in train- ing the organization benefits. Differences in approach are seen in methods used to assess the student's performance and PIOgress. Training uses criterion-referenced testing. In higher education norm-referenced techniques are used. Hence 18 assessment has a qualitative flavor, i.e., the degree of ex- cellence achieved and how well students perform relative to their peers. Computer Assisted Learning (CAL) and Computer Managed Learning (CML) are two traditional approaches to the use of the computer in education. In CAL the learning material is presented to the student through the computer, while in CML the computer is used to direct the student from one part of the course to another and learning materials themselves are not kept in the machine. At this point in time, it is dif- ficult to advance the use of educational technology until a better understanding of the learning process is developed (Rushby, 1979). The difference in application of computers to the learning process can now be described in terms of the way in which they mediate the flow of information and the levels of detail with which they are concerned. There is a changing realization that technology properly applied offers the possibility of improving the quality of learning. Thus, the cost effectiveness of education and training is improved due to greater effectiveness in learning rather than by a reduction in costs. The original hope that educational com- Puting could provide improved, individualized education at a low cost was unrealistic and this has led to a widespread disenchantment with technology. Educators now recognize that the computer is but one of the tools that can help with teach- ing and learning and understanding the problems therein. L p . 1 19 Further information and discussion of these and other aspects of computers in education and training can be found in the following publications: Carnegie Commission, 1972; Levien, R.E., 1972; Oettinger, A.G. with Marks, 8., 1969; and Rushby, N.J., 1979. Computer Assisted Learning According to Rushby, Computer Assisted Learning (CAL) is a general term most commonly used in the United Kingdom for teaching with the aid of the computer. Computer Aided Instruction and Computer Assisted Instruction (CAI) also mean teaching with the aid of the computer and are used ex— tensively in the United States. They are but two of a num- ber of synonyms for CAL but with a different connotation in Europe, and can imply tutorial or instructional CAL which is less general. This report follows Rushby's outline of the utilization of computers in education, therefore, it will be consistent to keep with this terminology. Computer Assisted Learning (CAL) may be perceived as Optimizing the student's performance towards specific goals. It is the use of technology in education to center the course On the subject material, or it can be thought of as a way of optimizing the instruction towards the student's own goals. PUt yet another way it is the use of technology in educatiOn to center the course firmly on the student. Before any further discussion on using CAL, it is neces- sary to examine four educational paradigms useful in relating 20 CAL to the general field of education: instructional, re- vealatory, conjectural, and emancipatory (Kemmis, 8., and Wright, B., 1977). Instructional CAL. A form of CAL akin to programmed learning in which the student is led through the learning material via a structured question and answer dialogue. The focus of the instruction is on the subject material and on the student's mastery of various concepts within it. Its main disadvantage is that it is a dialogue restricted between computer and student. Revealatory CAL. A form of CAL in which the user is guided through a process of discovery so that the subject matter and the underlying theory are progressively revealed to him as he proceeds through the CAL package. The computer acts as a mediator between student and a hidden model of some real-life situation. The main disadvantage is that care must be taken so as not to oversimplify reality so that the stu- dent can outguess the package. Conjectural CAL. The use of the computer to assist the student in the manipulation and testing of ideas and hypoth~ eses. It is based on the concept that knowledge can be cre- ated through a student's experiences; its emphasis is on the student's exploration of information on a particular tOpic. Emancipatory CAL. The use of a computer as a means of reducing the student's workload, for example, as an aid for numerical calculations or for information retrieval. l.—' 21 It is important to note at the beginning of this dis- cussion of using CAL that computer assisted learning is but one medium for the presentation of information to students. Others are: printed material, slide-tape, video, lectures, tutors, and instructors. Using the computer or any other media should be based on a decision which addresses the is- sues of appropriateneSs of use, financial constraints, and the need for variety. Each has its advantages and disadvan— tages in these respects. When it has been determined that computers are an ap— propriate medium the learning package should be considered. Should computer terminals be the sole device or should other media be included? These additional materials can be called courseware and, together with the computer hardware and soft- ware, form a triad or what can be described as 3 CAL program. The production of CAL packages (or programs) starts with a Specification of educational or training objectives of the whole package and is followed by the overall design. In pre- senting the GAL package the utilization of computer graphics should be considered. Computer produced graphics widens the possibilities for presenting complex information and improves the attractiveness and the quality of the teaching (although not necessarily the quality of learning) at a slightly greater cost. Another consideration in producing CAL packages is the choice of delivery mode, e.g., interactive versus batch com- puting. Interactive terminals provide an immediacy of access 22 and rapid feedback. However, it may not necessarily be the cheapest way of communicating with the computer. Batch mode processing should always be considered as a viable alterna- tive. The disadvantage is that output delays can be several minutes to several days. However, it is the least expensive ferm of processing information. Two other possibilities for increasing the efficiency of using computers in CAL packages are: l) grouping several students together to interact with a single terminal, and 2) utilization of microcomputer sys- tems . The significant difference between CAL and other media is the ability of the CAL package to break out of a linear sequence of material to recap or offer an alternative ap— proach. It allows the student to change something in his model or simulation and so to respond to his individual preferences. Usually CAL materials are put together by a production team, persons who Specialize in authoring these materials. However, teachers can produce their own material as can their students. Thus, CAL in itself can be an educa- tional experience. Further information about Computer Assisted Learning can be obtained from these sources: Ellis, A.B., 1974; H00per, R., and Toye, I., 1975; Kemmis, S., Atkin, R., and Wright, B., 1977, and Lecarnie, 0., and Lewis, R. (eds.), 1975. More information about using Computer Assisted Learn~ ing: Dyer, C.A., 1972; McDonald, B., Atkin, R., Jenkins, D., and Kemmis, S., 1977. £01 31' 23 Computer Managed Learning Computer Managed Instruction (CMI) or Computer Managed Learning (CML) entails the use of the computer to assist the teacher, instructor, and administrator with the routine management tasks in teaching and learning, such as assess- ment, guidance, record keeping, and reporting (Rushby, 1979). Computer Managed Learning (CML) puts the computer in a background or supportive role. In so doing it frees teachers of their administrative burdon thus allowing them to devote more time to their students. The CML cycle consists of four broad areas in which the computer provides management sup— port. One, construct, mark, and analyze tests for diagnostic or assessment purposes. Two, keep records of students' per- formance and progress. Three, provide guidance for each stu- dent, directing or advising their choice of route through instructional course materials. And four, from individual student records, report on the performance and progress of the student and of the course to either the student, instruc- tor, or manager of the educational or training institution. Many CML activities involve the use of modules or modularized instruction. The following discussion paraphrases Rushby's chapter on CML. The heart of most CML systems is assessment. Computer .Managed Learning endeavors to perform the mundane activities of test generation, test scoring or marking, feedback, and analysis. CML is particularly appealing to instruction '7' 24 involving large numbers of students where the material is fairly objective. It is worthwhile to make mention of some of the activities related to CML. First, multiple choice questions are generally used for they are amenable to automatic marking techniques. Secondly, computer aided marking techniques, e.g., optical mark reader, can be utilized to transform student responses into data bases for computer evaluation. Third, feedback in terms of test results can be turned into learning situations, perform— ance on one exam can suggest remedial work or influence the makeup of the next examination. Fourth, reliability and validity of tests can be statistically analyzed by the com- puter. A test administered to a group of students yields information not only about the students, but also about the course and test itself. These procedures can be built into the computer. Tests as a whole can be analyzed. ~Distribu- tions of scores can point out the good as well as the bad students. Both norm-referenced and criterion-referenced evaluation techniques can be utilized. Individual questions can be analyzed statistically employing measures of facility value and a discrimination index, again built into the com- puter. Finally, tests can be computer constructed via an item bank, a pool of questions, from which the computer can generate any number of "individualized" exams based on any number of criteria or feedback. In order to discuss CML completely, the three issues 0f routing, record keeping and reporting must be reckoned 25 with. Routing is the concept of using the computer to guide students through structured course materials. It is usually based on the assumption that the majority of students should follow one of a smaller number of paths. Educational psy- chologists point out that this is impossible since it is not understood completely how students learn (Ellis, 1974). Modules work where material can be highly structured, such as in math and science, and success in one module leads directly to the comprehension and success in mastery of the next. Routing rules can be made as complex as the system will allow, but these rules are still specified by the in- structor and need the versatility of being structured in a variety of creative ways. Record keeping is needed in CML to guide students but the maintenance of the records brings up the issue of pri— vacy. Confidentiality of records is not a technological issue but rather is sociological in nature. These types of problems can be minimized if these principles of privacy are adhered to: one, records should not be kept secret, par- ticularly from the students. Two, students should be able to challenge data and make corrections if the data contains errors. And three, students must be assured that the infor- mation held about them will be used only for educational Purposes and will not otherwise be used without their per- mission. Reporting involves maximization of communication be- tween those who offer the course and those who take it. Up 26 to date reporting is essential to manage teaching or train- ing effectively. The instructor needs information about the individual student's performance and progress or the lack thereof. One of the most alarming consequences of CML is the frustration students can experience when dealing with a com- puter when things are going badly. Procedures must be avail- able when problems occur and questions arise (see page 52). Some final thoughts about CML concern the possible ex— tensions of CML systems into other aspects of education. Computer Assisted Learning packages which can be used in a CML course could provide information directly into the CML records. It is possible to link computer based library cir- culation systems with CML. Student record systems and ca- reer counseling Operations could also be linked to CML. And if one wanted to go to the extreme, it is possible to imagine that a student's whole educational experience could be managed totally by the computer. Further information about Computer Managed Learning can be obtained from these sources: Anderson, Kulhavey, and Andry, 1971; Baath and Mansson, 1977; Byrne, 1975; Crocker, 1974; Guilford, 1973; McMahon, 1978; and Murray, 1975. Informatics and Education Informatics is the study of information, the way we manipulate and use it, and its impact on society (Rushby, 1979). 27 Currently there exists a misconception about who should use computers. Some feel that only those who understand how computers work or know how to program ought to use the com- puter. This is of course false. According to Nelson, those who know about computers constitute a kind of priesthood and they are likely to put things over on people who don't know about computers. This putting it over, as it were, has been labeled "cybercrud" and takes numerous forms. Firstly, the "computer is a magic word" and the "computer as a magic in- gredient," i.e., computer studies. "The truth is that the mere fact the computer is involved is something that has no bearing on its character or validity." Secondly, ”white lies: the computer made me do it." What this usually means is that the computer becomes a scapegoat and whatever has gone wrong is the computer's fault and not that of the bad software or program that one has written or is using. Lastly, "Yagottas: the computer as a coercer." The ultimate use of cybercrud is for making people do what you want. ”PeOple can be made to hand over personal information, secretaries can be intimidated into scouring the files and payment sched- ules can be artificially enforced.” Nelson has coined the word cybercrud to make others aware that those who use the computer can "confuse, intimidate or pressure" those who don't. Computer people do deserve respect and it is under- standable that they want to be appreciated. However, "no man has a right to be proud that he is preserving and manipulat- ing the ignorance of others" (Nelson, 1974). 28 Many people have proclaimed that computers are a panacea fer the ills of education. This can now be labeled educa- tional cybercrud. Realistically, the effectiveness of the use of computers in education and training depends on the educational context in which they are used rather than on the teacher having an empathy for the computer or knowing how to work it. However, this is not to say that there is no relationship between the study of computers and the use of computers in education and training. What needs to be empha- sized in order to more intelligently promote the greater use of computers in education and training are the concepts of computer awareness and computer literacy. The concept of computer awareness is simply an under- standing of what computers are and what they are capable of doing.. For those in education and training interested in using computers there are two notions to be cognizant of: 1) knowledge about computers is not dependent upon mathematics and physics, and 2) a comprehension of what is happening in- ternal to the computer is not mandatory. Computer awareness stated simply is: a computer is a general purpose machine. The concept of computer literacy has become much touted in education and training arenas. A definition of the con- cept follows: Computer literacy refers to a knowledge of the non- technical and low-technical aspects of the capabilities and limitations of computers, and of the social, voca- tional, and educational implications of computers. (Moursand, 1976) 29 Moursand was one of the first to develop a course at the University level on computer concepts or computer liter- acy. Briefly, it was a no prerequisite, low level, intro- ductory computer science course whose goal was to raise the level of a student's computer literacy. Its content was approximately one-third programming and two-thirds non- programming material. A well balanced course, he states, has the three overlapping areas: computer capabilities and applications, computer limitations and social, vocational and educational implications which are built upon the foundation of computer programming, and computer usage with hands on experience. Moursand has also written a textbook entitled: Basic Programming for Computer Literacy (Moursand, 1977) as well as a quiz for assessing computer literacy (Moursand, 1976). This sub section began with a definition of informatics, it will conclude with a summary on how it should impact stu- dents. Through informatics, students should be provided with sufficient knowledge Of computers so that they can make general and social inferences. In minimum terms, students should be provided with a chance for learning computer aware— ness. Ideally, all students should be given the opportunity to attain computer literacy. The greatest advantage com- puters can provide is their ability to carry out instructions precisely. Informatics and Computer Aided Learning tech— niques can be used to teach students the systematic approach to problem solving. Thus there is the potential to better 30 educate students. In a society where informatics is playing an increasingly larger role, the appropriate use of tech- nology ought to be based on the rudamentary knowledge of what computers can and cannot do. Section Two Computer Graphics Instruction Literature about computer graphics is for the most part divided into two major categories: one, hardware and software orientation, and two, applications orientation. Hardware is the physical equipment or components that make up computers and peripheral devices, including devices with graphics capabilities. Software is the instructions or pro— grams that underlie the capabilities of the hardware, i.e., allow it to do or to be something. The applications areas consist of the variety of uses of computer graphics. It was stated earlier that computer graphics was not an end in it- self but rather a general tool that can be utilized in a variety of ways. An attempt to classify these areas presents the following list of applications orientations: general man-machine communication, miscellaneous (areas and tech— niques), architecture, art and textile design, computer aided design and engineering, and management information systems (Machover, 1978). Although Machover's document is a relatively complete and recent collection of source information, it oddly lacks any mention of computer graphics education or educational 31 materials. One might say that this falls into the miscel— laneous category. Several possible explanations for this are in order: one, a possible reason for the lack of these materials is that computer graphics is new. Remember that the first computer graphics system was not developed until 1963 (Coons 1967). Two, the generation of imagery is highly dependent upon the hardware. Even at the time of this writ- ing, no two systems utilize the same instructions for image generation. Although attempts have been made at standardiza- tion, standards have yet to be adopted or accepted (ACM/ SIGGRAPH, 1978). Three, the computer graphics that have been taught at the University level until recently have been to upper division engineering and computer science majors and usually have been oriented towards advanced hardware/software considerations. Four, accessibility has been limited. Until the advent of fourth generation equipment, microcomputers, very few individuals even had access to devices with graphics capabilities. Lastly, demand for instruction has been mini- mal and is just now starting to be heard. The popularization of computer graphics via the entertainment media has created an awareness and an interest in computer based imagery. Films like "Star Wars" and others with futuristic themes have employed computer graphics sequences. Computer animation is beginning to compete with hand animation. Computer art has become recognized as a valid form of individual expression. Only now that a demand has been created for computer graphics are instructional materials beginning to appear. 32 The remainder of this section focuses upon computer graphics educational materials. In this discussion, the following terms will be used, therefore, distinctions must be made between computer graphics and interactive graphics. Also, the major application areas of these Operations vis-a- vis computer aided design and computer aided manufacturing. Computer Graphics (C.G.). The use of a computer to define, store, manipulate,’interrogate and present pic- torial output, essentially a passive operation. Interactive Graphics (I.G.). The use of a computer to prepare and present pictorial material. However, in I.G. the Observer can influence the picture as it is be- ing presented, an active operation. Computer Aided Design (CAD). Any use of the computer to aid inIthe design Of any—individual part, a subsystem, or a total system (does not have to involve graphics). Computer Aided Manufacturing (CAM). The use of a com- puter to are in the manufacture or production of a part exclusive of the design process. (Rogers and Adams, 1976) Computer graphics educational materials can be further divided into two categories based on their relative complex- ity and educational intent. Again, the focus is on the in- struction of computer graphics, interactive graphics, com- puter aided design, and computer aided manufacturing. Suc- cinctly, these divisions are: general and introductory ma- terials and engineering and technical materials. aneral and Introductory Materials Material in this category can be found in several forms. Firstly, occasional articles in magazines or journals. These articles describe programming logic and give listings of 33 specific graphics algorithms (Kolomyjec, 1977). Two, a series of articles describing graphics technique, such as three-dimensional rotation (Truckenbrod, 1980-1981). Lastly, whole issues devoted to computer graphics articles contain- ing a variety of algorithms and discussions on a variety of equipment, for example, the January 1981, July 1980, and June 1980 issues of Creative Computing magazine. These types of articles are usually aimed at the microcomputer hobbyists and are somewhat machine dependent but are almost always written in BASIC or FORTRAN and can usually be modified for use on most popular personal computers. Besides occasional articles introductory course ma- terials would fall into a separate category. To date there are only two texts published suitable for use in an intro- ductory course in computer graphics. The first, Computer Graphics (Demel, et a1, 1979) is a stepping stone between a student's first courses in computer science and technical drawing and higher level applications courses. The aim of the text follows: This book was written with the express purpose of devel- oping an individual's confidence in his ability to Wilte programs, to gain knowledge of computer graphics sys Ems, and to become familiar with the uses of computer grap ics . all parts of the book combine to give the beginning students a solid basis for pursuing advanced coursesdipO COmPUter graphics or for going into industrylprippge 1) use computer graphics systems. (Demel, et a , , p. The text itself is divided into two parts, two dimen- Sional graphics and three dimensional graphics. The f1r5t . C 0 .se part includes oveIViews of graphical systems, a conci T____f history of computer graphics, the anatomy of a computer - a brief discussion of computer architecture, as well as a dis- cussion of computer software. These areas provide a review and a foundation for the two dimensional programming exer- cises that follow. It is interesting to note that the empha- sis is on writing subroutines for each graphics Operation - a structured approach. The second half of the text deals with three dimensional graphics. It begins with the construction of a three dimen- sional data base which is utilized by the algorithms that follow. The graphical algorithms presented allow for ortho— graphic view drawing and pictorial drawing. The pictorial algorithms enable students to generate oblique, isometric, axonometric, and perspective drawings from a single data base. The text concludes with an interesting discussion of graphics system components andgraphicsapplications. Also, a glossary of terms and an extensive bibliography provide additional information to the students. It is noteworthy to mention that this text was the required book for the course upon which this study was based. A second text that would fall into this category is a self-teaching guide entitled Graphic Software for Microcom- puters (Korites, 1981). This book is noteworthy because it presents complete programs for a particular type of micro- computer having graphics capabilities (an Apple II plus, 48 K computer system). It utilizes vector mathematics and matri- ces and thus is more difficult to use than the previous text. 35 However, this type of approach is more suited to intermediate and advanced applications. This book begins with simple two dimensional plotting programs and a discussion of interactive techniques. Next are programs which show how to translate, rotate, scale, and clip two dimensional drawings. Rotations are expressed in matrix form. Three dimensional rotations are presented. Also included are programs and techniques for shading, hidden line removal, and perspective transformations. Engineering and Technical Materials The majority of formal literature available about com— puter graphics falls into this category. Since the emphasis of this study is computer graphics instruction, the focus was narrowed to cover the materials in this category that can be used as references or background material. ‘Anyone interested in more than a general overview or introduction to computer graphics would be well advised to begin further investigation with any or all works that will be highlighted in this section. A cursory review of five of the best works on the subject is appropriate. Each book has an extensive bibliography of its own. Each book has the potential to be required reading in an upper division college level engineer— ing or computer science course in computer graphics. .Most require some programming experience, knowledge of machine organization and/or machine language, a familiarity with data structures, as well as at least one year of college T—— 36 mathematics. Thus, these works are not recommended for those who do not have a math/science background. Principles of Interactive Computer Graphics (Newman and Sproull, 1979). This is the second edition of the 1973 clas- sic first edition. Prerequisites for use of this text would be some programming experience and an ability to be conversant with machine organization and data structures. The text is divided into six parts: basic concepts, 3 general introduc- tion to computer graphics as a whole, and an explanation of basic techniques, such as clipping, geometric transformations and incremental models; graphics packages; how to build a package of graphics functions or subroutines to support the writing of application programs; interactive graphics, tech- niques and devices that allow graphical interaction with com- puters; raster graphics, a different approach to computer graphics that uses TV-based raster displays; three dimensional graphics, the display and modeling of three dimensional ob- jects, also, removal of hidden surfaces, shading, the model- ing and display of curved surfaces; and advanced t0pics, dis— play processors, device independent graphics systems, and the design of user interfaces. Mathematical Elements for Comppter Graphics (Rogers and Adams, 1976). The text presents a mathematical approach to computer graphics techniques. Basically it organizes mathe- matical concepts that are useful for computer graphics appli- cations. One year of college mathematics is recommended prior to use of this text. 37 The material presented consists Of a discussion Of cur- rent computer graphics technology followed by the manipula- tion of graphical elements represented in matrix form using homogeneous coordinates; a discussion of existing techniques for representing points, lines, curves, and surfaces within a digital computer as well as computer software procedures for manipulating and displaying computer output in digital form; mathematical techniques for producing axonometric and perspective views along with generalized techniques for rota- tion, translation and scaling of geometric figures; and curve definitions for both explicit and parametric representations for both two and three dimensional curves. Curve definition techniques include the use of conic sections, circular arc interpolation, cubic spline, parabolic bending, Bezier curves, and curves based on B-splines. Geometric Principles and Procedures for Computer Graphic Applications (Chasen, 1978). This text, which covers and ex- pands upon many of the concepts set forth by Rogers and Adams, brings together the variety of geometric ideas which underlie much of computer graphics. It is written for two audiences: the systems programmer, and the user of graphics systems. The stated intent is to provide those concepts which will serve as a comprehensive instructional base and a useful ref- erence in courses in applied graphics and mathematics. Dis- cussion is broken down into three logical divisions: display- ing existing equations or math forms, creating a mathematical foundation to satisfy known or desired constraints, and three dimensional geometry. ”7—— 38 Computational GeometrX for Design and Manufacture Faux and Pratt, 1979). Computational geometry is the com- puter representation, analysis and synthesis of shape infor- mation. This text is slanted toward the application of com— puter graphics technique in Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM). It uses a geometrical rather than an analytical point of view. The work is fairly self-contained for the reader who is familiar with elemen- tary calculus to the level taught in most undergraduate engineering courses. Many types of splines as well as both Coons and Bezier patches are discussed. A comparison of surface fitting systems versus surface design systems is made relative to geometric and computational concerns. Interactive Computer Graphics: Data Structures, Al— gorithms, Languagps (Giloi, 1978). A fairly esoteric pre- sentation of computer graphics applications written for the graphics application program programmer, graphics systems programmer, graphics systems designer and CAD applications programmer. The emphasis of this book is on data structures suitable for computer graphics, or algorithms for picture generation and transformation and on the appropriate lan- guage constraints, in the widest sense, for the generation of graphics Objects. ‘ 39 Section Three The Use of Microcomputers in Education The use of microcomputers or personal computers in education is so new and decentralized that very little for- mal literature addressing its utilization in the educational setting is available. However, there has been a fair amount of informal rhetoric or articles on the use of microcomputers in education and these have appeared mostly in microcomputing periodicals. In a recently published listing of sources of information for those interested in educational applications of microcomputers, these sources were categorized as follows: magazines and newsletters, organizations, university educa- tors, and others involved in microcomputing in education, educational software and articles (LOpez, 1981). A review of many of the articles available on the sub- ject made it possible to identify several categories into which these types of articles might roughly fit. These coursely defined categories are: teachers' attitudes toward microcomputers, microcomputers in schools or computers in the classrooms, microcomputers and learning, microcomputers versus time sharing and general information about microcom- puter applications. The remainder of this section will elab- crate upon these categories with a discussion of ideas pre- sented in representative articles. It is also possible that an article could fit into more than one category, therefore, 4O placement into one category was purely subjective as was the category designation itself. Teachers' Attitudes Toward f Microcomputing Articles in this category generally reveal the atti- tudes educators presently have toward the utilization of the microcomputer. The first article dealt with a survey of 189 pre-service and in—service educators done in 1976 (Lichtman, 1979). Some general observations were that the educators seemed less enthusiastic about the computer's role in educa— tion than did the general public, those in administrative courses were more positive than other educators, and the "teachers" (those in education courses minus the administra— tors) showed the lowest percent in the feeling that computers will improve education. Lichtman's conclusions (which he qualifies with a statement to the effect that his study does not purport to represent an in—depth analysis) are fairly interesting and are general in the sense that they represent attitudes to— ward all computers, including microcomputers. One, teachers view computers in a more dehumanizing and isolating manner than do other segments of the population, especially school administrators. Two, teachers do not feel secure in their relationship with computers. Three, both teachers and ad- ministrators are more wary of computers in relation to jobs and skills than other people. Four, teachers see little im— provement in the quality of life through the use of computers, Ffi 41 administrators feel the Opposite. Lastly, a few teachers felt threatened by computers, i.e., that their jobs would be taken away from them. If this study is indicative, states the author, it may be of the possibility that a lot of com- ‘puters may be bought by administrators but they may not be used by the teachers, at least not to any large extent. In another article, Knight makes the statement that teachers' attitudes toward the computer are seen as an over- all resistance to technology. Educators are beginning to recognize the existence of math anxiety and math avoidance; "computer anxiety" has brought this into focus. Educational institutions have been delinquent in preparing themselves for dealing with this technological society. The way teachers must overcome their own anxieties is to spend time learning about computers and their capabilities. Using a microcom- puter, the author conducted workshops for teachers to help them overcome their ignorance and fear through lectures, simple programmed learning sequences, demonstrations, simula- tions, and game playing. By combining machine based activi— ties with non-machine based activities, the transition toward use of the computer was made easier (Knight, 1979). A more practical attitude toward the computer in the classroom is to use the computer as an aid rather than a re- placement for the classroom teacher (Spero, 1979). The author of this article shares his experiences with teachers using a TRS-80 level II system, a popular microcomputer. Computer Aided Instruction (Computer Assisted Learning) was 42 not discouraged, neither was it encouraged. Teachers can use microcomputers more cost effectively as well as peda- gogically effectively in their interaCtions with students. Training teachers to work with microcomputers outside of learning packages provides the best interaction between teachers, students, and equipment. The author obtained several grants to tutor teachers and found that there are many good materials (software) available that can aid in teacher training. In another article in this category, Martellaro states that in the fifteen years that CAI has been around, it has not been used much until recently. Computer Aided Instruc- tion has moved from larger machines to microcomputers. Why haven't more school children been exposed to computers? Generally the school systems themselves are resistant to change. There are three categories of teachers based on their attitudes toward computers: one, those that believe computers dehumanize education and should be kept out of schools; two, those that believe computers have potential but they are frightened of them; and three, those who believe in computers and want them in the classroom. Most teachers fit into the second category. There are attributes that affect the rate of adoption of any innovation, these attributes delineate barriers that must be overcome. Computers will have to have a perceived advantage over regular instruction. Their perceived complexity will have to be overcome. The ability to experiment with the computer on a limited basis 43 will have to be increased. The observability of successful utilization of computers in the classroom must be made visi- ble to those who don't use the computer. Finally the values, experiences, and needs of the teachers and students will have to slowly alter, making computers an integral part of learn- ing and life (Martellaro, 1980). The final representative article of teachers' attitudes toward computers and microcomputers by Koethe suggests that there is a mob attitude for a "back to basics" movement in education and teachers should not misinterpret its meaning. Back to basics does not mean a return to the 1940‘s. Rather, in the Information Revolution in which we now live, under- standing the computer is an essential key to survival. The "back to basics” desire expressed by the public must be taken as a responsibility by educators to provide students with the basic skills required in our increasingly complex technolOgi- cal society - to provide the skills of today using the most instructional tools of today. One of these tools is the microcomputer. The microcomputer represents the majority of the facets of computer supported instruction. Certain appli- cations of microcomputers are more well suited to "back to basics" principles, i.e., drill and practice. Microcomputers can also advance students from one skill level to another after a proficiency level has been obtained. After satisfy- ing public demands in the areas of the basics, microcomputers can be used for more loftier educational goals. Most educa- tors are in an unprecedented situation: the microcomputer is 44 the most important educational tool available, computer lit— eracy should be an essential subject in elementary and sec- ondary curricula, yet microcomputers did not even exist when most educators and teachers were in school. The microcom- puter has the potential to satisfy the public's cry for "back to basics" (Koethe, 1980). Microcomputers in Schools An article by Zinn of particular interest to this study primarily because it involves the use of microcomputers in higher education, is divided into two parts: one, descrip- tions of a single user, personalized system at the University of Michigan and two, the implications of the new technologies for improving access to higher education. Zinn states that microcomputers have evolved in a natural fashion from mini- computers. They are being used by students as word process- ing systems for text processing of their reports and in lab- oratories for instrumentation in their experiments. Train- ing has been advanced in significant ways by the introduc— tion Of microcomputers and they have been particularly use- ful in the following educational applications: graphics in engineering, linear programming in business, information systems in law, class record keeping, etc. Current planning at the University of Michigan suggests an important future role for microcomputers and personal computers in teaching and learning activities. Zinn mentions several points that indicate these implications. One, extr not puti rese 1,0l time in : bil the do per and abl Stu toc scl re- 3P1 te: Sp to an Th Ch Sc 45 extrapolation from typical computing in higher education is not possible. A revolution is going on that will put com- puting into everyone's home. Microcomputers are making CAI research obsolete. Higher education will enjoy the use of 1,000 more personal computers than the current number of time sharing terminals. Personal control will be a factor in increased use in education. Two, anticipated future capa- bilities and discontinuities are important. What will be the impact of new technology on education? Social implica- tions for planners to consider must include the impact that personal computers will have on learners, teachers, scholars, and the community. Three, as computing becomes more avail— able and personalized, it becomes increasingly useful to the student as a scholar. Using the computer as a scholarly tool, the student moves more easily into a community of scholars and learners. Dramatically lower costs lead to a re-thinking of what is useful to do with computers. Many applications are self-justifying. Lastly, the implications of personal computers for higher education will be dramatic even to the extent of has- tening the demise of some institutions. A shift in the re- sponsibility of learning will come about, courses will be- come more general and students will find considerable assist— ance in computer processing and improved learning skills. Thus, the traditional roles of the professor and student will change. Also, adult and continuing education will beaffected. Some institutions will not survive these changes (Zinn,l978). Bar: witi put put adv val use in fer the The 101 int scl pec 6111 SN co: to: Dr. Sh' t0 ha th pf 46 A contrasting article by the elementary school teacher Barstow warns that some underlying questions must be dealt with concerning values. TOO many people in educational com- puting accept too easily the assumption that any use of com- puters will improve schools. Every educational tool has both advantages and disadvantages. There is a need to clarify values and Opinions and better understand the effects of the use of computers on children. Computers should be utilized in a unique manner and not merely do the same work in a dif- ferent way. Computers are part of a wider issue concerning the use of technological devices of every kind in education. The issue precisely stated is: do computers or other tech— nological devices expand or do they complicate the lives of individuals? Realistically, computers will have a minor impact on schools, joining the ranks of available teaching aids. Some people will use them extensively, others will prefer differ- ent methods. Schools have a responsibility to help prepare students to deal with computerization. There will be serious consequences if the educational potential of this versatile tool is left unexplored (Barstow, 1979). Hakansson and Roach tell of an interesting approach to providing schools with computer classrooms. The Science Shuttle, a van with a dozen microcomputers and two instruc- tors, can turn any space into a microcomputer facility in a half hour. Taking advantage of the extreme portability of this type of equipment, the computer classroom can service up to I Law: cla: clai put: One den‘ The and 6nd tea: put. men“ the Ame“ lit. tOI tha‘ thi: r01. Cat. Stu. Yer 47 to 120 students per day. The project is supported by the Lawrence Hall of Science (Berkley, California) who, they claim, have pioneered hands on computer education. This claim is substantiated by an annual paid enrollment in com— puter activities of 40,000. The Science Shuttle allows two goals to be realized. One, to make computing accessible to large numbers of stu- dents, and two, to bring computers to the average classroom. The project is a success despite cutbacks in state funding and limited space at the Lawrence Hall of Science. Students enjoy the experience, teachers become excited by the new teaching method and administrators have a chance to see com- puters in the classroom before making a large capital invest- ment (Hakansson and Roach, 1979). A final article in this category by Billings suggests that computers have not realized their full potential in American education. In an information society, a computer literate society is important and there may be many a crisis to overcome as it deveIOps. The 1980's must be the decade of computer education. Computers have become so prolific that educators cannot ignore them any more. Teachers do not think they will be replaced by the computers but expect their role with students will change as they each learn to communi‘ cate with the machines. Computers are "mind multipliers" to students, via their interactivity, open-mindedness, and versatility in presenting text, sound, and graphics. 01 CC el p1 pe ca tl. tl. an in th de ti in th di mi 48 To use computers successfully, teachers must know what they want the machine to do in the classroom. Teachers need information on the development of microcomputer use in schools and on available educational software in order to make their own decisions. In postsecondary education, utilization of computers is at a higher level of SOphistication, whereas at elementary and secondary levels, the use lies more with drill and practice. Two obvious reasons are age of the students and varied nature of available computational equipment. The use of microcomputers is hindered by a lack of ap- propriate educational software and information on what other people are doing in education. What is needed is a clarifi- cation of roles. Those outside of education have to be shown that computers can do a better job in, and not just replicate, the learning process. Nevertheless, more and more educators are using microcomputers and are doing a variety of interest- ing things with them. An important issue that educators must address is that of control. Should students be in control of the computer or should the computer be in control of the stu- dent? Educators may be able to bring about tentative solu- tions to some of these issues by a show of support to soft— ware houses. If educators can, with a common voice, support the efforts of those who produce quantity software, more quality educational software might be produced. Also, new discoveries have yet to reveal the full potential of the microcomputer (Billings, 1980). 49 Microcomputers and Learning The first article in this category was written by Banet in which he discusses computers and early learning and makes the prediction that the typical instructional computing con- figuration in schools will evolve toward microcomputer-based, stand-alone systems with a variety of input/output modalities. Knowledge of a programming language will not be neces- sary for young children to interact with the computer. Com— puters in education will replace paper not peOple. Gamelike situations with immediate feedback hold the most promise for the young. Basic skills and concepts will be easily mastered through electronic systems. Thus, more time can be spent on more stimulating concrete experiences. Tutorial systems should be designed with features that give the students Op- tions of proceeding in certain directions rather than in pre- determined sequences. Computers will provide a means for representing complex processes in order to help students discover relationships and isolate variables. Computers will be used to construct self-assessment quizzes for elementary school children just as item pools are sampled by computers at the university level. In order to promote more utilization of microcomputers in schools we need: 1) good research and development among groups seeking to integrate computer systems and student- initiated learning, 2) to evaluate the utilization of com- puter systems in learning settings other than the classroom, 50 and 3) to establish information networks to keep one another informed about computer applications being developed all over the world (Banet, 1978). According to Dwyer and Critchfield, the significance of the home microcomputer for informal learning is that it provides a whole new range of interesting things for young people to create with the help of their parents. Programming a microcomputer can facilitate learning outside the school and bring a child and his parents closer together. Dwyer and Critchfield have been developing and teaching workshOps in personal computing. The syllabus is broad but particularly relevant to parents who wish to act as an informal teacher using a home computer. The workshops are designed to help anyone get started and derive satisfaction from "$010 com- puting." Learning to go solo means to be in charge. Putting a computer to work in the home as a ”solo learning” tool is one of the most existing educational ideas to come along in years. Helping parents learn how to exploit this idea is something every educational institution ought to consider (Dwyer and Critchfield, 1979). The final article in this category by Bell states that the computer revolution in schools which was promised as early as 1965 has yet to come. Learning takes place out of schools as well as in schools. What is happening with more and more personal computers in the home will directly affect their utilization in schools. Also, combined with the factor of low cost, microcomputers may be the catalyst of a learning 51 revolution in our schools. Historically, many technological innovations that are quite useful in promoting learning did not get much use in schools until after they were common in the home and on the streets. Two examples are television and hand-held calculators. The same will happen with small com- puters. The best way to motivate students to learn in school is to pay attention to the nature of out of school learning. Personal computers can motivate students by giving students some real control over what they learn and how they learn it. Bell states that it is fun to make a computer do one's bidd- ing; programming is physically and intellectually a creative activity; good programs impress people; "messing around" in a meaningful way can be relaxing and enjoyable. Therefore, personal computers in the hands of students in schools can remove some of the constraints of typical classroom environ- ments and replace them with some of the personal freedoms inherent in non-school learning situations. Students need to study each subject in a manner that permits them to function at all cognitive levels. Schools are good at imparting knowledge, however, schools are only moderately successful at teaching applications, analysis, synthesis, and evaluation. During the past fifteen years, it has been demonstrated that computers can be used in school to help teach knowledge, understanding, and application of various subjects. But what about cognitive abilities? Here- in lies the power of computers, especially personal computers, I»—nz 52 to really revolutionize learning and teaching in schools. Writing a computer program requires analysis and synthesis of a subject as well as the formulation of the appropriate instruction set. After several years of working on Project Solo at the University of Pittsburgh, it was found that many students and teachers could carry out independent research of their own choosing in computer-enhanced learning environments. Now, personal computers can bring the solo concept of high level, self motivated learning out of the research and devel- Opment laboratory and put it into the hands of large num- bers of students and teachers in school classrooms. Personal computers (with their low cost, easy accessibility, total dedication to the user, and on-the-street pOpularity) may provide the long awaited catalyst that is needed to make some dramatic changes in how computers are used in schools (Bell, 1979). Microcomputers and Time Sharing It is appr0priate here to include some practical obser- vations of Computer Assisted Instruction at the college level. Mowrer reports that student reactions to a time shar- ing approach were twofold: 1) computer terminals caused stu- dents to panic and 2) he, as the instructor who put together the CAL package, could not anticipate, and accordingly pro- vide for, all the correct answers. This latter item led to wrongly scored right answers which, in turn, frustrated the 53 students. Immediate feedback may act as a disadvantage. Mowrer found that computer quizzes were a disaster for mar- ginal and poor students. .Most problems were resolved with the removal of the grading feature, however, quizzes had to be given in the classroom which is contrary to the purpose of CAI. He states that research fails to support the supposed advantages of instructor-less classes. However, the combi- nation of instructor and CAI seems to result in an ideal delivery system (Mowrer, 1979). The articles discussed in the remainder of this sec- tion are not related to Mowrer's experiences with time shar— ing systems. However, the three articles are related to each other and discuss the statewide utilization of micro- computers compared with time sharing systems. First Ahl's article provided the background of the Minnesota Educational Computing Consortium (MECC). Second, a discussion by Hausman of the factors that led to a major commitment to microcom- puters, and lastly, a comparison, by Brumbaugh, of delivery modes. The Minnesota Education Computer Consortium is one of the largest and most highly respected organizations using computers in education. It was founded in 1973 and its major impetus was to meet the increasing demands for computer ser- vices and the desire to provide equality of educational 0p- Portunity. Negative influences on the MECC are funding con- straints and the educators' fears of non-educational agencies, The MECC Board of Directors establishes objectives and 54 policies carried out by a 54 member staff and a $5.6 million budget. The budget is divided into four areas: administra- tion and planning, management information services, special projects division, and instructional services (Ahl, 1981). Although utilization of microcomputers in education is relatively new, Hausman suggests that within the next few years an exponential increase in the number of microcomputers sold to educational institutions can be expected. Several factors are indicative: decreasing cost, increasing capa- bilities, rapid growth in their use,and their independence frommain frame systems (central computer facilities). The latter factor means an increased portability and the elimina— tion of communication-based problems. The Minnesota Educa- tional Computing Consortium (MECC) set out to investigate the possibilities for selection of microcomputers for their statewide educational system. The MECC set up a task force to survey current and future uses of microcomputers, to deter- mine strengths and weaknesses of microcomputers in various instructional modes and environments, to provide demonstra- tions of use, to coordinate and disseminate information, and to prepare recommendations for potential large-scale acquisi- tion and utilization. The recommendations of this committee were: 1) a Specific microcomputer system should be made available to all Minnesota educationally-related agencies through a state contract, 2) instructional service support for selected microcomputers should be defined and increased to the same 55 level as is currently available for large time sharing sys-_ tems, and 3) MECC should continue to analyze and evaluate microcomputer software technology and disseminate information to the Minnesota educational community. Based on these recom- mendations, on October 15, 1978, the MECC and Apple Computer Company signed a contract for 32K, disk based, Applesoft microcomputer systems. Educational computing service agencies must develop plans early if they are to cope with the fast growing micro- computer industry. Support in the following areas must be provided: purchase, installation, maintenance and documen- tation of the system; training in system operations and the use of application packages and programming languages; acqui- sition, conversion, develOpment, maintenance and documenta- tion of applications packages; and lastly, provide avenues for response to questions, problems and requests regarding all aspects of microcomputers (Hausman, 1979). An article by Ken Brumbaugh, Director of Instructional Services of the MECC elaborates on the experience the MECC has had with the use of microcomputers and time sharing sys- tems. To choose between either delivery mode the user must know what his computational needs are. One should not be too quick to disregard time sharing for microcomputers. Time sharing systems are the backbone of instructional computing. Telecommunications costs have been reduced but are still a factor in time sharing systems. Brumbaugh feels that cost and quickness cannot be used for comparisons. 56 Quite often maintenance and Operational aspects of microcomputers are overlooked. Approximately half of MECC's Apple microcomputers are in for repair. Usually these fac- tors are budgeted into time sharing systems whereas the micro- computer people have not considered this. The reported cost of maintaining a standard microcomputer system for a machine which is two or three years old was up to forty percent of the cost of the equipment. Microcomputing and time sharing go together very well. Some of the applications from time sharing could and should be transferred to microcomputers. For the heart of certain instructional programs, one delivery mode ought to be the backup for the other (Brumbaugh, 1981). Summary The computer is best used to facilitate the educational process. The peOple who use it do not need to have a know- ledge of its internal operation. Most of the computing done in higher education in the past fifteen years has been done on large computers, primarily in the areas of research and instruction on computers (computer science). The two tradi- tional approaches in the use of computers in education have been Computer Assisted Learning (CAL),which includes Computer Assisted Instruction (CAI), and Computer Managed Learning (CML). The basic difference between these approaches is that in CAL the learning material is presented to the student through the computer and in CML, the student is directed by the computer and other materials can be used. Computers 57 could be used more effectively in education if we understood more about the learning process. Recently, the only gains in cost effectiveness have been due to learning effective— ness. The original hOpe for low cost, individualized instruc- tion was not only unrealistic, it caused many educators to turn away from the general use of technology in education. Computer Assisted Learning seeks to use the computer to center the instructional material on the student. In this general sense it encOmpasses four paradigms or models. One, instructional CAL which includes computer programmed learn- ing and students‘ mastery of concepts. Two, revealatory CAL in which the computer uses discovery learning to guide and progressively reveal subject matter or a concept to the stu- dent. Three, conjectural CAL, which allows for the testing of ideas and hypotheses in which the computer provides an environment for the student to ask questions. Four, emanci- patory CAL where the computer becomes a tool, like a calcu- lator, and instead of simple numerical calculations, Opera- tions on data bases can be performed. CAL should be considered as part of a learning package or part of a total learning strategy. There are advantages and disadvantages to CAL but one of its advantages is that it can be utilized in the batch mode as well as interactively. The literature is beginning to suggest that CAL might be more appr0priate on smaller machines than larger systems, however,.most CAL is presently carried out on a time sharing basis. In any case, the significance of CAL is the ability 58 to break out of a linear sequence of material to summarize or offer alternative approaches. Hence, provide individualized instruction. Computer Managed Learning (CML) is the ultimate use of the computer in education and training. The intent of CML is to assist with the routine management tasks in teaching and learning. The CML "cycle" consists of: 1) making, giving, and marking examinations; 2) keeping records of student per- formance and progress, 3) providing guidance or advising to students based on performance, and 4) reporting to the in- structor or administrator on the student's performance and progress from records kept on the individual. This is usually accomplished using units or modules of instruction. The concept of CML assumes that courseware can be modu- larized and put into a format that allows mastery of one mo- dule to lead to the next. However, for most educators this approach represents the most dehumanizing aspect of tech— nology to education. In section three of this chapter one instructor's experience with such an approach was presented and it was not at all a positive situation. Yet, despite the negative reactions to CML, there are certain situations where CML properly applied is appropriate, i.e., in a large section of a mathematics or science class, say 300 students or larger. No intelligent individual involved in contemporary higher education in America could deny the trend toward both the increased involvement with information processing and use of computers, A dichotomy amongst educators is emerging: l 59 those who know about computers and those who don't. Through informatics, knowing about information processing and what impact the use and manipulation of data has on society, these differences can be lessened. Similarly, it is doubtful that any educator in higher education has not been intimidated or coerced by another individual_spouting "cybercrud." Computer awareness is simply understanding rudamental computer con- cepts enough to know what computers are and what they are capable of doing. Computer literacy goes slightly farther by combining a low-level working knowledge of computer pro- gramming with an understanding of social, vocational, and educational implications. Informatics should be a part of every college student's training, but this is an impractical suggestion. All educators ought to have some computer aware- ness and computer literacy but this also is impractical to suggest. Without these types of awareness, a division will continue to exist between those who know about computers and those who don't know about them. Computer graphics instructional materials are generally written for a technical audience but more and more graphic algorithms are being presented for the general user or recre- ational computerist. Also, text materials are being intro- duced that fill the gap between introductory programming and computer graphics applications programming. One must be aware that computer graphics is relatively new, graphic imagery and hence, instruction of generating graphic imagery, is hardware dependent and little standardization exists; 60 instruction has been limited to those majoring in engineering and computer science, accessibility to equipment with graphics capabilities has been limited until recently, and computer graphics is just now becoming popularized in the commercial media. The traditional applications for computer graphics have been in Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM) which, in themselves, are growing fields. The general and introductory instructional materials that exist can be found scattered in periodical literature and only two textbooks are available for instruction to lower division college students. Most formal literature and in- structional material demands a first year college math back- ground as well as the fundamental understanding of how com- puters work. In the preceding section several excellent examples of textbooks on the subject were reviewed that per- sons seriously interested in computer graphics should use to begin further investigation. The use of microcomputers in education is just begin- ning to have a major effect. In a few short years microcom- puting has evolved into a powerful force. Its major attri- butes are: low cost, portability, and increasingly powerful capabilities. These machines are not limited to game playing and recreational computing; they are putting computer power into the user's hands and freeing them from a dependence on a large system. .Microcomputers eliminate the problems of telecommunications and the spector of big brother . 61 accountability and eavesdropping usually associated with main frame systems. Microcomputing has made apparent the general nature of the computer. Although the focus of this study was limited to microcomputing in education, the breadth and variety of microcomputing applications in other areas is phenomenal. Microcomputing is impacting education in primary and second- ary schools. Most articles reviewed reported the effects of bringing the microcomputer into those learning arenas. How- ever, the number of microcomputers appearing in college stu- dents' rooms, faculty offices, classrooms, and laboratories is significant. Most students majoring in engineering and science have had previous experience with microcomputers be- fore entering college. If any observation can be made it is that this trend will increase in years to come. In summarizing the use of microcomputers in education it will be appropriate to use the several loosely formed categories described earlier. One must keep in mind that microcomputing did not flourish until the last few years. The first microcomputers were not commercially available un- til the mid-1970's. Therefore, microcomputer utilization in education is pretty much the state of the art. Teachers' attitudes towards microcomputing can be sum- marized as ranging from job threatening to a welcomed addi- tion to the classroom. Although no scientific evidence sup- ports this observation, it seems that those teachers who know the least about computers fear them the most. It is highly 62 unlikely that a computer will replace a teacher in the class- room. Computers should be used as merely another teaching tool available to the teacher. The microcomputer is more likely to be available than a time sharing terminal. Those interested in the use of computers in education are finding that the fears and reactions to computers lie not with the hardware but rather with an existence of an anxiety and avoidance of mathematics. Since some knowledge of mathematics is required for programming and using computers, the computer becomes an object txfbe feared. Whatis necessary is a train- ing program for teachers in the use of the computer, more research needs to be done to help teachers overcome their fears and preconceived notions about computers. Microcom- puters can help because their physical size is less ominous. It is also noteworthy to mention that microcomputers can take over the duties of larger machines, namely CAI. However, in order for microcomputers to become more accepted, certain barriers must be overcome. Computers will have to be inte- grated into learning and life. Lastly, understanding computers is the key to survival in the future. Schools have a primary responsibility for teaching basic survival skills for tomorrow's technological society. The microcomputer is just the tool because of its nature and ease of operation. Educators are in the predica— ment of having to provide these skills but not knowing how to use these tools themselves. A possible explanation may be 63 the fact that most teachers received their educational train- ing before the advent of the microcomputer and widescale computing. ‘ Presently in post secondary education there exists an evolution from minicomputers to microcomputers. Microcom- puters work well with large computers. Quite often what can- not be solved or processed on a small computer can, and should be done on the main frame (central computer facility). .Micro- computers can collect data and send it directly to larger machines with the appropriate communications interface and software. These are the kinds of innovative techniques that abound in higher education. Personal control will make microcomputers more widely used at the college level. Per- sonal computers will promote more scholarship, scholars will exchange software as well as dialogue. Low cost computers are affordable to those in higher education. Personal com- puters may have an effect to the extent that they change the formal classroom to a place where students no longer need the professor but rather a computer software director to manage their courses of study. Microcomputers in schools are forcing us to clarify our values. Some people are too willing to accept the assump- tion that any use of the computer will improve our schools. Computers should not be used to perform existing work, but rather unique applications must be found. Computers will not play a major role, but rather they will play a minor role 64 along with other available teaching aids. Nevertheless, schools have a responsibility to prepare students to deal with computerization. If schools cannot afford the equipment for teaching these new skills, there are organizations that will take these small portable computers to schools that need this ser- vice. This shuttling technique can make computers accessible to large numbers of students. Everyone concerned can experi- ence firsthand educational computing without making large capital outlays. The '80's must be the decade of computer education. Computers have become extremely prolific and can no longer be ignored. Microcomputer learning varies between elementary, secondary, and post secondary education in terms of the age of users and the nature of the computational equipment. There is a drastic need for better software and communication between people who utilize microcomputers in education. .Many discoveries need to be shared and others have yet to be re- alized. The potential of the microcomputer is enormous. Computers can take the drudgery out of learning. Basic skills can be tutored by computers and teachers can take more time to create concrete experiences. To promote more utili- zation of microcomputers in schools requires more integration between computer systems and student-initiated learning. Learning outside the classroom on computers needs to be eval- uated and information networks that facilitate exchanges be- tween computer educators need to be developed. Informal 6S learning promoted by computers in the home can improve family learning experiences which, in turn, can positively affect the performance of children in school. .Educational institu- tions should help parents exploit this idea and become part of the child's educational development. .Motivation in school is improved when educators pay attention to what happens in learning situations outside of school. Programming computers, especially microcomputers, is fun, it is a physical and intel- lectual activity, good programs impress peOple and playing with computers in a meaningful way can be relaxing and en- joyable. Personal computers remove the constraints of the classroom and replace them with personal freedom inherent to non-learning situations. Computing will promote analysis and synthesis of a subject as well as provide an exercise in for— malizing logic by way of writing computer programs. .Micro— computers may provide the long awaited catalyst that may make some dramatic changes in how computers will be used in schools. A final aspect of microcomputers in education is a case study of a statewide school system that made a substantial commitment to microcomputers. The Minnesota Educational Com— Puting Consortium began with a task force to survey the cur- rent and future uses of microcomputers. The recommendations Of this group were: 1) all state educationally related agencies should be provided with a specific microcomputer system if they need or desire such a system - an effort to standardize on a single machine. 2) Support for microcom- puters should equal that available for time sharing systems. 66 3) Analysis and evaluation of microcomputer software tech- nology should be ongoing. 4) Channels need to be maintained for information dissemination. Since MECC's founding in 1973, it has grown to a size worthy of an annual budget of $5.6 million. Their experi- ence in microcomputer based education puts MECC in a unique position from which to speak. Their Director of Instructional Services suggests that before choosing a delivery mode, the user must know what his computational needs are. Time shar- ing systems are still the backbone of educational computing and should not be discarded too quickly for microcomputers. Maintenance and other operational aspects of microcomputers are often overlooked and can be expensive. Microcomputing and time sharing systems complement each other and can serve to back each other up in times of outages. CHAPTER III RESEARCH QUESTIONS AND HYPOTHESES, DEFINITIONS OF IMPORTANT TERMS This chapter is devoted to a precise statement of the research questions and the research hypotheses that corres— pond to them. A brief discussion of significance level is presented. Observations relative to cost will be explained. The remainder of this chapter addresses important terminology that will provide a better understanding of the nomenclatUre used in the study. Introduction Described in this chapter are the issues that lie at the heart of this study. An educator or administrator may take up the task, either by choice or mandate, to evaluate the adequacy of technology-based instruction. This kind of review should be part of an on going process. According to Dressel, ". . . every aspect of the Operation of an insti- tution should be evaluated at some time and its relation- ships with other aspects should be included" (Dressel, 1976, p. 6). As the review of literature suggested, micro- computers are beginning to significantly impact instruction 68 in engineering and the sciences. These small user-oriented computers could have a major impact on the instruction of computer graphics technique. In the future, computer graphics in itself will play an important role in compacting the plethora of information computing in general has exuded. In evaluating computer graphics curriculum as a single course, and in light of the newer technology available, an evaluator might pose research questions similar to those pre- sented in this chapter. Directly related to the research questions were the research hypotheses. Both are presented. The remainder of this chapter is devoted to appr0priate discussion that will make possible the definition of impor- tant terms. Computer graphics systems are explained in terms of the equipment or hardware that can be put together in a variety of ways. Equipment categories are: graphical input devices, computer systems, program and data storage concepts, and output devices. Software and computer languages are also discussed. Terms relevant to the study are defined in order to clarify the research questions and hypotheses. The concept of processing mode and student type are described for these are the independent variables of the study. The official description of the course and related information is pre- sented. The distinction between opinion and attitude needed to be made for this provided the background to attitude measurement. Important terms related to the study of 69 attitude are defined. Finally, the initial concepts of cost assessment and costing modeling are presented as an intro- duction to the deveIOpment of devices in subsequent chapters. Research Questions The following three questions represent the primary thrust of this study: 1. What differences are due to either the effect of processing mode or student type by major on achievement scores in an introductory undergraduate course on computer graphics? 2. What differences are due to either the effect of processing mode or student type by major on stu- dent attitude in an introductory undergraduate course on computer graphics? 3. What observations can be made relative to the cost Of a particular processing mode in an introductory undergraduate course on computer graphics? Research Hypotheses The research hypotheses are derived directly from the research questions. The hypotheses are grouped together by dependent variable. The dependent variables are student achievement and student attitudes. The achievement measure was the final grade in the course, EGR 270. Attitude was measured on two scales: attitude toward computer graphics, and attitude toward computer systems. The information that led to the formulation of these scales is included in Chapter IV. 70 The major reasons for arranging the hypotheses in this format had to do with the design of the study and the presentation of the results. The statistical design that was used to analyze the achievement scores was an analysis of variance. The statistical design that was used to analyze the attitude measures was an analysis of covariance. The design is formally discussed in Chapter IV, the results in Chapter V. The independent variables are processing mode and stu- dent major type. The processing mode represented the "ex— perimental set" and had central batch facility users and microcomputer users as categories. The student type by major represented the "norm group" and had engineering and computer science majors as one category and engineering arts and other type majors as the other category. These major types represented two distinct subpopulations in the students who enrolled in the course, thus they added another dimension to the study. The hypotheses grouping reflected the fact that two experimental factors were studied. Also, two categories of each factor were present. Hence, hypotheses groups asked three questions of interest: 1) Are there systematic ef- fects due to experimental set alone (averaged over norm group)? 2) Are there systematic effects due to norm infor- mation alone (averaged over the experimental set)? And 3) Are there systematic effects due neither to norm 71 information alone, nor to experimental set alone, but attri- butable only to the combination of a particular norm group with a particular experimental set? (Hays, 1973, p. 492). Achievement Hypotheses 1. There will be no difference between the central batch facility users' achievement mean score and the microcomputer graphics users' achievement mean score. 2. There will be no difference between the achieve- ment mean score of the engineering and computer science majors and that of the engineering arts and other type majors. 3. There will be no difference in achievement mean scores between engineering and computer science majors using one processing mode for graphics than the achievement mean score of engineering arts and other type majors using the other processing mode for graphics. Attitude Hypotheses, Attitude Toward Computer Graphics 4. There will be no difference between the central batch facility graphics users‘ attitude toward computer graphics and the microcomputer graphics users' attitude toward computer graphics after adjustment. 5. There will be no difference in attitude toward computer graphics between engineering and com- puter science majors and engineering arts and other type majors after adjustment. 6. There will be no difference in attitude toward. computer graphics between engineering and com- puter science majors using one type of processing mode for computer graphics and the attitude to- ward. computer graphics of engineering arts and other type majors using the other processing mode for computer graphics after adjustment. 72 Attitude Hypotheses, Attitude Towafd Computer Systems 7. There will be no difference between the central batch facility graphics users' attitude toward computer systems and the microcomputer graphics users' attitude toward computer systems after adjustment, 8. There will be no difference in attitude toward. computer systems between engineering and com- puter science majors and engineering arts and other type majors after adjustment. 9. There will be no difference in attitude toward computer systems between engineering and computer science majors using one type of processing mode for computer graphics and the attitude toward computer systems of the engineering arts and other type majors using the other processing mode for computer graphics after adjustment. Significance Level Stated in conjunction with the hypotheses must be the decision rule that was used to test these hypotheses in the final analysis. In this study the conventional rule was used (Hays, 1973, p. 356). If the sample result fell among the highest five percent of means in a normal distribution then H given H was rejected; otherwise do not reject H 0 0 0' H0 represented the null hypothesis. The previously stated hypotheses are in the null form. HA was used to represent the alternative hypothesis. The conventional significance level was represented by alpha, a = 0.05. In practical terms this says that the sample result will be meaningful or significant if it falls beyond the five percent level. In other words, less than five percent Of all samples show results this deviant (or more so) from the ing: sign able poth th_e_ obse whit math tion the to u The allc duti Via for COS qu to Cc be 73 the expectation under HO, if H0 is actually true. By decid- ing in advance that samples falling into this region show significant departure from expectation to be called improb- able results, then doubt is cast upon the truth of the hy- pothesis, and H is said to be rejected (Hays, 1973, p. 337). 0 Observations Relative to Cost It was reasoned that the best methodology for making observations relative to cost was to formulate cost models which represented each processing mode. Cost models are mathematical models that allow predictions or cost projec- tions based on assumptions and historical data. In terms of the central batch facility, costs are directly prOportional to usage, i.e., costs usually increase with increased use. The users were given accounts and the class was given an allotment of computer funds. It was one of the instructor's duties to manage students' accounts, i.e., dole out the funds, via application software. The formulation of a cost model for the batch mode was relatively straightforward because all cost parameters were built into the billing procedure. On the other hand the microcomputer mode cost model re- quired careful thinking. No model existed for costs related to using microcomputers for instruction in higher education. Consequently, a cost model representative of this mode had to be developed. This development is presented in Chapter IV. The model is exercised in Chapter V and a comparison of.modes can be found in Chapter VI. 74 Definition of Important Terms Definitions for key terms used in the study follow to provide a common basis of understanding. Computer Graphics Systems Computer graphics systems use direct commands and pro- grammed instructions with a high speed digital computer to provide a visual display of data and objects. A computer graphics system will contain all of the elements of a com- puter system with additional provisions for data input, display generation, and viewing the results of computations, either as the program progresses, interactive, or after com— pletion of the program, passive. Graphics input devices are used to enter data into the computer. All graphics input devices can be put into these categories: touch sensitive devices, time dependent systems, and coordinate dependent devices. The following are examples of graphics input devices: teletype, joystick, light pen, and graphics tablet. Computer systems are the heart and brains of the com- puter graphics system. A distinction based on word size and core size can be made between the types: Large Computer (Main frame). A large computer, such as the IBM 370, AMDAHL, or CDC CYBER 750, which has a minimum word size of 32 bits and megabytes of core memory. .Minicomputer. A minicomputer usually has 16 bits per computer word (byte) and up to 256 K words of core memory. us] or] 11131 Put 75 Microcomputer. A microcomputer usually has eight bits per byte and up to 64 K words of core memory. .Microcomputers are composed of completely integrated electronic circuits with no external wiring in the logic system (Demel, et al, 1978, p. 78). One of the most distinguishing characteristics of com- puters is their ability to store data and programs. Com- puter storage can be divided into two main categories, namely short and long term storage. Short term storage means that data, intermediate calculations, and programs, have to be available for immediate access internally to the computer as a program is executing. Long term storage is required for data bases and application software that are used periodi- cally or need to be kept on file usually on a storage medium for future use. Core storage is for short term storage, usually used only when a program is in the machine. The other types are for longer, more permanent storage of infor- mation, namely hard disk, flOppy disk, and magnetic tape. Display devices or output devices that allow for the graphical representation of data or objects are the dis- tinguishing factor in a graphics system. Three categories of output devices can be used in computer graphics: printers, plotters, and cathode ray tubes. There are basically two types of printers, line printers and dot matrix printers. Plotters are a second major division of graphics out- ut devices and are used when a drawing needs to be recorded r preserved from a computer graphics system. This drawing, PG ca of 5)’ Co: Th. 76 normally referred to as hard COpy, is produced by a plotter. There are three basic types of plotters: drum, flat bed, and electrostatic. Cathode ray tubes (CRT) are the third category of graphics output devices. These are common display devices resembling television picture tubes that Operate using an electron beam focused on a phosphorous screen. CRTs can be categorized as "Refresh” or "Storage” based on the type of phosphor and technique used to hold the image on the scope. The above definitions refer to the hardware — the physical electronic components that make up computer graphics systems. Equally as important is the software that underlie or allow this hardware to function as graphics systems. Software is simply the set(s) of instructions or commands that constitute computer programs. A new term brought about by microcomputers should be mentioned. Firmware is a pro- gram or set of instructions that is contained in a hardware device like a ROM (read only memory which allows for the permanent storage of data or program on a single chip that cannot be erased even when the power to the machine is turned off). Software can be divided into two categories. The first type is called Applications Software. The users of computer systems normally enter the applications software, which are computer programs, into the computer via some input device. hese programs reside in the computer's memory and allow the USE for pla for lect fram sent the the ( Workj hare ware COmpu Ware. Serie % progr, Venier Cede 1 77 user to perform tasks relative to the application provided for by the software. One very good example of application software that played a major role in this study was the Statistical Package for the Social Sciences, SPSS. SPSS is an integrated system of computer programs de- signed for the analysis of social science data. The system provides a unified and comprehensive package that enables the user to perform many different types of data analysis in a simple and convenient manner. (Nie, et al, 1975, p. l) SPSS was used to analyze the experimental data col- lected by this study. SPSS resides in the University's main frame computer‘s memory. Hence, by way of example, it repre— sents application software that was particularly useful in the study. System Software is responsible for the operations of the controllers for the disk drives as well as the internal workings of the Central Processing Unit (CPU). System soft- ware is strictly hardware dependent. Once the system soft— ware is established, the user can overlook its presenCe. Computer languages are usually facilitated by system soft- ware. The execution of a program generally consists of a series of computer language transformations. The source program is the actual program written by the user. Source prOgrams are usually written in general terms to permit con- venient utilization on different machines. Examples of source Code languages are FORTRAN, BASIC, PASCAL, APL, and others. Thes word Proc more inst schc ning star wheI ing with SGCC shaI sim and the Tett This the Back and USeI acti 78 hese are called high level languages and many use English ords as commands. 'rocessing Modes Even though computer graphics systems are becoming Tore sophisticated and less expensive, many educational -nstitutions still cannot afford them. However, almost every school can have access to a computer that is capable of run- ning batch mode. This term, as well as time sharing and stand alone systems, will now be defined. Batch Mode Processing. An approach to data processing where a number of similar input items are grouped for process- ing during the same machine run. It is generally associated with a single person using a computer at one time (for several seconds or a longer time period) as contrasted with time shared computing in which many users appear to be making simultaneous use of a machine. In the batch mode, programs and data are keypunched and submitted as a deck of cards to the computer center as an entity or "batch" to be run and returned with results after some turn-around-time has elapsed. This type of system is considered to be passive. Time Shared Computing. A method of operation in which ’the computer facility is shared by several users concurrently. ach user typically has access by means of a remote terminal, nd the resources of the computer system are shared by the sers. This type of computing is considered to be inter- 79 Microcomputer Mode Processing. A small personalized computer system, self contained and dedicated to a single user. A microcomputer is a stand-alone system, that is, it does not require or depend on any other computer facility to process and execute programs, just plug it into a wall plug and begin computing. Microcomputer systems contain the appr0priate components to process programs input directly into the computer's memory to be executed real-time. This type of system is considered to be interactive. ProcessingMode. The context in which this term is used in this study refers to the use of any of the above techniques for processing data via a computer. In terms of computer graphics, it is the use of a particular system hav— ing graphics capabilities for the generation of computer imagery and graphics problem solving. Time sharing was not ‘considered in this study. The rationale for this is given in Chapter VI (page 174). Student Types As previously stated in the limitations section of Chapter I, the population was those college students inter- ested in computer graphics and was limited to those people who enrolled in EGR 270 during Winter and Spring Terms, 1981. Although these students did not represent the typical Uni- versity student, they could nevertheless be divided into two istinct subpopulations. As stated earlier in this chapter, ne strata consisted of engineeringand computer science naig mati tot of e stud that For stru Mech fore najc gran stuc dime arts orit thai dis1 dit: font 8O majors. These students could be identified from the infor- mation presented on the class enrollment documents provided to the instructor by the Registrar's Office at the beginning of each term. Specifically, along with the student name and student number, a code is presented, two two-digit numbers that represent the student's curriculum code and major code. For example, if the numbers 36 70 appear it tells the in- structor that the student is enrolled in a degree program in Mechanical Engineering. This type of student would, there- fore, belong in the group of engineering and computer science majors. A complete list of these codes for the degree pro- grams in the College of Engineering is provided in AppendixIL Similarly, the other strata consisted of engineering arts and other type majors. This group could be identified by the designation 39 50 or any other designator indicating a major type outside of the College of Engineering, or not on the list presented as Appendix A. The reason for this stratification or disaggregation of student types based on major was to allow for an additional dimension in the student population in EGR 270. Engineering arts and other type majors are somewhat less technically oriented in terms of the courses they are required to take than engineering and computer science majors. By making this distinction in the students who enrolled in the course, ad- ditional patterns and interactions based on major type were found that would not have emerged without stratification. Introdi in Com] E61 Sp: or Use 502 the Mic academi only bl Term as credit nechani ductory 1 din B, may. as a s Studen Sites. Stude: in? a to at is b: Prob IESP 81 Introductory Undergraduate Course in Computer Graphics EGR 270 Computer Graphics Spring 3 (3-0) EGR 160 or EGR 161; CPS 110 or CPS 120; or approval of department. Use of computer controlled display systems for the solution of multi-dimension problems. (MSU, 1980) The above description of EGR 270 is reproduced from the Michigan State University Description of Courses for the academic year 1980-1981. It is normally offered Spring Term only but special permission was acquired to offer it Winter Term as well for the purpose of this study. It is a three credit class, lecture format, and has, as prerequisites, a mechanical drawing course, EGR 160 or EGR 161, and an intro- ductory computer science course, either CPS 110 or CPS 120. The course outline and syllabus are included as Appen- dix B. EGR 270 is administered by the Department of Metal- lurgy, Mechanics and Material Science, College of Engineering as a service course. A service course can be taken by any student,regardless of major,who meets the required prerequi- sites. Student Attitude The distinction between asking for Opinions and measur- ing attitudes needs to be made before definitions relative to attitude measurement are stated. The‘study of Opinions is basically more troublesome than facts. There are several problems to consider. One, the uncertainty of whether the reSpondent in any meaningful sense, "knows" the correct answe‘ be mm upon ' Three ansne‘ in we: tual r are u. 92ini what j a giv. furth. asses. as a 1 dents t0 a i is th: The p: this . tion. at th 82 answers. Two, a person's Opinion on virtually any issue may be many sided. The answer the respondent gives will depend upon the aspect of the issue that is uppermost in his mind. Three, intensity of opinions vary between individuals. Four, answers to opinion questions are more sensitive to changes in wording, emphasis, sequence, etc., than are those to fac- tual questions. In the face of these problems, two distinct approaches are used in opinion and attitude inquiries. One, asking for ppinions is the most common and allows for the estimate of what proportion of the survey population say they agree with a given Opinion statement. Two, measuring attitudes goes further by including a number of Opinion statements and assessing the reSpondents' answers to the set of questions as a whole. This approach attempts to measure the respon— dents' attitudes, which typically means combining the answers to a set of Opinion questions into some sort of score. This is the method of attitude scaling (Moser and Kalton, 1972). The procedures used in forming the attitude scale used in this study are discussed in Chapter IV in the procedures sec- ion. However, several salient definitions are appr0priate t this time. Attitude Instrument. A device, in this case a series f questions, that provides a means of identifying factors nd assessing student attitudes relative to these factors in n attempt to measure student attitudes relative to the va- riables under study. fact niqu iron Ject A fa $01116 Cost nit‘ hno nah edr Ti 83 Student Attitude Measures. The scores on a series of factors which were determined by using a statistical tech- nique called factor analysis and the resultant responses from the preliminary study of the attitude instrument. Factor Analysis. A statistical technique that allows the orderly simplification of a number of interrelated mea- sures. The primary aim of factor analysis is the discovery of common factors. Factor. An identifiable trait, sometimes highly sub- jective, that emerges from the process of factor analysis. A factor is said to exist when a group of variables has, for some reason, a great deal in common. Cost Assessment The basic principle behind finding the costs associated with the various aspects of higher education is that this knowledge will provide another important dimension to decision making. Complete cost assessment of any program in higher education includes: Direct Costs. The direct dollar expenditures in sala- ries, supplies, and equipment. Indirect Costs. Costs which are not reflected in ex- penditures, yet they are real and interpretable in dollars, e-g., depreciation of facilities and equipment. Time Costs. Costs which are quantifiable although not eadily expressable in dollars, e.g., student and volunteer ime, use of facilities, and contributed services. efiec puhli most dealt prov: tion: cost pute but stud pain as t Thi: dir var re: 84 Non-quantifiable Costs. Faculty and student morale, effect on programs, reputation of institution, academic, and public reactions to program and personnel. However, a complete cost assessment is unlikely within most evaluation studies in higher education. Costs.must be dealt with and "every reasonable effort should be made to provide estimates of program costs and alternatives in rela- tionship to benefits" (Dressel, 1976, p. 131). This study did not attempt to address the total program costs involved in presenting an introductory course on com- puter graphics. Not only is it beyond the scope of the study, but according to Dressel, it is practically impossible. The study did take a narrower view of cost, a more practical point of view. The basic unit of comparison that was decided as being most appropriate was cost per student per term. This figure contained several aspects of both direct and in- direct costs. Salary of instructor was not included as a variable. Cost models were formulated that represented the respective processing modes and enabled comparison. John Keller, an early proponent of models, noted that model building is useful in four ways: 1. Development of the model automatically forces a deeper understanding of the interactions within the system under study. 2. .Models permit the evaluation of a wide range of alternatives 4 surely a key feature of cost/benefit analysis. 3. Models help provide a hedge against risk and un- certainty. 4. Where low confidence is associated with the.most probable values of key input parameters, models are supposed to help cope with uncertainty. (Skubal, 1977, p. 10) sip tior latt [Sh C051 cert per tren edut 85 Models, a term which can be used interchangeably with simulations, in their simplest forms are "mathematical [equa- tions that] approximate the representation of reality formu- lated to capture the crux of the decision making problem" (Skubal, 1977, p. 10). This was the exact intent of the cost models that are presented in Chapter IV. A cost model is a model that allows for the prediction and extrapolation of cost data. Later it will become apparent that based on certain reasonable assumptions, a figure representing cost per student per term can be Obtained. Cost models are ex- tremely valuable as an aid to decision making in higher education. tions: the st1 treatme delinea the pre T] qUestior deSign a ism that Testing each nul hate 11th SBS wolm be addres A s OVer tWO “der Stu the fOTma CHAPTER IV DESIGN AND PROCEDURES The material in this chapter is divided into two sec- tions: the design Of the study and the procedures used in the study. The design indicates the relationships between treatment groups, the measurements, and time. The procedures delineate the sequence of steps that were followed during the preparation and execution of the study. Introduction The design for this study was dictated by the research questions stated on page sixty-nine in Chapter III. The iesign and statistical procedures sought to provide a mechan- ism that would allow the research hypotheses to be tested. 7esting the research hypotheses entailed either accepting rach null hypothesis or rejecting it in favor of the alter- ate hypothesis. Accepting or rejecting the research hypothe- es would, in turn, enable the original research questions to e addressed. A sample was taken from two classes of EGR 270 taught fer two terms which was representative of the pOpulation rder study. The design over time will thoroughly explain Le format of the experiment. There were two independent 86 majo achi vari pernf Offit grade of Co This analy of Ta: permit approi Commit APprov this e Studen granter dic c z adminis mlysj sUtes. isolate the cov Variablr tw° Sta: 87 or design variables: 1) processing mode, and 2) students' major type. The dependent variables were twofold also: achievement scores and attitude measures. Originally the plan had been to do an analysis of co- variance on both achievement and attitude. However, when permission was requested from the Michigan State University Office of the Provost for the prerequisite course final grades, permission was denied by the Committee on the Release of Confidential Information and Approval of Questionnaires. This memorandum is included in Appendix C. As a result, analysis of achievement scores were done using an analysis of variance rather than an analysis of covariance. Also, permission to use the questionnaire for this study had to be approved by another University committee, the University Committee on Research Involving Human Subjects (UCRIHS). Approval of utilization of the questionnaire was critical to this experiment as the measures Of attitude were based on student responses to these instruments. This permission was granted and the apprOpriate memoranda can be found in Appen- dic C also. Thus, two versions of the attitude survey were administered, a pretest version and post testversion, and an analysis of covariance could be done on the attitude mea- sures. As will follow in this chapter, two factors were isolated and scales were formulated to measure them. Hence the covariates were the pretest measures and the dependent rariables adjusted by them are thepost testnmmsures of these :wo scales. 88 The design section of this chapter includes the infor- mation relevant to the design of the study. Both the design over time and design over variables will be elaborated upon. Lastly, validity and reliability concerns will be addressed. The procedures section explains the details of what was done. Specific procedures were followed in preparation for, and in the execution of, the many aspects of this study. This section will address the concerns of population and sample, including the enrollment profile in EGR 270 over the past five years. The treatment section will explain a novel technique used to pick the sample and explain how the mea- sures were administered. The body of the procedures section deals with instrumentation and data collection and will be divided into three parts: achievement measure, attitude measure, and formulation of the cost models. Design Design Over Time Design over time refers to Campbell and Stanley's no- tation used to describe the experimental design (Stanley and Campbell, 1963). This notation indicates relationships between groups, measurements, and time. It also shows that stratified random assignment has been used to initially equate the groups. The specific purpose of presenting the design over time using this notation in this section is to specif sectio addres Design evalua‘ to eat at the shown 1 random Popula naj 0r 1 sure t] node i] PTOCes: ETOHp ( batch . “mutt eXeCut( graphi‘ 89 specify the experimental design over time, in a subsequent section the validity concerns of these designs will be addressed. Design Over Time — Achievement The dependent variable, achievement, was studied by evaluating the composite final grade in the course relative to each treatment group. Each treatment group was formulated at the beginning of the term by major type. The deSign is shown below using the previously mentioned notation. batch SR X1 01 proceSSIng mode microcomputer SR X2 01 ”SR” indicates that groups were assigned by stratified random technique. This technique will be discussed in the population and sample section of this chapter. Again, the major purpose of proportionate stratified sampling is to in- sure the presence of each type of student in each processing mode in a proportionate manner. "X1" indicates the treatment of the use of batch mode processing (present practice). Students in this treatment group executed their assignments on the University's central batch facility. "X2" indicates the treatment of the use of the micro- computer graphics system. Students in this treatment group executed their assignments on the "personal computer" with graphics capabilities. an achi In tern mean al Design A score c measure observe in this able. notatic T as prev ll Were us tUde. in 3 ma cWipers H by the t" all A "02" ar 90 "01" indicates the post hoc Observation. In terms of an achievement measure this was the final grade in the course. In terms of a group's achievement mean score this was the mean average grade of those students who comprised that group. Design Over Time — Attitude Attitude was studied in terms of the composite mean score of each group in terms of the two scales or attitude measures. Unlike the design over time for achievement, two observations were made. Pretest attitude was the covariate in this design. Post test attitude was the dependent vari- able. The design is shown below using Stanley and Campbell's notation and was used for both factors of attitude batch SR Ol X1 0 processing mode microcomputer SR 01 X2 02 The terms "SR" ”X1" and "X2” have the same indications as previously stated in the achievement section. ”01" indicates pretest attitude. The pretest responses were used as covariates for each factor that made up atti- tude. Basically, these scores adjusted the post test score in a manner analogous to a "handicap" in bowling or golf :ompetitions. "02" indicates post test attitude. After being adjusted >y the pretest responses, thepmwt testscore for each fac- :or allowed for the examination of attitudes. At this point it is noteworthy to mention that "01" and '02" are two different versions of the same attitude survey or ins ing or simple of the Desigr Two-W: was s analy State into pende siste pende engin and o expeI Varia desig gory 0that all] Unmb‘ hYpo 91 or instrument. The difference between the two is a reorder- ing or "shuffling" of the questions (see page 101). This simple technique purports to increase the internal validity of the instrument. Design Over Variables Two-Way Analysis of Variance In the research hypothesis section of Chapter III it was stated that the statistical design that was used to analyze achievement was an analysis of variance. It was also stated that the independent variables would be disaggregated into an ”experimental set” and a "norm group." For the inde- pendent variable processing mode, the experimental set con- sisted of batch mode and microcomputer mode. For the inde- pendent variable, student major type, the norm groups were engineering and computer science majors and engineering arts and other type majors. This contrived arrangement of two experimental factors is known as a two-way analySIs of variance. The design matrix took the form of a completely crossed design, two experimental factors were present and each cate- gory or level of one factor occurred with each level of the other. Furthermore, the experiment was balanced, that is, all possible combinations of factor level occur an equal number of times. This particular design determined the context of the hypotheses as was noted on page seventy. This study can be ‘ set pro to two exp mai des of bee eff int era the fee Twc var Alt des th: per thr Vic 92 be viewed as conducting separate investigations on the same set of subjects: 1) there were two groups differing only in processing mode (averaged over the norm group). This came to be known as the group type main effect. 2) There were two groups differing only in major type (averaged over the experimental set). This came to be known as the major type main effect. By virtue of using a crossed and balanced design a third inquiry was possible. 3) The unique effects of combinations of treatments (after each main effect has been deleted). This was referred to as the interaction effect. In contrast to the main effects, the first two investigations allowed separate experimental variables to be examined, interaction effects showed differences due only to the unique combination of treatments. This was the desirable feature of two-way analysis. Two-Way Analysis of Covariance The variable matrix for the two-way analysis of co- variance was exactly the same as for the analysis of variance. Although the statistical model was somewhat different, the design matrix is identical. Dosign.Matrix Table One illustrates the format of the design matrix that was used to report achievement and attitude factor ex- perimental data. It classified the components that comprised the matrix. As was previously stated, this study can be viewed as testing three sets of hypotheses about the same —¥—‘ H m4m~ paw EHOH meH meH owma whoa wnma nmma whoa am so so am am am no filial .. 1 .IOH .IA .Iom ]_1 I '1’ .lom mpcopsum "lini was a 110 111C sampl the 1 the 1 mize< are 1 for l and and Trea doml with in l list ques Shut flir iter all 101 It was stated early in Chapter I, in the section on "limitations" that only one microcomputer graphics system was available. In that same section it was also stated that no more than eight students ought to interface such a system as a processing mode in a single term. Hence, our goal in sampling was to insure that: 1) those students who made up the processing mode groups were stratified proportionate to the total sample; and, 2) the microcomputer group was maxi- mized to contain eight students. The results of this attempt are presented in Chapter V in a series of three tables, one for Winter Term, 1981, a second one for Spring Term, 1981, and a third showing both terms combined (Tables Seven, Eight, and Nine). Treatment Two aspects of the study data had to be shuffled or ran- domly reordered. The first application of shuffling occurred with reordering the questionnaire. The algorithm presented in flowchart form in Appendix E produced a randomly reordered list of items which was used for the post test version of the questionnaire. Table Three presents the original and re— shuffled list of item numbers as an illustration of reshuf- fling. Reshuffling simply rearranges the order of any set of items in a random fashion. stions but in a different order. Table Note that the post test contains all twelve original que Thre< that W35 1 from Popuj type then thus Proc. ing was The of e the micr c133 that 102 Three represents the results output from a computer program that was used to generate the reordered item sequence which was utilized in the post test questionnaire. TABLE 3 PRETEST AND POST TEST ITEM ARRANGEMENT GENERATED BY _SHUFFLE ALGORITHM Original Order of Items 1 2 3 4 5 6 7 8 9 10 ll 12 Unshuffled (Pretest) Reordered Questionnaire 2 10 6 8 l 7 12 5 ll 9 4 3 Items after Shuffling (Post Test) The shuffle algorithm was also used to pick the samples from the respective major type pools formed from the class pOpulation in EGR 270. The student numbers of each major type were entered into the shuffle algorithm, shuffled and then the appropriate sample was "taken off the top" namely, those first n students from the reordered sequence. This procedure is exactly like shuffling a deck of cards and deal- ing (face down) the number of required cards. This procedure was followed for each strata of the population each term. The result then was a randomly selected proportionate sample of each class. Students picked in this manner who comprised the sample group were assigned to treatment group A, the microcomputer mode. The remainder of the students in the class became treatment group B, the batch mode. For the exact numbers, refer to Tables Seven, Eight, and Nine. Instrume Tl divided cost. dent rev achiever sun of ' group. which i forward that war here in cluded accompl the dir be pres they We 103 Instrumentation and Data Collection The instrumentation or measures in this study can be divided into three categories: achievement, attitude, and cost. The achievement measurewnusthe final grade each stu- dent received in the course, EGR 270. The mean average achievement score of a group is the arithmetic mean of the sum of the scores of those individuals who comprise each group. The grading procedure is outlined roughly in the text which follows. The measurement of attitude is less straight- forward than the measurement of achievement. The process that was utilized to derive the attitude factors is presented here in a chronology of steps, and supporting material is in- cluded in Appendix D. Observations relative to cost were accomplished by the formulation of cost models replicating the direct costs of each processing mode. These formulae will be presented in this section along with the process by which they were derived. Aghievement Measure In the course outline presented in Appendix B, the "basis for grade" defines the procedure followed in deriving the final grade. The final grade represents the individual student's reported grade in the course. Group achievement mean scores were a function of the arithmetic mean average of these grades for those individuals who made up each group. Three components constituted a final grade: homework, com— puter graphic assignments (the best six out of seven of these were too these cc cent of tial rat ed grade cedure what of 1 course, mary da study. student Syllabt ask qu assign the pr might have t dates Appenc ments given twent we of hers Outlj 104 were counted), and a "special" assignment. Respectively, these components represented twenty, sixty, and twenty per- cent of the final grade. Each student could receive a poten- tial raw score of 250 points in the course. Final or report- ed grades were determined based on the norm-referenced pro- cedure of curving the raw final scores in order to get some- what Of a normal distribution of grades. In addition to the course outline and syllabus of the course, Appendix B contains the assignments, achievement sum- mary data, and distribution of scores for each term of the study. At the beginning of each term of the experiment, each student was presented with a copy of the course outline and syllabus and, in discussion, was given every Opportunity to ask questions or seek explanations about its contents. The assignments were identical each term with the exception of the problem option. Several problems have choices, Option B might have been given Winter Term, whereas Option C might have been given in the Spring Term. Also, the assignment due dates were changed according to the term. All assignments in Appendix B are from Spring Term, 1981. As stated, all assign— ments, with the exception of number five, the special problem given at the middle of the term, have a potential value of twenty five points. Assignment five had a maximum point val- ue of fifty points. The lowest score on an assignment, num- bers one to four, and six to eight, was not counted. Assignments were graded according to specific criteria, outlined by the assignments, and by due dates. As was previous the fact the same measure: veal th larly, coded s stateme ambigui was ma) , score 1 "bad d dents demand the gr 0n Stu benefj One, 5 cours. ment, uPOn TEQUI a bon Iowa: then 105 previously mentioned in the limitations section of Chapter I the fact that the researcher and the instructor were one and the same person had a minimal effect on student achievement measures. A review of the assignments in Appendix B will re- veal that specific, objective solutions were required. Simi- larly, homework assignments required numerical solutions or coded solutions based upon specific computer programming statements. These kinds of assignments leave little room for ambiguity. Thus, the validity of the achievement measure was maximized and its credibility maintained. At this point an explanation of "dropping the lowest score phiIOSOphy"is noteworthy. In an attempt to provide for "bad days" or to take into consideration the fact that stu- dents may have periods during the term when other courses demand more of their time, this provision was included into the grading procedure. This provision has a positive effect on students and was discovered several years ago to be of benefit. This device has three effects on the students. One, it tends to improve the student's overall grade in the course. Two, by eliminating or not counting a missed assign- ment, students are relieved of some of the pressure placed upon them when they get ill or several courses simultaneously require material from them. Three, it is sometimes used as a bonus by those students who have worked hard all term, a reward. If a student's first six assignments are acceptable, then the last assignment need not be done as a reward for success: to be a motivat F summary homewor cent of ments 2 special twenty putatic presen' tribut achiev Attitu 0f thi ferenc From t resear in Ch: Cessil Sign 1 dents the c ment. 106 successful performance over the term. This device has proven to be a simple technique for improving morale and increasing motivation of the students in EGR 270. Following the individual assignments in Appendix B, the summary data for bOth terms are presented. There were nine homework assignments given which accounted for twenty per- cent of the grade, the best six out of seven regular assign- ments accounted for sixty percent of the grade, and the special assignment (assignment number five) accounted for twenty percent of the grade. In the summary data, the com- putations which led to the final raw score in the course are presented. Lastly, at the end of this information the dis- tribution of scores and ranges which determined the final achievement score in the course are presented. Attitude Measures The endeavor to measure attitudes was a major effort of this study. As was stated in Chapter III, there is a dif- ference between asking Opinions and measuring attitude. From the very beginning of this study, as indicated by the research questions stated in Chapter I and formally stated in Chapter III, there was an interest in the effect of pro- cessing mode on students' attitudes. The experimental de- Sign provided for the initial measure of attitudes of stu- dents coming into the course and once again at the end of the course utilizing a modified version of the same instru— ment. The analysis of covariance adjusted responses using the outl tude beer Mose 107 the pretest measure as a covariate. This section then will outline those procedures taken in the development of an atti- tude instrument. The majority of the information presented here has been taken from Survey Methods and Social Investigation by Moser and Kalton, 1972. This investigator tried to develOp an instrument based on the considerations and procedures dis- cussed in this text. Those issues pertinent to develOping a survey instrument will be presented here under Considera- tion and Procedures. The Implementation section will address the specific develOpment of the attitude questionnaire, in- cluding the issues of reliability and validity. The results of using the instrument as an experimental tool are presented in Chapter V. Considerations There are certain considerations required of a re- searcher in the initial formulation of attitude scales. Scal- ing devices are used to try to combine the answers given by respondents to various questions into a measurement of the extremity and intensity of their overall attitude. This is an analytical approach. A single belief is a poor indica- ° ‘ to measure tion of a person's more general att1tude_and so, this more accurately, a sample of beliefs covering a range 0f aspects of the attitude needs to be obtained and the set of answers combined into some form of average. Averaging reduces the effects of ideosyncracies of particular respond! The set chosen ‘ samplin E wishes each It The at the rj attiti lowel (item selec Cess: 108 respondents in respect to particular aspects of the attitude. The set of statements used for ascertaining beliefs must be chosen with this in mind. Choosing statements is a kind of sampling problem from the universe of content. Essentially, scaling methods come into play when one wishes to utilize simultaneously a number of Observations on each respondent. There are three difficulties in scaling: 1. To decide what are the appropriate pieces, en- suring that they are logically related and all refer to the same attitude dimension (which be- tween them they span adequately). 2. To fit these pieces into a meaningful whole. 3. To test the properties, particularly the relia- bility and validity of the scale. The advantages of combining answers to many questions is that the risk of bias through wording is thereby reduced. Most attitude scales attempt nothing higher than the interval level of measurement. Procedures The first stage is to assemble a set of statements (items) from which those employed in the final scale will be selected. There are four basic considerations In this pro- C655! 1. Principles of wording. Avoid complex items, ambiguous items, items involving double negatives, vague items, etc. ~ items eral t and Se this c in whj (Thur: 3 pant using 109 Items should be chosen to differentiate between those with favorable and those with unfavorable attitudes. For this reason items to which all respondents answer in the same way are unsatis- factory. Thus, deliberately sensing items, formu- lating a balance of positive and negative ques— tions, will achieve this purpose. Avoid factual items. Items must range broadly Over all aspects of the attitude. After having assembled an item pool, choose from it the items to be used in the final scale. There are several gen- eral types of scales to consider: Likert, Guttman, Thurston, and Semantic Differential. With all but the Thurston scale this choice can be made on the basis of an exploratory study in which a group of individuals are asked to respond. (Thurston scales are made on the basis of the assessment of a panel of judges.) An analysis of items should be performed using the following guidelines: 1. Discard items that do not "hang together” with the general pool. Discard items with wording faults. After removing the unsuitable items, a selection is made from the remaining items in the pool to form the final scale. Care should be taken to make sure the universe of content is adequately covered. appr scal with wide ment agrt The tit! of - of Sun thi ite dre are att C011 was The 110 4. Pilot study and analyze based on the preceding. 5. Once the scale is formed, checks can be made on reliability and validity; if these prove satis- factory, the scale can be administered. Implementation It was decided that the Likert scale would be the most apprOpriate scale for the attitude instrument. Likert scales do not seek a respondent's agreement or disagreement with an item. Rather, several response categories are pro- vided indicating various strengths of agreement or disagree- ment. The response categories are five in number: strongly agree, agree, uncertain, disagree, and strongly disagree. The categories are assigned scores and the respondents' at- titudes are measured by their total scores, which is the sum of the scores of the categories they have endorsed for each of the items. Because of this, Likert scales are known as summated or summated rating scales. The further aspects of this type of scaling, involving item sensing, item pooling, item analysis, etc., is addressed later in this section. Twelve question areas were compiled that seemed to ad— dress student attitudes in the area of study. The two major areas of attitude upon which it seemed relevant to focus were attitudes towards computer graphics and attitudes towards computer systems. Advice in developing these question areas was sought from the professor who previously taught EGR 270. The aim of these question areas was to adequately cover the universe both comp the centr sented in of formul that the ficient. the queS‘ Th' into two the atti struct we sent nor be reckc 0r QUEStiOI sentatir two 56p: most ap] area as eVidenc ing the Study 0 reQuisi T are 115 111 universe of content. It was reasoned that attitudes towards both computer graphics and computer systems combined were the central issues worthy of study. The question areas pre- sented in Table Four represent the first phase of the process Of formulation of the attitude instrument. It was decided that the six questions from each attitude area would be suf- ficient. Also indicated in this table in the proximity of the question is the appropriate attitude area. The fact that the question areas were initially divided into two attitude areas will be utilized later in isolating the attitude factors, also it was an attempt to provide con- struct validity; The isSue that these questions might repre- sent more than distinct attitudes in these areas also had to be reckoned with. This concern is addressed later. Once the question areas were decided upon, two separate questionnaires were developed. Both had one question repre- sentative of each question area. The purpose of develOping two separate questionnaires was twofold. One, to find the most appropriate statement (item) to address each question area as given and to make this choice empirically based on evidence. Two, to minimize the respondent's time in answer- ing the questionnaire. Since the population utilized in this study of the preliminary questionnaires was students in pre- requisite courses, the use of class time had to be minimal. The initial two versions of the preliminary instruments are listed in Appendix D and are entitled Form 1 - Red Ver- sion and Form 2 - Blue Version. Separate colored mm 10 11 016363063 112 TABLE 4 .QUESTION AREAS G = attitude towards computer graphics 8 = attitude towards computer systems 1. the appeal Of the process of doing computer graphics 2. the perceived adequacy of present practice (batch mode graphics) in an introductory computer graphics course 3. motivation to take the course 4. preconceived notions about large systems 5. turn-around time 6. programming language 7. perceived complexity of computer graphic algorithms 8. 1 computer graphics as work or fun 9. perceived applications for computer graphics 10. apprehension or eagerness towards computer graphics 11. attitudes towards the use of a microcomputer in the course 12. system size: convenience vs capabilities Group G Questions: 1, 3, 7, 8, 9, 10 Group S Questions: 2, 4, 5, 6, 11, 12 questionn he quest positive of each q sions. For the sense strength particulz the grea‘ more fav computer Pittlimin Th both ini domly sc courses, tiVe to Hence, ; dents f approxi decided five pa in the 113 questionnaires and response sheets were used. The senses of the questions were modified by careful wording to be either positive or negative. Table Five illustrates the sensings of each question as they appeared in the two initial ver- sions. For the purpose of scoring it was important to know the sense Of each question. The goal was to measure the strength of intensity of each respondent's attitude to a particular question. Thus, after a process of ”recoding," the greater the individual's score on the questionnaire, the more favorable the student's attitude shOuld have been toward computer graphics and small systems. Preliminary Study - Form 1 and Form 2 .The second phase of the preliminary study involving both initial versions of the questionnaire was given to ran- domly selected sections of students in the prerequisite courses. Generally, the rule of thumb for sample size rela— tive to questionnaire size is: n = 2(I)+1 (Child, 1970). Hence, for twelve questions or items, I = 12 n = 2(12)+1 = 25 i This investigator tried to get equal numbers of respon- dents from each prerequisite course. With an enrollment of approximately fifty students per section in CPS 120, it was decided that two sections of EGR 160 (approximately twenty five per section) would be given the questionnaire. Students in the two CPS 120 sections and four EGR 160 sections were Ques Questio asked f0 not to p 1“ APPen tionHair ing data Which th so, it M courSes n°t to j 114 TABLE 5 QUESTION SENSES Questiofi7 Sense On Sense on Question Area Form 1, Red Version Form 2, Blue Version 1 + + 2 .. - 3 - + 4 - - 5 + - 6 — + 7 + - 8 + - 9 + - 10 - + 11 + + 12 - + asked for their voluntary participation and given the option not to participate if they felt so inclined. The statement in Appendix C was read before the administration of the ques- tionnaire in compliance with UCIRHS guidelines. The follow- ing data shows the number of respondents and the course in which they were enrolled when the survey was conducted. A1- so, it was possible for a student to be enrolled in both courses in the same term, therefore, these students Were asked not to fill out the questionnaire a second time. Thefbllowing figures identify Th cents, s son's pr terms of Ti the comt D). The were f0] item 0V6 Those v5 combine¢ T} giving 1 dents. EGR 160 Were Pre the 5am} 115 figures represent complete and unspoiled questionnaires. No identifying data, i.e., name or student number, was solicited. FORM 1 FORM 2 CPS 120 students 33 32 EGR 160 students §§ 43 Total (sample size) 89 5? The response data, including relative frequency in per— cents, summary statistics, interitem correlation using Pear- son's product moment correlation, and overall reliability in terms of Cronbach's Alpha, is given in Appendix D. Preliminary Study Combined Form The third phase of the process was the formulation of the combined questionnaire from the response data (Appendix D). The guidelines established in the Procedures section were followed. The most helpful statistics in choosing one item over another were relative frequencies (percent). Those versions of the items that were chosen to make up the combined questionnaire are listed in Table Six. The fourth phase of the preliminary study involved giving the combined questionnaire to randomly selected stu- dents. Again, one section of CPS 120 and two sections of EGR 160 were utilized. The procedures outlined previously were precisely followed. The following numbers represent the sample size. COMBINED FORM CPS 120 students 37 EGR 160 students 43 Total '§6 intGTES‘ tistictt' Form 2. Alpha 5 for FOr CaSes a that th ity SCQ 116 TABLE 6 COMBINED QUESTIONNAIRE SOURCE AND SENSE ‘Combined Form Source Item Number Version (Form) Sense 1 1 + 2 2 - 3 l - 4 l - 5 2 - 6 2 + 7 2 _ 8 1 + 9 2 - 10 2 + 11 2 + 12 2 . + The response data is presented in Appendix D. It is interesting to note that the combined questionnaire is sta- tistically a much better instrument than either Form 1 or Form 2. The overall reliability improved to a Cronbach's Alpha score of 0.585 compared to 0.456 for Form 1 and 0.148 for Form 2. Also, the interitem correlations in all but two cases at the significance level (P) is 0.001. This shows that the items "hang together" fairly well. Along with the concern for a moderately high reliabil- ity score which 0.585 represents, validity concerns needed to be kep was menti validity. attitudes puter sys be taken collecte' form and However, made of the next Cc these 01 into tht scales ' ment of data an in the 0' each that ft 3 SCale three j bility terms within 117 to be kept in mind. In an early section of this chapter it was mentioned that the goal would be to achieve construct validity. By originally formulating questions in two areas, attitudes toward computer graphics and attitudes toward com- puter systems, a close look at these attitudinal areas had to be taken. The approach that was chosen was to use the data collected from the preliminary study of the combined survey form and eventually to do a factor analysis on the scales. However, before this could be done, an inspection had to be made of those questions that made up each scale. This was the next to the last phase of the preliminary study. Construct validity would be achieved by going back to these original question areas and separating the questions into these two areas by dividing the response data into two scales representative of these two areas. Further refine- ment of these scales was accomplished using the response data and a feature in SPSS entitled "Alpha if deleted" with- in the RELIABILITY procedure. Hence, the optimal reliability of each scale was achieved. The result of this process was that for the computer graphics scale, by deleting one item, a scale reliability of 0.599 was achieved. Also, by deleting three items in the computer systems scale, a scale relia~ bility of 0.617 was achieved. Scale reliability was given in terms of Cronbach's Alpha. As a result of this process, two scales were isolated within the overall instrument; their separate reliabilities greater These sc of the r to good tions of Th the comb The fact scale ar in the a cedures not that inciden1 routine: rESpons< dueed t] Scale h; than on. The COmj account W measuri tudes t Sefims t attitud Statist 118 greater than the overall reliability of the whole instrument. These scales reflect the originally postulated expectations of the researcher to what was felt would be a satisfactory to good measure of the attitude areas. The actual calcula- tions of "Alpha if deleted" are presented in Appendix D. The last step in evaluating the preliminary study of the combined questionnaire was the factor analysis itself. The factor analysis would reveal the homogeneity of the scale and the weightings of item responses to be used later in the analysis of experimental data. Factor analysis pro- cedures produce the same results on response data whether or not that data is standardized. This was found out in an incidental fashion bytflfiesresearcher using factor analysis routines in the SPSS package on both raw and standardized response data. Nevertheless, both analyses on the data pro— duced the same results, namely that the computer graphics scale had one significant factor present (eigenvalue greater than one) accounting for 75.3% of the variance in the scale. The computer systems scale had one significant factor present accounting for 100% of the variance in the scale. What this means is that the computer systems scale is measuring a single attitude having to do with students' atti- tudes toward computer systems. On the Other hand, there seems to be the slight influence of another factor in the attitude toward computer graphics scale but it did not seem statistically significant enough to cause great concern. The p0: tic am] Ho1 spl thi be in or da th 51 pe in La On It de me Inc St 8T 119 possible explanation for this might be the different percep- tions students might have about computer graphics and certain ambiguities and imprecision in the questions themselves. However, since most of the variance is explained by the re- sponses to the five items in the questionnaire representing the computer graphics attitude, this scale was considered to. be a good indicator of a student's attitude toward computer graphics. The factor analysis of scales computations are presented in Appendix D. The variables are the items. Along with overall computations of item communality and factor analysis data, the factor score coefficients are included. It was these coefficients that were used as weightings in the analy- sis of covariance of the experimental data. Cost Model The history of costs in EGR 270 is illustrated in Ap- pendix F. The data was obtained by a search of accounting information kept in the Michigan State University Computer Laboratory. The charges were located by problem number. Only the data for 1977-1980 could be located. This data al- lowed the calculation of the average computer cost per stu- dent in EGR 270 over the past four years which was of im- mediate use in one of the cost models. 'There are two cost models. The batch model which is based on fixed cost per student, and the microcomputer model which is based on fixed and variable costs. Batch of stt HSC, : where and H Micro hddE‘ ‘ the t COSIS Plain first tions 111g e where 120 Batch Processing Mode Cost Model The cost per term, batch mode, was simply the number of students, N, multiplied by the historical student cost, HSC, as derived in Appendix F. CPTBATCH = N * HSC where, N = number of students required to use the batch mode and HSC = the historical student cost. Microcomputer Processing Mode Cost Model The cost per term, microcomputer mode, is the sum of the total equipment and maintenance costs, TEMC, plus labor costs. CPTMICRO = TEMC + L Determining these costs was fairly involved and is ex- plained as follows. TEMC, included several parameters that first had to be identified and, more importantly, their rela— tionships with each other had to be determined. The follow- ing equation accomplished both. TBMC = (N/NPS)*(CS+(YM*YU)) (YU*TYT’ where, N, was the number of students enrolled in the class. NPS, was the number of students designated to use one station or microcomputer graphics system. CS, was the cost per station, the hardware cost of one station. YM, YU, TY, 121 was the maintenance cost per station given as a yearly expense for the maintenance of the micro- computer graphics system. It can be the cost of a yearly service contract for the equipment that configures each station. It can be zero if main— tained by labor (see, L, labor cost). was the years of expected use. The variable was based on an assumption that the hardware would be in service for some predictable number of years. was the terms of use per year. For institutions on the quarter system, TY would equal three if utilized for the three quarters of the academic year. For institutions on semesters, TY would equal two. was the labor cost per term. This variable can be determined any number of ways. Labor costs can vary from zero to thousands. Generally, as the number of students that require access to the equipment increases, the need for labor will in- crease as well. For a single station serving a few students, it does not make sense to hire a full time attendant. However, a facility of three or more stations may need to be supervised. Ten or more stations might demand full time super- vision. In this case the person in charge of the facility might be able to perform the maintenance function, thus affecting the variable YM. TEMC ‘ per t can b Thus, subst divic requi numer 122 Hence, the cost model can be expanded by replacing the TBMC variable with the function of its components. The cost per term, microcomputer mode, was represented as: CPT = (N/NPS)*(CS+(YM*YU)) MICRO (Yu*TY) + L An example of how this model can predict cost per term can be illustrated by making the following assumptions: Variable Assumption N 60 students NPS 6 students CS 2000 dollars (hardware cost) YM 195 dollars per year YU 5 years TY 3 terms per year L 600 dollars per term Thus, CPTMICRO = TEMC + L = (N/NPS)*(CS+(YM*YQ))+ L (YU*TY) substituting = + 600 (60/6)*(2000+(195*5)) (5*35 1983.34 dollars per term To find the cost per student per term in this model d' . lVlde CPTMICRO by N. CPSMICRO = CPT = 1983.33 60 = 33.06 MICRO N Incidental to the cost model, the number of stations required can be determined by the first component of the model taker The n equi; less, C0515 cost facui Also: 123 Number of stations required = N/NPS, in this example, 60/6 or 10 stations are required. These models are exercised in Chapter V. The batch model is simple because all direct and indirect costs are taken into account when computer charges are established. The microcomputer model is based on several assumptions and equipment purchase prices that vary fairly often. Neverthe- less, these models will provide a reasonable handle on the costs associated with each processing mode. One final word about these models. The instruction cost is assumed to be the same in either mode. That is, the faculty person's salary would be the same in either mode. Also, all other costs are considered equal. r—a CHAPTER V RESULTS Introduction Sampling results demonstrate the effects of stratified random sampling. Since the study was conducted over two terms, the question of combining data had to be addressed. For each dependent variable the summary information is pre- sented in the following order: I) The restatement of the re- search hypothesis. 2) The translation of the research hy- pothesis into statistically testable form. 3) The tables showing the results of the statistical techniques. 4) The decision based on the presented evidence on whether or not to reject the null hypothesis. 5) Graphical representation of the results. Finally, a statement is made about cost relative to the cost models that were formulated. The models are exercised by changing the various parameters. Sampling Results In the treatment subsection of Chapter IV, the concept of proportionate random samples was introduced and explained. Proportionate random sampling insured a fully crossed and balanced design matrix. Cases representative of each grOUp type and major type were necessary for analysis. Although 124 125 the sample size for the microcomputer group had to be set at a maximum of eight students, the makeup of this group had to reflect the overall proportions of major types in each class. As the data presented in Tables Seven through Nine illus- trate, these prOportions were maintained as closely as pos- sible. When the data was combined (see Table Nine) both group sample prOportions were within two percent of the com— bined terms totals. These results are based only on students who completed the course. Students who drOpped out of the course did not provide complete data sets. Consequently these students were not counted in these and other results that follow. Combining Data The first concern prior to the analysis of data col- lected in this study was the issue as to whether or not the data from each term could be combined. The obvious reason for aggregating these data was to improve the analysis sta- tistically. Equally important was a determination of the reliability of the study. If the data could not be combined, then the replication aspect of the study would have been jeopardized. As was found, the data gathered during both terms of the study was able to be combined. Based on the assumption that the enrollment in EGR 270 represented a larger population of students interested in computer graphics, it was assumed that the students who en- rolled Winter and Spring Terms were independent samples of 126 whommz maze hocuo paw mpH< wcflaoocwwcm t macaw: oocoflom Housmeou paw Mawyoocfimcma o.ooa em o.ooa ca o.OOH w mpOmmz HH< 5.09 ea w.wo HH m.No m macaw: Hospo\ onoom NmH.o Hee.o wAN.m wH shoe wefludm . . N @3090 men.o ow mm.o- mom 0 «H H HVH.o omo.o wom.m «N EHoH HouaHz H @3090. .moam mm ous> .noam osHm> Hoehm >om cwoz mommu HHwH-N H , HHmerm m , ppm ppm H0 z mmmoom Hzm2m>mqu< OH mHm pcovaomonv HH.m .e .m .N mEopH momzomm -om umos umom 131 om¢.o mmw.H oom.HN wH .Eeob mcHemm eeH.o ow Ne.H- 00H.o ea.H N msoeo mHm.o nmm.m oom.o~ «N Show nouaHz H @5090 HoumHHm>oov UH.a .w .m .H; mEopH momcomm -om,MmoHon .noym mm ous>, .nomd osHa> pophm >om amoz momwu HHmH-N H HHmH-~ m ppm cum Ho 2 .umm mocmHHm> onoom, mqua accused -omv OH .N .H mEouH momcomm tom umoe umom ome.o omo.N mmw.w wH Shoe wcapmw wuu.o ow wN.o NNm.o Nm.H N adowo Hom.o ch.H ooo.m «N shoe Houcaz H msowu. monHHw>oov o .m .N. mEouH momcomw -om umouoem .nopm mm ous> .noem osHm> pogpm >09 amoz mommu . HHmhuN H HHmH-N m bum ppm H0 z .umm oocmem> @oHOOQ mZmemsm - moeumoHao< .m ohsmHm mmomne to minus one. The "one" is one standard deviation and :he data cannot be made to fit this scale exactly because of :he weighting factors involved in transforming the data (see factor analysis of scales in Appendix D). However, a sense >f relative magnitude can be obtained. The greater the ver- :ica1 distance above the mean line (zero line) indicates a rore positive attitude in regard to the attitude factor. The greater the vertical distance below the mean line indicates 1 less favorable attitude in regard to the attitude factor. 1 cell mean on or very near the mean line is indicative of an indecided, uncertain, or neutral attitude. As previously stated in the discussion of Figure Three, :he vertical displacement indicates a difference between major types relative to processing mode. The slant or SlOpe 3f the line increases with differences within major types >etween processing modes. The cell mean data can be found 1n \Ppendix H and Table Fifteen. Discussion and interpretation .5 reserved for Chapter VI. 146 mEHoH wocHnEou use meH waHHmm .meH Hochz .ogsmmoz moHnmmpo HoasmEou .mcwoz_ow3uHuu< .v opsmHm meowdels which represent whole University structures is stag- 1g. The models developed in Chapter IV and exercised are a microcosm by comparison, yet their purpose is .ar. Based on a set of assumptions it was possible to .ct the various costs relative to the several variables trameters contained within the model. A restatement of the models that were used and that W111 1e subject of the following analysis are: = * CPTBATCH HSC N 150 mEuoH vocHLEou can meH mcHumm .meH HoucHz .oHSmmoz.mEoum>w HousmEou .mnmoz.ow:uHup< .m oHSMHm 2022 mmmaoém + 1 1.. + $822 23me x 1 * 5.22 8qu not; 8on 5:5 2qu . oé- 1 .61? \1. E \ \ \\ x \ _ \ \ 1\ \ o . o l.—.. +|II|I /X m . o anz wH ll 2 <1- N ll 2 151 ere’CPTBATCH’ the cost per term utilizing batch mode pro- ssing is a function of the historical student cost times e number of students enrolled. CPT = (N/NPS)*(CS+(YM*YU)) MICRO (YU*TY) + L ere, CPT the cost per term utilizing microcomputer MICRO’ de processing is a function of the quantity represented by e numerator: N, the number of students enrolled divided by S, the number of persons designated to use each station, at quantity times the quantity, CS, hardware cost per sta- on plus the product of YM, yearly maintenance times YU, ars of utility. The numerator is divided by the denomina- r, which is the years of utility times TY, the number of rms per year. Finally, L, the labor cost added to quotient. .If the cost per student per term, CPS, is required, it simply the cost per term divided by N. CPS HSC*N _ HSC BATCH N — CPSMICRO = CPTMICRO N st Strategy The strategy that was employed when exercising the cost dels was to formulate a series of basic assumptions and nipulate a single parameter while the others were held at e assumed valueS. It should be noted that there are no ntrols over costs in the batch mode. The purpose of a cen- 31 batch facility is to provide services and the cost for 152 11 these services is reflected in a single statement of barges. Hence by loading the batch mode, the cost increases irectly proportionate to the number of users. The best stimate of the cost for these services in this mode was the istorical student cost. Since HSC was represented by a ixed cost of $55.68 per student per term, it, in turn, was re basis for comparison, i.e., all comparisons will be made 1 terms of cost per student per term. The following basic assumptions were made and, with the (ception of the parameter under consideration, were held >nstant during all computations. These values were: Variable Basic Assumption N (Number of students enrolled) 30 NPS (Number of students desired per station) 6 CS (Cost per station, hardware cost) 2,320 YM (Yearly maintenance contract cost per station) 195 YU (Years of utilization, estimate of the effective service life) 5 TY (Term per year of expected use) 3 600 L (Labor cost per term) It should be noted that these assumptions were realis- .c figures. Thirty students was the enrollment limit for 1R 270 over the past three years. Given a choice, six stu- ents per terminal is optimal. In terms of the number of :rminals required, five terminals would have been necessary 1 this course based on the first two assumptions. 153 Cost per station and yearly maintenance were actual figures. Cost per station was the purchase cost of the sys- tem that was used in the study. The dealer would have main- tained that equipment for a yearly maintenance contract cost as stated. Years of utilization was a compromise between the worst guess on equipment life of three years and the best guess of seven years. It was reasonable to expect five years of service from the hardware with proper maintenance. Terms of use per year assumed the institution was on the quarter system and the course was offered three times a year. The final assumption, labor cost, was based on the fact that five systems would have required some attendance. If these ma— chines were kept in a classroom or laboratory, their use could be supervised twenty hours a week by a student employee paid $3.00 per hour for ten weeks. Hence, the figure $600.00 per term was obtained. Cost Graphs Figures Six through Eleven represent the predictions in terms of cost per student per term based on the given set of assumptions. All but Figure Six apply only to the micro- computer model. Figure Six allows for a comparison between modes in terms of cost per student per term and the number 0f students enrolled. The point at which the plot of the microcomputer mode crosses the batch mode (a horizontal line by virtue of a fixed cost per student) represents the pOint where the costs were equal. Based on the assumption set, 154 ris point was about thirty three students. Any larger class Lze would have indicated a savings in terms of cost per stu- ant (all other variables held constant). When inspecting Figures six through Eleven, these no- Lons must be kept in mind: 1) all variables are the assumed rantities with the exception of the parameter under consid- 7ation. 2) The parameter under consideration was always the )scissa, the independent variable, and was located along the -axis. The units varied and were always given along with 1eir respective range. 3) The ordinate, or dependent vari- >le, was always given in cost per student per term (dollars) Long the Y-axis. 155 owe: comm :H mucovspm Ho Honesz %n 5905 Hon psousum Hem umoo Ho :omHthEou. .o opsmHm omH ooH mezmmnem om o r — — P b — L F — P — — . P — ION Hpmou oHanHm> , I use eoxHHO tees omqu v 4? b1 1L-1 n u b . IO? . Epou pom 11% I. #COVDvm H09 #mOU Humoo OomeO oOoE IUH .O oasmHm mm .OH oesmHm mm OH O O N o m e m N H O _ _ P H b _ H _ _ Jr “ I IOm Epoe yea I pcovspm l Hon umou mm@#37 38 38 938 938 I38 #38 39 581139 ‘Credit toward graduation will not be given for mathematic Major Code Engineering 00 Engineering — No Major Preference 90 Engineering for International Service Aggicultural Engineering 10 Agricultural Engineering Chemical Engineering 30 Chemical Engineering Civil 5 Sanitary Engineering 40 Civil Engineering Electrical Engineering 60 Electrical Engineering ,Mgghanical Engineering 70 Mechanical Engineering Metallurgy 80 Metallurgy Engineering Sciences 00 Engineering Sciences - No Major Preference 10 Computer Science 20 Systems Science 30 Materials Science 40 Mechanics Engineering Arts 50 Engineering Arts except in the Engineering Arts major. @At the present time, no new students who have this major. IThe number of students admitted as jun 'OPhomore engineering students to be recorded with the p Official Record. **Because employment opportunities cannot be gua Cooperative Engineering Education program IJY Effective: Source: Fall 1979 Memorandum from Dr. Jack B. Kinsinger, Registrar, dated July 30, 1979. AEBREVIATION ON RECORD CARD ENGR.-NO PREP. ENGR. FOR INT'L SERV. AGRICULTURAL ENGR. CHEMICAL ENGR. CIVIL ENGINEERING ELECTRICAL ENGR. MECHANICAL ENGR. METALLURGY ENGR.SCI.-NO PREP. COMPUTER SCIENCE SYSTEMS SCIENCE MATERIALS SCIENCE MECHANICS ENGINEERING ARTS iors to this major is limited. be limited. COLLEGE OF ENGINEERING 181 August 1979 DEGREE OFFERED (FréSo Only) B’s. & BOA. B.S. (FrbSo Only) (ErsSo Only) 3.5. 3.5. 3.5. 3.5. 3.5. Page 28 CREDITS REQUIRED FOR GRADUATION 225 180 180 180 180 180 180 180 180 180 180 3 courses below Mathematics 112 reached junior standing are being admitted to All Freshman and refix "PRE" on the student's ranteed for all applicants, enrollments in the Associate Provost, to Dr. Horace C. King, APPENDIX B COURSE OUTLINE SYLLABUS ASSIGNMENTS ONE THROUGH EIGHT ACHIEVEMENT SCORE SUMMARY DATA WINTER TERM 1981 ACHIEVEMENT SCORE SUMMARY DATA SPRING TERM 1981 DISTRIBUTION OF SCORES APPENDIX B COURSE OUTLINE Computer Graphics EGR 270 Instructor: Bill Kolomyjec 236 Engineering B1dg., 355-9577 Graphics Consultant: Dan Belfer - ECCO 269 Engineering Bldg. Texts: 1. Computer Graphics, Demel, Wilke, COppinger and Barr, Creative Publishing Company, 1979 (available at bookstore). ' 2. Userguide VOl. VII packet. Course Objectives: 1. Obtain a familiarity with computer graphics as another form of output from a digital computer. 2. Explore the requirements for a generation of two- dimensional figures and graphs by developing graphic algorithms. , 3. Explore the theory of systems for representing three-dimensional imagery through pictorial repre- sentations and algebraic functions. 4. Utilize existing graphic systems and software, i.e., plotter (CALCOMP). Basis for Grade: This course will be problem oriented. The final grade will be determined by computed average score derived from the suc- cessful completion of problems that are tO be solved graphi- cally both from the workbook and on the computer. Workbook assignments will comprise 20% Of the final grade. Programm- ing assignments will be Obtained from the best 51x out of seven computer graphic output problems for 60% of thegrade. There will be nO formal mid-term or final examination. How- ever, a special assignment given near the middle Of the term will have the value of two counted assignments and Will have a value Of 20% of the final grade. 182 183 Point breakdown for basis of grade: Max. Possible Points Homework (8-10 assignments from workbook) 50 Computer Graphic assignments (25 points each) 150 Special assignment 50 Raw final score 'TSU Course grade will be determined by curving (norm referencing) all raw final scores. Prerequisites: It will be assumed that students who are enrolled in this course have completed the required prerequisites. Students who have not will be at a distinct disadvantage. .Most prob- lems will be oriented toward mechanical (technical) drawing. The mathematics required will not be complex; however, a. good understanding Of algebra and trigonometry is essential. Format: Monday and Wednesday classes must be attended by all students. Fridays will be less formal, i.e., no new material Will be presented. If you need help, Fridays Will be extremely pro- ductive. However, since a fairly large number Of programm- ing assignments are expected tO be completed, we Will deSig- nate these days for dealing with the production system at the computer laboratory. See note. Note: This course has been modified and updated based on end of term evaluations by students from preVIous terms. There- fore, some modifications may have tO be made as the term pro- gresses. If this action needs to be taken, changes Willlle announced. Your suggestions and corrective critic15m Wi be welcome. Week 184 SYLLABUS ACtivity Course introduction, system re- qUirements. 2-dimensional Graphics 2 System graphic software drawing theory. Subprograms, plotting, rectangle subroutines and problems. Polygon subroutines and problems, circle and art subroutines and problems. Dashed line, graphic subroutines and problems. Sizing and scaling, Michigan map problem. 3-dimensiona1 Graphics 7 8 9 10 Orthographics. Pictorials, Oblique. Isometrics. Perspective. Assignment Acquire course materials, read 1-12 workbook. 13-20 workbook. 21-29 workbook. 30-37 workbook. 28-43 workbook. Handouts 44-53 workbook. 54-58 workbook. 59-61 workbook. 67-70 workbook. 185 EGR 270 - COMPUTER GRAPHICS Assignment One From Computer Graphics page 20 dO problems: Using Option ONU'I-P-LNNl—l DOWUOW Requirements: All output must be on a single page,for each 2 or 3 dimen- sional shape the following must be given: Name . _ . Parameters as spec1fied by the Option Solution (area or volume) Default Format: ................ New Page Area = 20.000 Rectangle: Height = 4.000 Width - 5.000 3 lines Etc. Numerical data must be significant tO three places. 20 points for correct solutions output using Scorin : default format. 1-5 extra points for crea printer formats. tive one page layout using various Due: April 13, 1981. Hand in program and output. 186 EGR Z70 - COMPUTER GRAPHICS Assignment Two From Computer Graphics page 28 dO problems 3, 5, and 6, include axes. Requirements: Follow instructions on sheet 28. Group A Three separate plots plus program listing. One grid square equals 25 pixcels. Group B In 21" Of plotter paper plot the three required drawings (align x-axis Of each plot on a horizontal line, but do not draw this line, and plot in numerical order left to right). One grid square equals 0.5". Use queue 0. Scoring: 25 points. Due: April 20, 1981. Hand in program and output. 187 EGR Z70 - COMPUTER GRAPHICS Assignment Three From Computer Graphics page 37, do problem 3, parts A and B. Requirements: In part A make the diagonal straight line segment between the two arcs 45°, adjust radius accordingly. Print out X,Y coordinate location Of: P0 (center of top arc), P (tOp Of sloping line), P (bottom of sloping line), and P (center Of bottom arc). Also Rl (radius Of top arc), and R2 radius Of bottom arc). Group A Two separate plots plus program listing. .One grid square equals 15 pixcels. Locations and radii given in integer numbers (round to nearest pixcel). Group B Plot on queue 0. Put both plots together (align XSaxes on same horizontal line). One grid square equals 0.5 .. Hand in plot and program printout. Locations and radii given in decimal inch to 3 places. Scoring: 25 points. Due: April 27, 1981. 188 EGR 270 - COMPUTER GRAPHICS Assignment Four From Computer Graphics page 43, do problems 1C, 2C, and 3C. Requirements: Follow instructions on sheet 43. All graphs must be drawn via separate subroutines from main program. . Group A Three separate plots run from the same program. Use one inch equals 30 pixcels as indicated by problems on sheet 43. Group B In less than 30” of paper plot the three required graphs (align the bottoms of the graphs on a continuous horizontal line - do not draw this line). Plot in order left to right. Use sizes as indicated by problems on sheet 43. Use queue 0. Scoring: 25 points. Due: May 4, 1981. Hand in program and output. 189 EGR 270 - COMPUTER GRAPHICS Assignment Five - State Map, Midterm Plot Phase I (10 points homework) 1. During the state lottery you will pick a state. Name of your state: 2. As home due May 11, 1981, gather the appropriate data and do the following: A. Plot on graph paper for digitization the out- line of your state. Indicate map origin and locate the 25 largest cities on this sketch with a "plus" symbol. B. Formulate a data base of the 25 largest cities, their populations and X and Y coordinates (relative to your map origin). Present this information in the following format: Col 1-16 Name of city NOTE: Use a coding Col 17-26 Population form or graph Col 27-31 X coordinate paper, print Col 32-36 Y coordinate legibly. C. Hand in both the digitizing sketch and data base. (Make sure you make a COpy for yourself) D. Cite the source of your data. Phase II (50 points) 1. Plot a digitized map of your state. Group A: This must fit on screen. This must fit on queue 0 output. (Final plot should be on queue 5.) with a "plus" or any other a circle in the given range scale the circles (ra- Group B: 2. Locate each given city center symbol and draw for the 9 largest Cities, dius) based on population. 190 Assignment Five Continued Group A: Radius range 3-20 pixcels. Write a "plus” drawing subroutine for a 5X5 pixcel plus. Group B: Radius range 0.25 - 1.00 inches. Use "CALL SYMBOL" for center symbols with a size of 0.2". 3. Plot the name of all cities and scale the height of the letters in the given ranges very near the centered symbol indicating each city's location. Group A: Use LETTERSUB height range between MULT = 0.5 (4 pixcels) and MULT = 2 (l6 pixcels). Adjust location to elimi- nate overlap. Group B: Use CALL SYMBOL. Height range between 0.125 and 0.75 inches. Angle parameter may be used to change plotting angle Of names. 4. Print out requirements. Printer output. A. The data base in alphabetical order. B. The data base sorted into descending order by population (name of city and pOpulation). C. The calculated radii for the population cir- cles, relate to given ranges. D. The calculated letter heights, relate to Organize this information (items B, C, read table. given ranges. and D) into an easily Print out and plot due May 18, 1981. 191 EGR 270 - COMPUTER GRAPHICS Assignment Six - Sheet 53-2, Option A 1. From the following list and algorithm, determine your object: Object Number Textbook Location Scale (Full size) 1 49-1 Group A: 1 grid square 2 49-2 equals 15 pixcels 3 49-3 Group B: 1 grid square 4 49-4 equals 1 inch. Algorithm: (Number of characters in your last name modulo 4) + l 2. Main program should draw two sets of orthographic views from the apprOpriate subroutines and follow this form: Initialize NOTE: Draw views Call Dinput (data input) subroutine full size and to Define initial scale the scale indicated Call front (scale) by the Option Call t0p (X space, scale) deSignation. Call Rside (Y space, scale) Redefine necessary parameters (move origin or clear screen) Call front (...) Call top (...) Call Rside (...) Terminate Group A Clear screen between multiview displays. Group B Put both multiview drawings on one plot, 1e to subroutines through COMMON. queue 0. 3. Data should be availab Due: May 20, 1981. (plot and printer output) 192 EGR 270 - COMPUTER GRAPHICS Assignment Seven - Parallel Projection Pictorials Using the same 3-D data base (and DINPUT subroutine) from asSIgnment 6, plot the following pictorials of your object Via oblique and isometric subroutines. Oblique l - Cavalier oblique with a receding angle of 240° Oblique 2 - Cabinet oblique with a receding angle of 60° Oblique Note: See sheet 57 Isometric l - Standard axes configuration. Isometric 2 - Substitute: Z for Y axis Y for X axis X for Z axis Isometric 3 - Substitute: -Z for X axis X for Z axis Y no change Isometric Note: See sheet 61, problem 2 for examples Of these. NOTES: 1. Scale requirements. Overall data bases should be scaled down as follows for these problems: Oblique l 70% Of full size ‘ Oblique 2 30% of full size Isometric l 66% of full size Isometric 2 50% of full size Isometric 3 33% of full size 2. Again, main program should do nothing more than locate origins and call subroutines With appro- priate parameters. . . Oblique parameters requ1red: receding angle, re- ceding axis scale factor, overall scale factor. Isometric parameters required: overall scale factor, axes assignments. 3. To improve visualization use circle or arc sub- routine to identify object origin by draWIng a 0.25 dia. circle around origin. (Group A, 8 pix- cel die.) 4. 'Group A DO this in 2 parts, 2 oblique drawings on one screen display, clear screen, draw 3 iso- metrics. Assignment Seven Continued Grou B DO all pictorals on single plot, adjust PLIMIT if necessary, use queue 0. Due: June 1, 1980. Plot and printer output. 194 EGR Z70 - COMPUTER GRAPHICS Assignment Eight - Perspective Array Using your previously assigned Object and DINPUT subroutine generate two arrays of translated (rotated) perspective pictorals. NOTE: Centers, i.e., XTR, YTR can be generated using sizing- scaling technique. Array Group A Group B 1 NX = 7, NY = 5 to fit screen NX = 7, NY = 5 to fit Within 2 NX = 8, NY = 6 NX a 9, NY = 7 12X12” plot area Also, let your draw loop indices JX, JY determine rotation angles as follows (note: radians). THETA = TWO PI*((JY-1)/(NY-1)) watch for mixed—mode PHI = TWO PI*((JX-1)/(NX-l)) Thus, each object will have a unique rotation (the excep- tions will be at the outer rows and columns Where 0 and 360° are the same angle). Use: XE = one-half width, YE = one-half height and appro- priate (you decide) values for ZE and ZDIST (move back distance). ZE and ZDIST may be the same or different for each array. pictoral array as shown in accompanying O timize ers ective p p p ld not overlap (too much). example. Views shou Due: June 5, 1981. UNOPTIMIZED PERSPECTIVE ARRAY, FTR = 0.2 :..o&-@?@@% Wgflflfigfig Wflfiflfimflfl 90 35! IBIIB &@@@@&¢ 0° 240° 300° 360° OPTIMIZED PERSPECTIVE ARRAY, FTR= 0. 25 196 m.m 5O O.©HN Om me n.5m Om ON O.m NO O.opm .cwm< HmHOH xpozoEom xhoz OHUHZ 3mm mo 55m HwHuopm mEmHQOHQ wmuamHoz -050: meH ZKmH mmHsz . m<223m mmoum HZmZm>mHEU< uanSHm .Hommm uxmcv mmhoom we cOHuangpmHm ommO .ogoum chHw 3mg u Hm.O+0.0«ZDmOHzH HmmmucH. umohmoc ow wmo OCDOH .O.O >9 :OHumuHHmHuHDE kn OOH-O ow OmN-O Eopm mmeumx 197 .Hmuow xhoszo: OmpanoB n O\m x prou MhozoEOEO 0.0 «O O.mmN Om mmH 0.0¢ OO OH O.v «O N.OON OO OOH N.H¢ ON NH 0.0 NO 0.0HN Om ONH O.m< ON OH O.m mO O.HHN OO OOH 0.0m Oh mH m.m NO 0.0NN Om OOH O.Nv mm «H m.m NO 0.0NN Om NmH O.NO OO OH O.w mO O.NmN Om OOH O.m< On NH O.v OO H.m¢N Om OOH H.O¢ mO HH O.N ON O.NOH Om OOH O.Nm Om OH m.m OO N.mNN Om OmH n.Om OO O O.m OO 0.00N Om HNH O.Nm OO O 3090 :uumm m.m mO O.NmN Om OOH O.mv NO 5 O.¢ NO m.HvN Om mvH N.Oq «O O O.m OO N.m~O .cOm< HOHOH Hwoszo: Mgoz OHUHZ 3mm no 55m HmHuomm mEoHOOHO OOpHOHoz -oEo: pamwzym HOOH 2mmh oszmm I m<223m mmoom Hzm2m>mHIU< 198 DISTRIBUTION OF SCORES Winter 1981 Grade Range Frequency 4.0 100-92 5 3.5 91-87 7 3.0 86-82 6 2.5 81-73 4 2.0 72-64 1 1.5 63-54 1 Spring 1981 Grade Range Frequency 4.0 100-94 5 3.5 93-89 5 3.0 88—84 S 2.5 83—76 1 2.0 75-68 2 APPENDIX C REQUEST FOR CONFIDENTIAL INFORMATION REPLY FROM OFFICE OF THE PROVOST MEMORANDUM TO UNIVERSITY COMMITTEE ON RESEARCH INVOLVING HUMAN SUBJECTS (UCRIHS) REPLY FROM UCRIHS _ 1‘ “ '1 APPENDIX C ~ REQUEST FOR CONFIDENTIAL INFORMATION i ON INDIVIDUAL STUDENTS j'i-on Requesting Information: William J. Kolomyjec 1e: ’Instructor \itution: Michigan State University ,rmation Required; Prerequisite course final grades for EGR 270 students W'8l and SP'81. EGR 270 prerequisites are: EGR 160, EGR 161, EGR 162, CPS 110, CPS 120 ior LP: 500. arch Purpose;Doctora1 thesis in higher ed. & admin. entitled: A Comparison of the of Microcompu er rap 1c ys ems ctory *se in Computer Graphics. Specifically, I need these grades in Order to do an [3515 of covar1ance to meaSUFe—EEHTEV -"- er of Students Involved: w '81 24; Sp '81 approximately 27. Class lists attached. archers using data from student records are reminded of their responsibility totect students by adhering to the following regulations: [- Research reports (whether preliminary or final) will not identify students by name or report data in a form or so extensively that knowledgeable persons may be able to identify individuals. L- Neither the student records nor personall) identifiable data therefrom will be released to other persons or agencies in the institution, nor will the records be used for any other purpose tixan the one indicated above. 3. The student records and any personally identifiable data there— from will be destroyed as soon as the research purpose has been fulfilled. I. A copy of research findings will be sent to: Director of Institutional Research Michigan State University 320 Administration Building East Lansing, MI 48824 'irm that we will adhere to the four provisions stated above: Signmi: (A)b1141114LO ObfiUZLG‘~¢~4\F"> Person requc‘ting informatiom L}. [ 'SLg' Ccnfirmod WW : dfififlrfif*¥?tfiééefiF I 94' R Ju r2. €110 FCSSC‘fl" y W (’a mm 1166- W O L, ,,L ._,I,, _ 7, ,,7< 199 osTFOR .. cabwn ISTRfiV 5! TM” GENcoA? E OEMI IGAN STATE UNIVERSITY 200 F THE PROVOST EAST LANSING ° MICHIGAN ° 48824 RATION BUILDING May 6, 1981 COMMITTEE ON RELEASE OF CONFIDENTIAL INFORMATION and APPROVAL OF QUESTIONNAIRES Dorothy Arata, Chairperson equest described below and attached to this Memorandum has been / / approved denied by the Committee. Reques;er: William J. Kolomyjec Dated: . February 15, 1981 Subject° Prerequisite course final grades for EGR 270 students W'81 ard SP'81. Need grades in order to do an analysis of covariance to measure achievement relative to processing mode. /7’ " f/ iii/”é @A:/ For the Committee ts: The committee is unanimous in its reluctance to approve the use of Lential information from student records without the prior approval of tudents. It is felt that this is personal dissertation research and not uired University research. The Committee suggests that you seek signed see from each individual student or ask the Student to furnish the data. 201 13, 1981 ANDUM University Committee for Research Involving Human Subjects William J. Kolomyjec QCT: Request for Information on Student Grades Relative to Doctoral Research Proposal 1 instructor in Metallurgy, Mechanics and Material Science, College of 1eering and a doctoral candidate in the College Of Education, I am working dissertation for a degree in Higher Education and Administration. I am 19 a comparison between the use of a microprocessor graphics system versus :entral batch facility in an introductory undergraduate course on computer iics, EGR 270. I propose to divide the students enrolled in this class into groups based on student type and from these groups form a sample group for purpose Of investigating differences between these groups; the sample group 9 the microprocessor with graphics capabilities, the majority of the class 9 present practice on the central batch facility. The comparison will be ue experiment and will involve pretesting and posttesting student attitudes a l2 question attitude survey as well as using grades in prerequisite ses to measure student achievement. Analysis will be obtained by an analysis variance on this data. requirement for the subject population is any one enrolled in the course 270. The students in this course were chosen because EGR 270 is designed e the first or introductory course in computer graphics. Since we are rested in evaluating the technology of microprocessors for graphics instruc- , this is a logical choice. is no potential risk to the student in physical, psychological, social, . economic or other terms. All students have knowingly enrolled in EGR for the purpose of obtaining computer graphics instruction. ata collected will be strictly confidential and aggregated so that nO 1dual student's identity will be revealed. benefits to be gained by the individual subject is what we seek to determine. re evaluating delivery mode in a course that utilizes computer hardware. ddition to evaluating the effect on students in terms of differences in evements and attitudes, we are also looking at coSt of instruction. Our 1al model suggests that as we load each mode with more students, the micro- essor mode lowers the cost of instruction per student over a certain time e. The benefits to departments and colleges in these terms is obvious. apolating that to society one might say that to those who support public itutions would certainly be in favor of getting better quality instruction lower cost. 202 :onsent procedure is a request by the instructor Of the student to participate 1e attitude survey on a voluntary basis. At the beginning of the term, the ants are asked to take part in the experiment by filling out a questionnaire. / student has the right to refuse to participate, this same procedure is awed at the end of the course. armal consent form will be utilized. The attached statement is read out before the students fill out the questionnaire. ched are documents to support this request. k you for consideration of this request. chments: 203 13, 198l ANDUM University Committee for Research Involving Human Subjects NO T. Bell, Ph.D. Dissertation Director Research Proposal for w. J. Kolomyjec {e reviewed the research proposal for William J. Kolomyjec's doctoral TS and hereby state that it meets with my full approval. 204 ‘IIGAN STATE UNIVERSITY SITY COMMITTEE ON RESEARCH INVOLVING EAST LANSING ' MICHIGAN ' 48824 SUBJECTS (UCRIHS) TINISTRATION BUILDING JW6 May 5, 198] dilliam J. Kolomyjec llurgy, Mechanics, 8 Materials Science Engineering Building Mr. Kolomyjec: Subject: Proposal Entitled, “A Comparison of the Use of Microcomputer Graphic Systems Versus a Central Batch Facility in an Introductory Undergraduate Course on Computer Graphics” ' above referenced project was recently submitted for review to the UCRIHS. 'e pleased to advise that the rights and welfare of the human subjects 3r to be adequately protected and the Committee, therefore. approved this act at its meeting on May 4, l981- In approving this project, however, as strongly recommended that a statement ensuring the students that no 5 or identifying pieces of information are reqU35ted3 ParthEPBtion is y anonymous. acts involving the use of human subjects must be reviewed at least annually. >u plan to continue this project beyond one year, please make provisions bbtaining appropriate UCRIHS approval prior to the anniversary date noted . you for bringing this project to our attention. If we can be of any '6 help, please do not hesitate to let us know. Sincerely, Henry E. Bredeck Chairman, UCRIHS TITS Dr. Norman T. Bell APPENDIX D QUESTIONNAIRE STATEMENT QUESTIONNAIRE PRELIMINARY VERSIONS, FORM 1 (RED VERSION) QUESTIONNAIRE PRELIMINARY VERSIONS, FORM 2 (BLUE VERSION) RESPONSE DATA, FORM 1 (RED VERSION) RESPONSE DATA, FORM 2 (BLUE VERSION) COMBINED PRELIMINARY QUESTIONNAIRE (PRETEST VERSION) RESPONSE DATA, COMBINED FORM IMPROVING RELIABILITY OF SCALES FACTOR ANALYSIS OF SCALES COMBINED QUESTIONNAIRE (POST TEST VERSION) EXPERIMENTAL DATA APPENDIX D QUESTIONNAIRE STATEMENT This is a study that will allow us to measure student atti- tudes towards the possible introduction of new techniques in an introductory course in computer graphics. We are asking for your voluntary participation. If you do not wish to partake in this attitude survey simply do not complete the distributed questionnaire. Your responses and identity will be completely confidential. It is important that you read the questions carefully and indicate the response that comes closest to your Opinion or feeling stated in the question. If you have no opinion or cannot decide a neutral or undecided choice is provided. Use pencil only (#2 or softer) to mark the provided scoring sheet. The response codes are: 1 or A STRONGLY AGREE 2 or B AGREE 3 or C UNCERTAIN 4 or D DISAGREE E S or STRONGLY DISAGREE Definition of Computer Graphics: the formulation, display or processing of visual imagery by way of a digital computer and peripheral devices having graphic capabilities. 205 206 Use Red Response Sheet ponse Code: 1 or A STRONGLY AGREE 2 or B AGREE 3 or C UNCERTAIN 4 or D DISAGREE 5 or E STRONGLY DISAGREE Personally, the idea of using a computer to generate visual imagery appeals to me. The present practice of generating computer plots by punched card decks (batch mode) seems like an adequate technique in an intro- ductory computer graphics course. I would enroll in an introductory course in computer graphics only because it was required for my major. Large central computer facilities like the system presently in use at MSU are best suited for computer graphics. Ideally, computer graphics should be an immediate process where programs are executed in real-time and visual output is plotted or dispalyed at once. An introductory computer graphics course should use FORTRAN and a more general purpose programming language like BASIC should not be considered. Graphic algorithms, such as the instructions for drawing a circle, are generally simple and easy to understand. Writing computer programs to generate computer imagery seems more like it would be fun than work. Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM) are application areas where knowledge of computer graphics is essential. The thought of programming a computer to do computer graphics scares me to death. A microcomputer (personal computer) with graphic capabilities that has the ability to execute programs immediately and generate computer imagery in real—time seem like a more appropriate approach for learning in an introductory course in computer graphics. Given the choice of using a large computer system with its greater capacity, variety of output but lesser convenience versus a micro- computer system with graphic capabilities whicn are limited in terms of capacity and variety but, is extremely convenient to use, I would prefer the larger system to learn computer graphics. OF QUESTIONNAIRE. PLEASE CHECK THAT YOUR RESPONSES ARE CLEARLY MARKED. —+ 207 Use Blue Response Sheet sponse Code: l or A STRONGLY AGREE 2 or B AGREE 3 or C UNCERTAIN 4 or D DISAGREE 5 or E STRONGLY DISAGREE It would be very interesting to generate graphics using a computer that has this capability. Keypunching and submitting these cards as program decks to the computer in the batch mode is appropriate for the instruction of computer graphics in an introductory course. I would enroll in an introductory computer graphics course if I had the option. The nature of generating computer graphics dictates the need for large central computer facilities with sophisticated graphic output devices. Having to wait several hours for processing programs and return of graphic output is inherent to the computer system and can be tolerated. If a more general purpose programming language was used, say BASIC compared to FORTRAN, I feel it would improve an introductory computer graphics course. Generating complex computer imagery through the use of graphic algorithms, subroutines which perform specific graphic functions, would involve a better than average understanding of computer programming. I would rather use computer graphic programs than write them. Computer graphic applications are limited to engineering and science. I would have no fear programming a computer to do computer graphics. The immediate execution of programs and display of computer imagery possible with a microcomputer (personal computer) seems like an ideal approach for learning in an introductory course in computer graphics. If I could, I would choose a computer system with speed and convenience over greater capacity and capabilities for problem solving and learning in an introductory course in computer graphics. OF QUESTIONNAIRE. PLEASE CHECK THAT YOUR RESPONSES ARE CLEARLY MARKED. 208 RESPONSE DATA FORM 1 - RED VERSION N = 89 RECODED Relative Frequency (%) Summary Statistics Item SD D N A SA X SD VAR. 1 — 1.1 7.9 52.8 38.2 4.281 0.657 0.432 2 5.6 32.6 30.3 27.0 4.5 2.921 1.003 1.005 3 9.0 24.7 21.3 33.7 11.2 3.135 1.179 1.391 4 2.2 12.4 67.4 15.7 2.2 3.034 0.682 0.465 5 - 2.2 22.5 44.9 30.3 4.034 0.790 0.624 6 13.5 23.6 40.4 16.9 5.6 2.775 1.063 1.131 7 - 3.4 59.6 30.3 6.7 3.404 0.669 0.448 8 2.2 15.7 18.0 50.6 13.5 3.573 0.987 0.975 9 1.1 1.1 29.2 41.6 27.0 3.921 0.842 0.710 10 2.2 5.6 6.7 50.6 34.8 4.101 0.918 0.842 11 - 6.7 31.5 42.7 19.1 3.742 0.846 0.717 12 2.2 29.2 29.2 33.7 5.6 3.112 0.970 0.942 PEARSON CORRELATION COEFFICIENTS EACH ITEM WITH TOTAL (INTERITEM CORRELATION) N = 89 p P 1 0.3900 0.001 2 0.3453 0.001 3 0.5713 0.001 4 0.3828 0.001 5 0.5061 0.001 6 0.1271 0.118 7 0.3040 0.002 8 0.4087 0.001 9 0.3173 0.001 10 0.5269 0.001 11 0.4686 0.001 12 0.3607 0.001 {ELIABILITY N OF CASES = 89, N OF ITEMS = 12 Ironbach's Alpha = 0.48551 209 RESPONSE DATA FORM 2 - BLUE VERSION N = 95 RECODED Relative Frequency (%) Summary Statistics Item SD D .N A SA X SD VAR. 1 - 1.1 7.4 45.3 46.3 4.368 0.669 0.488 2 9.5 40.0 28.4 10.5 11.6 2.747 1.139 1.297 3 3.2 9.5 38.9 28.4 20.0 3.526 1.019. 1.039 4 13.7 31.6 42.1 11.6 1.1 2.547 0.908 0.825 5 4.2 22.1 15.8 35.8 22.1 3.495 1.184 1.402 6 2.1 18.9 34.7 26.3 17.9 3.389 1.055 1.113 7 15.8 47.4 32.6 3.2 1.1 2.263 0.802 0.643 8 20.0 35.8 30.5 11.6 2.1 2.400 1.004 1.009 9 2.1 5.3 18.9 48.4 25.3 3.895 0.916 0.840 10 2.1 12.6 28.4 40.0 16.8 3.568 0.986 0.971 11 1.1 5.3 26.3 42.1 25.3 3.853 0.899 0.808 12 7.4 15.8 24.2 33.7 18.9 3.411 1.180 1.394 PEARSON CORRELATION COEFFICIENTS EACH ITEM WITH TOTAL (INTERITEM CORRELATION) N = 95 p P 1 0.2316 0.012 2 0.5872 0.001 3 0.2285 0.013 4 0.1964 0.028 5 0.3194 0.001 6 0.1936 0.030 7 0.2387 0.010 8 0.2364 0.011 9 0.2944 0.022 10 0.4355 0.001 11 0.4599 0.001 12 0.2830 0.003 RELIABILITY N OF CASES = 95, N OF ITEMS = 12 Cronbach's Alpha = 0.14779 210 sponse Code: l or A STRONGLY AGREE 2 or B AGREE 3 or C UNCERTAIN 4 or D DISAGREE 5 or E STRONGLY DISAGREE Personally, the idea of using a computer to generate visual imagery appeals to me. Keypunching and submitting these cards as program decks to the computer in the batch mode is appropriate for the instruction of computer graphics in an introductory course. I would enroll in an introductory course in computer graphics only because it was required for my major. Large central computer facilities like the system presently in use at MSU are best suited for computer graphics. Having to wait several hours for processing programs and return Of graphic output is inherent to the computer system and can be tolerated. If a more general purpose programming language was used, say BASIC compared to FORTRAN, I feel it would improve an introductory computer graphics course. Generating complex computer imagery through the use of graphic algorithms, subroutines which perform specific graphic functions, would involve a better than average understanding of computer programming. Writing computer programs to generate computer imagery seems more like it would be fun than work. Computer graphic applications are limited to engineering and science. I would have no fear programming a computer to do computer graphics. The immediate execution of programs and diSplay of computer imagery possible with a microcomputer (personal computer) seems like an ideal approach for learning in an introductory course in computer graphics. If I could, I would choose a computer system with speed and convenience over greater capacity and capabilities for problem solving and learning in an introductory course in computer graphics. OF QUESTIONNAIRE. PLEASE CHECK THAT YOUR RESPONSES ARE CLEARLY MARKED. PRETEST 211 RESPONSE DATA FORM 3 - COMBINED FORM N = 80 RECODED Relative Frequency (%) Summary Statistics Item SD D N A SA X SD VAR. 1 — 3.7 6.3 55.0 35.0 4.212 0.724 0.524 2 6.3 35.0 27.5 20.0 11.2 2.950 1.124 1.263 3 8.8 18.8 22.5 40.0 10.0 3.237 1.139 1.297 4 - 16.2 58.7 17.5 7.5 3.162 0.787- 0.619 5 2.5 22.5 15.0 36.2 23.8 3.563 1.157 1.328 6 6.3 18.8 31.3 27.5 16.2 3.287 1.138 1.296 7 11.2 52.5 21.2 13.7 1.2 2.412 0.910 0.828 8 12.5 30.0 21.2 25.0 11.2 2.925 1.230 1.513 9 - 12.5 20.0 40.0 27.5 3.825 0.978 0.958 10 2.5 8.8 25.0 47.5 16.2 3.662 0.941 0.885 11 1.2 1.2 15.0 51.3 31.3 4.100 0.789 0.632 12 2.5 13.7 17.5 30.0 36.2 3.837 1.141 1.302 PEARSON CORRELATION COEFFICIENTS EACH ITEM WITH TOTAL (INTERITEM CORRELATION) N = 80 p P 1 0.4795 0.001 2 0.4581 0.001 3 0.5529 0.001 4 0.2321 0.019 5 0.5285 0.001 6 0.4121 0.001 7 0.1886 0.047 8 0.3577 0.001 9 0.5184 0.001 10 0.3890 0.001 11 0.3673 0.001 12 0.5702. 0.001 RELIABILITY N OF CASES = 80, N OF ITEMS = 12 Cronbach's Alpha = 0.58492 212 IMPROVING THE RELIABILITY OF SCALES COMBINED DATA VARIABLES = 11 to I12 SCALE (WHOLE) = 11 to I12 Item Alpha if Item Deleted 0.54808 0.55959 0.53242 0.59102 0.54049 0.57291 0.60624 0 0 0 0 0 8 .59446 .53922 .56850 .56815 .52704 0 'N OF ITEMS = 12 1...: OKDOONO‘U'l-D-LNNH l—‘H NH ALPHA = 0.58492 N OF CASES SCALE (GRAPHICS) = 11, I3, 17, 18, 19, 110 1 0.48651 3 0.41852 7 0.59935 8 0.48052 9 0.50815 10 0.49445 ALPHA = 0.54882 N OF CASES = 80 N OF ITEMS = 6 SCALE (SYSTEM) = 12, I4, 15, I6, 111, 112 2 0.54357 4 0.58569 5 0.41549 6 0.52520 11 0.58252 12 0.49338 ALPHA = 0.57641 N OF CASES = 80 N OF ITEMS = 6 213 VARIABLES = 11, I3, 17, I8, 19, 110 SCALE (GRAPHIX l) = 11, I3, 18, I9, 110 Item Alpha if Item Deleted 1 0.51152 3 0.46924 8 0.58093 9 0.57609 10 0.57602 *ALPHA = 0.59935 N OF CASES = 80 N OF ITEMS VARIABLES = 12, I4, 15, 16, 111, 112 SCALE (SYSTEM 1) = 12, 15, I6, 112 2 0.55423 5 0.41445 6 0.54072 12 0.61666 ALPHA = 0.60743 N OF CASES = 80 N OF ITEMS VARIABLES = 12, 15, 16, 112 SCALE (SYSTEM 2) = 12, 15, I6 2 0.52534 5 0.45945 6 0.56296 *ALPHA = 0 61666 N OF CASES = 80 N OF ITEMS *Best possible - deleting any further items does improve the reliability (alpha) ll U1 II p ll LN not 214 FACTOR ANALYSIS OF SCALES - SUMMARY STATISTICS Graphix 1 - Attitude Towards Computer Graphics Vari- Commun- Eigen- Pct of Cum. able ality Factor value Var Pct. 1 0.41461 1 1.41367 75.3 75.3 3 0.45268 2 0.46377 24.7 100.0 8 0.28435 9 0.59593 10 0.12988 Factor Score Coefficients Factor 1 Factor 2 1 0.34796 0.04138 3 0.38936 0.07782 8 0.29614 -0.06357 9 -0.13832 0.71215 0 0.04826 0.10516 System 2 - Attitude Towards Computer Systems Vari- Commun- Eigen- Pct of Cum. able ality Factor .value Var Pct. 2 0.32948 1 1.06646 100.0 100.0 5 0.46511 6 0.27186 Factor Score Coefficients Factor 1 Factor 2 _OHnbo 0.31329 0.46587 0.26198‘_ 215 sponse Code: 1 or A STRONGLY AGREE 2 or B AGREE 3 or C UNCERTAIN 4 or D DISAGREE S or E STRONGLY DISAGREE Having to wait several hours for processing programs and return of graphic output is inherent to the computer system and can be tolerated. Personally, the idea of Using a computer to generate visual imagery appeals to me. If I could, I would choose a computer system with speed and convenience over greater capacity and capabilities for problem solving and learning in an introductory course in computer graphics. . The immediate execution of programs and display of computer imagery possible with a microcomputer (personal computer) seems like an ideal approach for learning in an introductory course in computer graphics. . Writing computer programs to generate computer imagery seems more like it would be fun than work. .. I would enroll in an introductory course in computer graphics only because it was required for my major. If a more general purpose programming language was used, say BASIC compared to FORTRAN, I feel it would improve an introductory computer graphics course. 5. Large central cOmputer facilities like the system presently in use at MSU are best suited for computer graphics. 1. I would have no fear programming a computer to do computer graphics. )- Keypunching and submitting these cards as program decks to the computfir in the batch mode is appropriate for the instruction of computer grap 1C8 in an introductory course. L. Computer graphic applications are limited to engineering and science. 7 Generating complex computer imagery through the use of grighic alg:r:thms, subroutines which perform specific graphic functions, wou invo better than average understanding of computer programming. I i E . ND 0F QUESTIONNAIRE. PLEASE CHECK THAT YOUR RESPONSES ARE CLEARLY MARK D POST TEST 216 EXPERIMENTAL DATA rLrLC.rLt_t.r.c.rLrLC_C.r.-LCrLrLrLC.B_:L§_Lrttcfl([taanl.rLr.r—.ILCC...C.rLrLfLr: 7 Id? 1..."... 3.150. 1.r¢4-_1 7c? liar/1 r(7..7 er 1.617 1 1.01»r(7x7.h~r(r£a7 1caa~a_7.. 1111. .1111111111111111111105.72?20 222229 27.22? d.1.1.2222?31.26-. 794.131.121.10:.2?1? COMMENT: THIS ROUTINE WILL SHUFFLE ANY ARRAY OF N NUMBERS AND PRINT THE RESULTS 1 . COMMENT: THE ARRAY ID CONTAINS N ITEMS TO BE ] SHUFFLED. PASSED TO SUBROUTINE THROUGH COMMON (UNNECESSARY IN BASIC) 1 COMMON ID(N) _ I [COMMENT: THE ARRAY NS CONTAINS A SEQUENCE OF INTEGER 1 THROUGH N . I L‘DIMENSION NS(N) _ I , COMMENT: THE FOLLOWING LOOP BUILDS THE SEQUENCE ARRAY NS I _ \ J — 1 TO N [as I NS(J) = J l [ COMMENT: DEFINE SHUFFLE LOOP INDEX N1 217 218 COMMENT: SHUFFLE LOOP I [————4- J = 1 TO N1_j>———~<::> _ I LCDMMENT: USE RANDOM FUNCTION TO RANDOMLY OBTAIN A NUMBER ,K, IN THE RANGE J THROUGH N A I L IN BASIC: K INT(RND(1)*(N-J+1))+J) IN FORTRAN: K IFIX(RANF(0)*(N-J+l))+J) I COMMENT: USE TEMPORARY VARIABLE L TO SWITCH THE CONTENTS OF THE Kth LOCATION IN THE SEQUENCE ARRAY WITH THE CONTENTS OF THE Jth LOCATION ____. .___l .4 l__l I L = NS(J) NS(J) = NS(K) NS(K) = L 9 COMMENT: USE SHUFFLED NUMBER ARRAY AS THE SUBSCRIPTS TO REORDER OR "SHUFFLE" ORIGINAL ARRAY OF NUMBER AS IN THIS PRINT LOOP I J = 1 TO NH RETURN) ___ I PRINT 1D(NS(J))/7 APPENDIX F HISTORICAL COST DATA EGR 270 219 .zwoumsonma HousmEoo oz“ kn pmou Hmpou Eomm mumou woupoam can Mopqflhm oomwowmmwmfiw op newcoEOHmEH mm: osswouosm mcwucsouom oposmmom wow 3o: m U .EhOp mafia umoo swan ocu wow pcsooum mucoEcMMmmw o>fioowsou:w Hmpo>om was omssou may op Hoflam pcoemoao>ow omssou Hmwunmpmnsw n .wmox mvsouow mcfipcsouow so: woNHHHp: opoz mosowOUOMQ wcfiumou wonpflo: when: moowfi EmHV zpflaflowm ampsmaou Hoapocw :0 Doom one: mucoecmflmmm may Mammm wo.mmm n manuasopmflsv ucoosum pom umou NOH owmfi-aamfi Haemasum mo Hones: H6069 o~.mao.ma u owmfi-kaafi mamou soasaaou Hauoe mmhsou mm Hm ON mm GEDOHSEOU on: mucoonpw mo 4. mm.ommH mm.maafi .aN.HeA ow.ama.fi Hapoe Um~.wwa a\: a\: 6\: mamaaao .ea<. mm.HsmH mbm.maafi maN.HsA ow.Hma.H Hauoanam a~.aa~ .mN.HHo -am.ma~ om.m~m ON\E-HN\m am.kmm Hk.smw mm.eam mfi.ook om\m-HN\a Ha.awfi mH.mNm ae.HNH mo.oaH oa\a-HN\m -o- ma.mo~ -o- -o-. ONxm-HN\N . wofisom whouom. mcfihoocflmam memu mo>m .HumcH , 22: mo :woq mHm\zzz mcflpoocfiwcm ammo wafisomaomm mooofimo aomaws omoeamo omaawo sonesz soflnosa mcflhmm mcflpmm wcflamm wnwwmm Esme omafi akafi mama Rama A66» mom