Proceedings of NECC/2 National Educational Computing Conference 1980 (Norfolk, Virginia, June 23-25, 1980)

DOCUMENT RESUME

ED 194 060 IR 008 912

AUTHOR Harris, Diana, Ed.: Ccllison, Beth, Ed. TITLE Proceedings of NECC/2 National Educational Computing Conference 1980 (Norfolk, Virginia, June 23-25, 1980). INSTITUTION Iowa Univ., Iowa City. Computer Center. PEPORT NO IsBN-0-937114-00-6 PUB DATE Jun 80 NOTE 306p.

EDRS PRICE MFOI Plus Postage. PC Not Available from EDRS. DESCRIPTORS *Computer Assisted Instruction: Computer Assisted Testing: Computer Oriented Programs: Ccnputer Programs: *Computers: *Computer Science: Elementary Secondary Education: Games: Higher Education: Humanistic Education; Learning Laboratcries: Material Development: Mathematics Instruction: *Microcomputers: Science Education IDENTIFIERS *Computer Literacy

ABSTRACT This proceedings, which includes 52 papers and abstracts of 13 invited and nine tutorial sessions, provides an overview cf the current status of computer usage in education and offers substantive forecasts for academic computing. Papers are presented under the following headings: Business -- Economics, Tools and Techniques for Instruction, Computerd in Humanistic Studies, Computer Literacy, Science and Engineering, structured Programming, ACM Elementary and Secondary Schools Subcommittee, Computer Science Education, Integrating Computing into K-12 Curriculum, Mathematics, Testing-Placement, Pre - College' Instructional Materials, Minority Institutions--ECM/, Computer Laboratories in Education, Computer Gt.mes in Instruction, and Computing Curricula. Abstracts are provided for 13 invited sessions dealing with such topics as microcomputers in education: research in microcomputer uses: personal computing: educational ccmputing: past, present, and future: the Open University: CAUSE program and projects: funding academic computing programs: computer-based resource sharing: computers and instruction: improving utilization cf 2-year college computer centers: teaching computer ethics: data sets available from the federal govenment: and MIS education, as well as nine tutorial sessions designed to provide attendees with the opportunity to expand their appreciation of and involvement in educational computing. (CHC)

*********************************************************************** * Reproductions supplied by EDRS are the best that can be made * * ircm the original document. * *********************************************************************** U S DEPARTMENT OF NEACTN. EDUCATIONE *waits NATIONACINSTITUTE OF EDUCATION THIS DOCUMENT NAS owl REPRO- ovceo exam., AS RECEIVED FROM THE PERSON OR ORGANIZATION ORIGIN. MING IT POINTS of view OR OPINIONS STATED 00 NOT NECESSARILY 'TERRE set:TORSION. NATIONAL INSTITUT ECF EDUCATION POSIT/04 OR POLICY

,Proceedings of NECC/2 National Educational Computing Conference 1980

Edited by Diana Harris Beth Collison

Hosted by Christopher Newport College Newport News, Virginia

"PERMISSION TO REPRODUCE THIS Held et MATERIAL INMICROFICHE ONLY Holiday Inn/Scope HAS BEEN'ORANTED BY Norfolk, Virginia Ted Sjoerdeme

23, 24, 25 June 1980 TO THE EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC).' International Standard Book Humbert 0-937114-00-d Copyright19Sgs SZCC Published by The University of Iowa Wag Computing Center Iowa City. Iowa 52242 for NAtIonal aluestIonal Computing Conference II guns 1980

Cover dealgn by Madeline WIn4auer 1

NECC 1980 Steering Committee

Alfred Bola University of California-Irvine Bobby Brown University of Iowa William S. Dorn University of Denver Karen Duncan Gerald L. Engel Christopher Newport Co.iege Norman Gibbs College of William and Mary John W. Hamblen University of Missouri-Rolla Paul Hazen Johns Hopkins University Harry Hodges Michigan State University Lawrence A. John University of Dayton Sister Mary Kenneth Keller Clark* College William E. Knabe University of Iowa David R. Kniefel New Jersey Computer Network Doris K. Lidtke Towson State University Elisabeth Little EDUCOM/EDUNET Sister Patricia Marshall Xavier University of Louisiana Richard W. Pogue Medical College of Georgia James Poirot North Texas State University Nancy Roberts Lesley College Theodore J. OPerdeme University of Iowa David L. Stonehill University of Rochester 'David B. Thomas University of Iowa

General Chairman: Gerald Engel Program Co- chairmen: Richard H. Meting, Dori:: K. Lidtke Local Arrangements Committee: Robert Mathis, Michael Stamen Publications Committee: Diana Harris. Theodore Sjoerdema, Madeline Windauer The National Educational Computing Conference wishes to thank the following people for their contribution of effort, time, and knowledge as referees for the papers submitted for presentation.

Robert Aiken, University of Tennessee David Maharry, Wabash College William Atchison, University of Maryland Sister Patricia Marshall, Xavier University Emile Attala, Polytechnic State University Donald H. McClain, University of Iowa R. P. Banaugh, University of Montana Donald E. McLaughlin. Augustana College Bruce Barnes. National Science Foundation Edward D. Meyers Jr.. Center for the Study of Joyce Damao, Southern University Health Development Alfred Bork, University of California Percy L. Milligan, Southern University O. R. Boynton. Indiana University Robert D. Montgomery. North Carolina Central Mary Jane Brannon, Huntingdon College University Hans &trey, Tennessee Tech. University Catherine Morgan, Kensington. Maryland Tom Carroll, Michigan State University Mike Moshell, University of Tennessee N. John Castellano Indiana University Robert Noonan, College of William and Mary Sylvia Charp, School District of Philadelphia Bruce E. Norcron. SUETY at Binghamton Ronald Collins, Eastern Michigan University Linda Petty, Hampton Institute Frank W. Connolly, Montgomery College Charles Pfleteger, University of Tennessee Michael J. D'Amore, New Jersey Educational Richard Pogue, Medical College of Georgia Computing Network W. W. Porterfield, Hampden-Sydney College Charles Davidson, University of Wisconsin James Powell, Burroughs-Wellcome Co. Herbert Dershem,\Hope College C. A. Quarles, Texas Christian University David Evans, VIMS Joe Rabin. Queen College of BURY Selby Evans, Texas Christian University Ottis Richard, University of Denver Stefan Yeyoch, College of William and Mary David Rine, Western Illinois University F. T. Fink, Michigan State University Peter Riese. Control Data Corp. Fred Gage, Texas Christian University Leroy Roquemore. Southern University Norman Gibbs, College of William and Mary R. C. Rosenberg. Michigan State University Sheldon'P. Gordon, Suffolk County Community Ted Sjoerdsma, University of Iowa College David Smith, Duke University Keith Hall, Ohio State University Philip F. Spelt. Wabash College John W. Hamblen. University of Missouri Elliot A. Tanis. Hope College Marry Hodges, Michigan State University Robert Tannenbaum, Ancram, New York Darlene Heinrich, Florida State University David Thomas. University of Iowa Jews A. Higgins. Digital Equipment Corp. Robert Thompson, University of Dayton A. AA J. Hoffman. Texas Christian University John Van Iwaarden. Hops College Lawrence Jehn, University of Dayton Rita Wagstaff, Temple University Vincent H. Jones. Baton Roque, Louisiana T. C. Willoughby, Ball State University Basheer Khumawala, University of North CarolinaGary Wittlich, Indiana University Joyce C. Little. Community College of BaltimoreAllen Disbur, SUNK at Binghamton Richard W. Lott, Bentley College

And special thanks to the authors whose cooperation kept this publication on schedule. .,.-Diana Harris

5 Foreword

The Second National Educational Comput- equally able hands of Doris K. Lidtke of ing Conference (NECC/2) builds on the Towson State University. We give them our success of last year's conference at the sincerest thanks for the many hours spent University of Iowa. In organizing the on these tasks. We also thank the authors program for NECC /2 we have attempted to for submitting their works and working with implement many of the suggestions we have us to maintain our schedule. received during the year. Of special Thanks are also due to the entire NECC/2 significance are the tutorial sessions Steering Committee for their excellent which are designed to provide attendees guidance in preparing for the Conference. with the opportunity to expand their Special acknowledgements go to Ted appreciation of and involvement in Sjoerdsma of the University of Iowa who educational computing. coordinated publicity and Diana Barris NECC/2 is a broadly based conference' also of the University of Iowa who edited bringing together, in the common inter- these proceedings. ests of computers in education, a great Finally we thank all those individuals number of individuals with diverse back- who came to NECC/2 and helped insure that grounds and a great number of cooperating the concept of this series of conferences societies. It is hoped that in this is a success. forum the diversity can be focused to improve our common interest. This volume presents the papers presented at the conference and summaries of Gerald L. Engel most of the special sessions.The coordi- Chairman, NECC/2 nation of the contributed papers was in Christopher Newport College the able hands of Richard H. /mating of Newport News, Virginia 23606 the University of Maryland, and the coordination of the special sessions was in the

CV Table of Contents

TUTORIALS 1 CAI-An Introduction Michael Aronson and Harold Rabmlow 2 How to Choose a Microcomputer for Educational Use Kevin Hausmann 3 Turning Students on to the Computer: The Introductory Course Gary Shelly 4 Instructional Design Gary Stokes

INVITED SESSION 5 . Microcomputers in Education Chaired by Murali R. Varanasi

TUTORIAL 6 Program Development Techniques A. J. Turner

BUSINESS/ECONOMICS 7 Computer Science and MIS College Students: An Investigation of Career-Related Characteristics Eleanor W. Jordan 12 Force- Feeding SPS8 in Market Research and Analysis at Hampton Institute Howard F. Wehrle III 16 Short-Run Forecasting of the U.S. Economy Minims R. Bowman

TOOLS MD TECHNIQUES FOR INSTRUCTION 19 Hypertext: A General-Purpose Educational Computer Tool Darrell L. Ward and Steve Bush 25 A Dynamic Process in Teaching Techniques Jamil E. Effarah 31 Considerations and Guidelines for Developing Basic Skills Curriculum for Use with Microcomputer Technology Robert M. Caldwell

INVITED SESSION 37 'Research in Microcomputer Uses in Education Chaired by David Xdiefel

CCMPOTERS IN HUMANISTIC STUDIES 39 Creativity through the Microcomputer George M. Bass, Jr. 42 Giving Advice with a Computer James W. Carson 46 ?roma Theory of Reading to Practice via the Computer Dale M. Johnson and R. Scott Baldwin 54 Non - Harmony: A Vital Element of Bar-Training in usic CAI Joan C. Groom-Thornton and Antoinette Tracy Corbst 7 vi COMPUTER LITERACY 58 A Case for Information Literacy Bruce B. Schimming 62 A Byte of BASIC. Judith Am Hopper 65 A Computer Workshop for Elementary and Secondary Teachers Herbert L. LW:rah:to and John T. Whittle 68 Microcomputers and Computer Literacy: A Case Study' Robert J. Wilson

INVITED MISSION 73 Personal Computing: An Adventure of the Mind-- Paul Hazen

INVITED SESSIONS 74 Educational Computing: Past, Present, and Future Ronald W. Collins 75 The Open University Prank Lovis and William Dorn

SCIENCE AND ENGINEERING 76 Demographic Techniques in Ecology: Computer-Enhanced-Learning A. John Gets, Jr. 81 Microcomputers as Laboratory Instruments: Two Applications in Neurobiology Richard F. Oliva Classical Mechanics with Computer Assistance A. Douglas Davis 90 Computer-.Augmented Video Education in Electrical Engineering at the U.S. Naval Malawi Tian S. Lim, Michael W. Magee, and Richard A. Pollak

STRUCTURED PROGRAMMING 96 The Use of Programming Methodology in Introductory Computer Science Courses Elisabeth Alpert 103 FORTRAN 77:Impact on Introductory Courses in Programming Using FORTRAN Prank L. Friedman 112 Using Model -Based Instruction to Teach PASCAL Bodgan Csejdo 119 Structured Machine-LanguagesAn Introduction to Both Low- and High -Level Programming David G. Hannay

ACM ELEMENTARY AND SECONDARY SCHOOLS SUBCOMMITTEE 125 ACM Elementary and Secondary Schools Subcommittee Progress Report David MOursund 130 Computing Competencies for School Teachers Robert P. Taylor, James L. Poirot, and James D. Powell

INVITED SESSION 137 CAUSE Program and Projects Chaired by Lawrence Oliver

TUTORIAL 138 Databases - What Are They? Arlan DeKock COMPUTER SCIENCE EDUCATION 139 Required Freshman Computer Education in a Liberal Arts College David E. Wetmore 143 Development of Communications Skills in Software Engineering John A. Seidler and John G. Meinke 147 Systematic Assessment of Programming Assignments Judy M. Bishop 152 Data Structures at the Associate Degree Level Richard F. Dempsey

INTEGRATING COMPUTING INTO K-12 CURRICULUM 155 The Scarsdale Project: Integrating Computing into the K-12 Curriculm Thomas Sobol and Robert P. Taylor 168 Panel Beverly Hunter and Catherine E. Morgan

INVITED SESSION 169 Funding Academic Computing Programs Sheldon P. Gordon and Lawrence Oliver

TUTORIAL 170 Techniques for Instructional Software Development Using Microcomputers Kevin Hausmann

MATHEMATICS 171 A Method for Experimenting with Calculus Using CAI Frank D. Anger and Rita V. Rodrigues 179 Computer Applications in a Finite Mathematics Course G. piegeri. K. Abernethy. and A. L. Thorsen 1B4 A Computer-Assicted Course in Biomathematics Pui-Kei Wong 194 Computer Symbolic Math David R. Stoutemyer

INVITED SESSIONS 197 Computerilased Resource Sharing Donna Davis Mebane and Rodney Mebane 198 Computers and Instruction:Development, Directions, and Alternatives Chaired by William Gruener

TUTORIAL 199 Videodisc Bobby Brown and Joan Sustick

TESTING/PLACEMENT 200 MicrocomputerAssisted Study and Testing System Hugh Garraway 205 RIBYT - A Database System for Formal Testing and SelfAssessment F. Paul Pubs 214 Computer-Managed Placement in Mathematics Instruction for Health Occupations Students Thomas A. Boyle and Peter Magnant

9 INVITED SESSIONS 220 Improving Utilization of Two Year College Computer Centers Chaired by Joyce Currie Little 0 221 Teaching Computer Ethics (Workshop) Walter Mauer

TUTORIAL 222 PASCAL H. P. Haiduk

PRE-COLLEGE INSTRUCTIONAL MATERIALS 223 Computer-Based Instruction for the Public Schools: A Suitable Task for Microprocessors? Timothy Taylor 230 Microcomputer/Videodisc CAI - Some Development Considerations Ron Thorkildsen and Kim Allard 236 The Won-Technical Factors in the Development of CAI Michael Mocciola

INVITED SESSIONS 237 Data Sets Available from the Federal Government Chaired by Thomas E. Brown

MINORITY INSTITUTIONS ECMI 238 Academic Computing:A Sampler of Approaches in Minority Institutions Sr. Patricia Marshall 245 Computer Use in Chemistry at a. Minority Institution James D. Beck 249 Educational Use of Computer* in Puerto Rico Frank D. Anger

COMPUTER LABORATORIES IN EDUCATION 250 Microcomputers in the Teaching Lab Robert F. Tinker 256 The Computer Lab of the 80s Guy Larry Brown 258 The Education Technology Center Alfred Bork, Stephen Franklin, and Barry Kurtz

INVITED SESSION 260 MIS Education: Industry Needs and Educational Solutions Chaired by Eleanor W. Jordan

COMPUTER GAMES IN INSTRUCTION 261 Shall We Teach Structured Programming to Children? Jacques E. LaFrance 266 Structured Gaming:Play and Work in High School Computer Science J. M. Moshell, G. W. Amann, and W. B. Baird 271 Tapping the Appeal of Games in Instruction Peter O. McVay

o COMPUTING CURRICULA 276 An Sducational program in MedicalComputing for Clinicians and Health Scientists Albert Hybl and James A. Reggie 281 A Secondary Level Curriculum inSystem Dynamics Nancy Roberts and Ralph M. Deal 287 The Computer Software TechnicianProgram at Portland Community College David M. Hata 290 A Computer Science Major ina Small Libe.al Arts College Joerg Mayer

293 Author Index

X Tutorials

CAI - AN INTRODUCTION

Michael Arenson Dept. of Music Univ. of Delaware Newark, Delaware 19711 (302) 738-8485 Harold Rahmlow Abacus Learning Inc. 531 Lancaster Avenue Wayne, Pennsylvania 19087 The American College Bryn Mawr, Pennsylvania 19010

ABSTRACT The NECC-2 Computer-Based Instruction Tutorial Session is designed for persons who have had little or no experience working with computer-based instruction (CBI). This session will give basic information that will help them get started in CBI. The topics for the session include:

(1) An examination of terms related to CBI and the differences between computer assisted instruction, computer- managed instruction, and computer test generation (2) Characteristics of today's educational environment and the place of CBI in it (3) Alternate CBI systems and characteristics of each (6) Examples of successful CBI (5) Questions one should consider before getting into CBI (6) Where to get additional information or assistance concerning CBI

12 2 NECC 1980

HOW TO CHOOSE A MICROCOMPUTER FOR EDUCATIONAL USE

Xevin Hausmann Minnesota Educational Computing Consortium 2520 Broadway Drive St. Paul, Minnesota 55113

ABSTRACT --Milypeople are realizing the tremendous Once the range of systems has been potential that microcomputers have for narrowed, ways to provide or acquire education; however, it is becomming increas- support should be considered. The ingly difficult to stay abreast of all the following points were deemed important varieties of microcomputers currently in the Minnesota plan for microcomputer available. One solution to this problem support: is to define the components that make up - One specific microcomputer (chosen a minimal educational microcomputer system through a bid process) should be and then consider only those systems which available to educational agencies meet minimal criteria. by a state contract. The minimal system, as defined by - Instructional service support for Minnesota educators, consists of the the selected microcomputer should be following: defined end increased to the level - An input device must be a typewriter currently available for timesharing. keyboard and output a multiline monitor or printer. - Continuous analysis and evaluation - a permanent file storage must be some of hardware and software must keep form of disk storage. up with the changing technology. - The BASIC language must be supported. The technical evaluations and the - At least 12X of user memory must be invitation for bid used by the Minnesota availalbe, excluding operating Educational Computing Consortium are system and language processor(s). available from XECC Publications, 2520 Since BASIC is the most often used Broadway Drive, St. Paul, Minnesota 55113 language, an evaluation of BASIC language Ask for the 79 -80 Microcomputer Report. features and capabilities should be made. BASIC features typically considered include sequential file handling, random access file handling, chaining, special functions, matrix operations, formatted Output, and good graphics commands.

13 Tutorial3

TURNING STUDENTS ON TO THE COMPUTER: THE INTRODUCTORY COURSE

Gary 8. Shelly Anaheim Publishing Company 1120 East Ash Fullerton, CA 92631 (714) 879-7922

ABSTRACT --7Flai long as computing has been taught in high schools, colleges, and universities, there hii been controversy concerning the first course in the curriculum.Some have advocated a high-level course dealing with algorithms and programming, while others have taken a computersn-society approach. With the increased enrollment in this course, notonly in colleges and universities but in high schools and even junior high schools, guidance must be given to teach a worthwhile course which can satisfy the large number of students from all disciplines who will be taking it. This tutorial will present a suggested course content and methods for implementing this course at all levels of education. It will include justification for the subject matter which the author views as critical for the course and unique and effective ways in which this subject matter can be taught effectively.

14 4 NECC 1980

INSTRUCTIONAL DESIGN AND COMPUTER EDUCATION

Gordon Stokes Computer Science Department Brigham Young University Provo, Utah 84602 (801) 374-1211 Ext. 3027

ABSTRACT mace of performance objectives, course organization, instructional strategies, and evaluation procedures will be discussed in this session. The instructional design of an introductory FORTRAN class will be the case history that illustrates the instructional design process in a classroom. The presentation will give the participants enough information and examples to help them organize their own classes. The evaluation procedures will cover both formative and summative exams. Some recent research results on individualized instruction in an introductory class will be presented, and their Applications for instructional design will be discussed.

15 Invited Session

. MICROCOMPUTERS IN EDUCATION

Chaired by Murali R. Varanasi Dept. of Electrical Engineering Old Dominion University Norfolk, VA 23508 (804) 440-3742

ABSTRACT 7-the introduction of the first microprocessor in 1972, advances in the microprocessor field have so accelerated that universities have been faced with a aerious educational challenge.To take advantage of the cost and flexibility offered by the programmable LSI devices, the digital designers have to be educated in software engineering; computer scientists have to learn digital systems at the gate and subsystem level. Even though microprocessors, they must be dealt with as any other computer--i.e., interfaces have to be built, buses have to be designed, and programs written.There- fore the following challenges arise: 1. How are design skills to be taught?, 2. What sort of courses are needed? 3. What sort of laboratories are needed? 4. How does one introduce a Computer Science/Engineering student to application areas? This panel will address the above questions from the perspective of experienced education, industrial, and government experts knowledgeable in computer education. Future applications of microprocessors will also be discussed.

5 -1 Tutorial

PROGRAM DEVELOPMENT TECHNIQUES

A. J. Turner Clemson University

ABSTRACT Computer programs are often developed by users without considering the effort that will be required for someone other than the author(s) of a program to understand it modify it, or correct an error. However, several techniques are available to assist in the development of computer programs that are easier to road, understand, debug, and modify. Since tee techniques also facilitate the initial implementation of most programs, they are valuable even if a program is intended to be used without modification. Techniques for three aspects of program development are considered in this tutorial: design, programming, and implementation. Topdown design is discussed as the basic approach to program design. Module independence, module function, information hiding, and the HIPO technique are considered as modularization criteria and paradigms to -id in the development of the design.The two programming techniques discussed that facilitate the development of program code are the use of pseudo-code and stepwise refinement. Also included are techniques for improving the readability of program code, such as structured coding and program formatting and commenting conven- tions. The use of iterative. enhancement and module stubs are discussed as implementation techniques that complement the desigr and programming techniques and facilitate the use of a topdown approach to program development. Emphasis is placed on the use of these techniques by a small implementhtion team or a single individual. Examples in the BASIC and PASCAL programming languages are included.

6 Business/Economics

COMPUTER SCIENCE AND MIS COLLEGE STUDENTS:

IS THE STAMPEDE FOR THE MONEY?

Eleanor W. Jordan Department of General Business University of Texas at Austin Austin, Texas 78712 512-471-5322

INTRODUCTION graduates with an adequate education College programs in computer science for business applications.He suggests and management information systems (MIS) that computer companies spend less are generally experiencing increasing time raiding each other's DP employ- enrollments in spite of declining or ees and more time interacting with stable university-wide enrollments. educational institutions in order to Since the demand for DP professionals improve the DP personnel situation. is currently very high, enrollment Several-educators have joined increases in these two areas may be an industry recently in complaining that indication of the practical orienta- computer science education is often tion of the '70s college students who irrelevant to industry needs (2). supposedly left the liberal arts and Resulting recommendations have often demonstrations of the '60s to pursue been to add particular courses or safer career paths in a declining job stress particular languages (2), but market. if this is true, will colleges there has been an increasing interest be producing more computer graduates in in new programs designed for the the '80s who are money-oriented rather broadly defined fields of software than computer-oriented? engineering (7) or information systems industry is unlikely to be dis- (4). mayed at the prospect of more practical Before too many new programs are computer science graduates. Bruce developed, it might be valuable to Gilchrist (5) predicts that the short- consider whether it is possible to age of OP professionals will continue relieve industry's frequently expressed through the *80s mainly because educa- frustration with computer science tional institutions are not producing graduates by making curriculum changes.

7 .18 8 NECC 1980

Paul Anagnostopoulos (1), a participant Are people who choose DP careers markedly in the 1979 Brown University conference different from people in general? Couger on the difference between software and Zawacki found that practicing DP pro- theory and practice, decided that the fessionals have lower social needs than much talked-about software crisis is others and concluded that this trait had due mainly to programming personality important implications for how DP work defects. His suggestion that more environments and job requirements should be attention be paid to the psychological designed. If computer science and MIS aspects of programming and systems students differ from other students in design is similar to Daniel Cougar job-related attitudes and achievement moti- and Robert Zawacki'e (3) suggestion vations, it could have implications for that DP professionals consider behav- program design, course design, and student ioral aspects of personnel responsibi- advising in the computer science and MIS lities as well as the required technical areas. skills. In their investigation of more than 600 DP professionals (analysts, STUDENT SURVEY PROJECT programmer/analysts, and programmers), In spring 1977 I was involved in the Couger and Zawacki found that the development of an undergraduate MIS program stereotype of programmers as loners in the College of Business Administration had considerable basis in fact: DP at The University of Texas at Austin. A professionals reported significantly strong computer science program already less need for interaction with others existed at the graduate and the undergrad- than all six of the other professional uate level, but odr business school committee groups studied. thought that the orientation of the computer

TABLE 1. Mean Ratings of Importance of Possible Considerations In Career Choice for Student Sub-Groups

Career Considerations Business MIS ComputerLiberal Natural Engineering in RInk Order of Science Arts Sciences Importance (N=326) (N=21) (N2r20) (1440) (N=45) (N=15)

1. Personal Interest* 3.4 3.3 3.0 3.6 3.5 3.4

2. Personal Abilities 3.1 3.2 3.2 3.3 3.4 3.1

3. Career Potential 3.1 3.3 3.3 2.8 2.9 3.5

4. Job Security 3.0 2.6 3.1 2.4 2.7 2.7

5. Salary** 2.7 2.4 2.9 2.2 2.3 2.5

6. Amount of Time Allowed Outside Work 2.6 2.4 2.4 2.5 2.3 2.9

7, Flexibility 2.5 2.5 2.5 2.5 2.4 2.4

8. Variety of Job Requirements 2.4 2.3 2.3 2.4 2.2 1.9

9. Opportunity for** Community Service 1.8 1.5 1.3 2.2 2.5 1.5

10. Prestige 2.0 1.5 1.9 1.9 1.6 1.7

Items are measured on a five-point scale with Osindicating "not at all important* az.d 4 indicating "extremely important." Asterisks indicate that an analysis of variance resulted in a statistically significant difference among sub-group means at the following levels: *--p4.05, **--14.01 19' Business/Economicsg

TABLE 2. Mean Scores for Student Sub-Groups On Work and Pamily Orientation ?actors

Factor Business MIS Computer Liberal Natural Engineering Science Arts Sciences (A=326) (N=21) (N=20) (N=40) (N=45) (N=15)

Mastery 20.1 20.3 19.0 19.2 19.6 19.7 Work 19.7 20.7 18.4 20.4 20.5 19.9 Competi- tiveness** 13.7 12.1 11.2 12.1 12.3 12.9 Personal Unconcern 10.3 10.3 11.2 10.9 10.7 10.1

The overall group means for these factors were very close to the Helmreich and Spence normative data for 1455 college students. **An analysis of variance resulted in a statistically significant difference among sub-group means at the .01 level. science undergraduate program was more cided group had specific career goals, suitable for aspiring graduate students, while the other half indicated that they systems programmers, or scientific pro- had decided on a general area but were not grammers than business applications sys- sure about what their specialization might tems analysts. The MIS program was de- be. signed to provide the education for a On the survey students were asked to knowledgeable business user who can ef- rate the importance of ten elements in fectively define his information needs their considerations for choosing a career as well as analysts who can design the field on a scale of 0 to 4. The ten ca- requested' software without the usual reer choice items included on the survey conflict between user and DP personnel. are listed in Table 1 in the order of aver- Pall 1979 was the first semester age importance to the entire sample. for the proposed MIS program. Since it Personal interest in a career field is a computer program in the business had the highest mean rating for all stu- school I was interested in whether the ..J'aftt 'sub-groups except the computer science students attracted by the program would students and engineering students who tended have work-related attitudes and motiva- to rate the importance of career potential tions similar to business students or more highly than any of the other items CS students. I therefore included stu- listed. dents in two of the required MIS classes The generally high ratings given -to in a Pall 1979 survey project that mea- personal interests in career choice deliber- sured attitudes from a large sample of ations are not consistent with a popular business students at the undergraduate view of '70s students as obsessed with fi- and masters level.A smaller sample of nancial security, but the rest of the items liberal arts, natural sciences (includ- in the top half all fit this view.Matching ing computer science), and engineering a career with personal abilities ranks sec- students was also included for a compar- ond for most groups and career potential, ison with the MIS group. The resulting job security, and salary follow. sample included 467 undergraduates) the To determine what reliable inter-group large majority were business majors in differences might exist in deliberations accounting, management, or marketing. about career choice, I performed a separate analysis of variance for each of the ten RESULTS items. Statistically significant differences among means were found for three of Determinants of Career Choice the items:personal interest, salary, and Most of the students participating For all in the survey reported that they had opportunity for community service. three of these items the mean for computer decided on a career goal, even though science students was at one end of the the sample included about as many fresh- range of sub-group means and the man and sophomores as it did juniors and seniors. Approximately half of the de-

20 10 NECC 1980

means for liberal arts students and the mean score on competitiveness for MIS natural science students were at the other students was similar to the middle scores extreme. The mean for MIS students was of liberal arts and natural science stu- closer to the other sub-group means for dents. all three of these career choice items. For these three considerations in making SUMMARY AND CONCLUSIONS a decision about a career, the MIS stu- The sample of computer science and dents appear to be more like other stu- MIS students participating in this sur- dents than the computer science students. vey was quite small so any conclusions For the other considerations included that can be drawn from the results must in the survey, the computer science be tentative.Where comparisexe could students, like the MIS students, appeared be made to other mtudies the data did to rate each item of similar importance appear to be representative of college or lack of importance to the other sub- students and consistent with the Couger groups. and Zawacki (3) study of DP professionals' personal needs. The additional implica- Achievement Motivation tions of my study may then be worth consi- The su=vey also included four dering at least for purposes of discus- measures of achievement motivation. sion. These measures were taken from Robert For the most part computer 'science He]areich and Janet Spence's (6) Work and MIS students appear to make career and Family Orientation Questionnaire choices in a manner similar to other since considerable evidence exists for students and have similar achievement the statistical reliability and valid- motivations. However, where statisti- ity of this instrument. The four fac- cally significant differences do exist, tors identified in Helmreich and the computer science students were found Spence's analyses are designated as consistently to have an extreme posi- work, mastery, competitiveness, and tion, while the MIS students consistently personal unconcern. The first two indicated choices and motivations simi- would seem to be highly relevant to lar to the other student sub - groups. realistic motivations of aspiring DP professionals in the rapidly chang- Implications for Industry ing computer environmentswork is If the results of this study in an intended to measure the desire to work academic setting are combined with the hard and mastery measures desire for Couger and Zawacki (3) study, the impres- intellectual challenge. The items sion of the DP professional is that of composing the competitiveness factor someone who has low needs for social are related to a desire to succeed in interaction and little desire for inter- competitive, interpersonal situations. personal competition relative to other A high score on the personal unconcern professionals or aspiring business factor indicates a relative lack of executives. Computer science students concern about the possible negative are just as interested in an intellec- interpersonal consequences of achieve- tual challenge ae other students, accord- ment. In validation studies these facing to my study, and have a greater need tors have been found to be appropri- for personal growth than other profes- ately related to measures of scienti- sionals, according to the Couger and fic achievement, college grades, and Zawacki study. But challenge and growth income. are apparently not defined in terms of Computer science and MIS students, social interaction or interpersonal according to my results, are similar competition. If the business execu- to business, liberal arts, natural tive understands this then it seems highly sciences, and engineering students in plausible that it would be possible to terms of work, mastery, and personal capitalize on the task orientation of unconcern (Table 2). In a comparison the traditional DP professional and of the six sub -group means the results benefit.from the lack of possibly des- of an analysis of variance test was not tructive interpersonal competition. statistically significant for any of Some of the closed-door complaints that these three factors. However a Statis- are sometimes heard about DP shops may tically significant difference among stem from an assumption that the source means was found for competitiveness. of the isolation policy is a know-it- The highest mean factor score was for all egocentricity, when in fact it is the business students and the lowest likely to be a difference in task orien- was for the computer science students, tation.

21 Business/Economics11

The results of this survey related REFERENCES to MIS students seem very hopeful for a future reduction of the frequency of 1. Anagnostopoulos, P. "Software Crisis: conflicts between computer experts Method or Psychology?" Computerworld, and business managers. For all ana- October 29, 1979, pp. 23, 28. lyses where differences in motivations or importance placed on career choice 2. Cook, J. R., Gallagher, M. C., Johnston, considerations were found to be statis- M. A. 'An Analytical Study of Industry's tically significant, the MIS students Computer Education Needs." Interface, tended to take a middle position. Vol. 1, No. 1, Winter 1979, pp:37. Hiring a combination of MIS and computer science graduates may very gall result 3. Conger, J. D. and Eawacki, R. A. "What in a DP shop that has better intro- Motivates DP Professionals?" Datamation, company relationships and still delivers Vol. 24, No. 9, September 197 8, the technical expertise that may re- pp. 116-123. quire an almost exclusive orientation toward the task at hand. 4. Davis, G. B. "Information System Another hopeful note is that the Curricula in the Business School." programmer's big ego desoribed in Interface, Vol. 1, No. 1, Winter 1979, some of the stories of Gerald Weinberg 113727:177- (8) in The Psychology of Computer ?rIgraw,mEg doesn't seemo pre- 5. Gilchrist, B. "Wanted: DP Professionals va ent among either MIS or CSstudents. or the '80's: Computerworld, January Weinberg's observations may be more 7, 1980, pp. 6 -T. based on egocentricity than egotism. 6. Helmreich, R. L. and Spence, J. T. Implications for Educators "The Work and Pamily Orientation Ques- Gordon Davis (4) has discussed tionnaire: An Objective Instrument to the conceptual differences in the func- Assess Components of Achievement tions of information analysts and system Motivation and Attitudes Toward Family analysts as a basis for designing and Career."JSAS Catalog of Selected curriculum in MIS and computer science. Documents in Psychology, Vol. 8, 1978, The results of my study provide some pp. 35-55. additional support for the distinction between broadly defined programs 7. Jensen, R. W. and Tonics, C. C. "Soft- in information systems and the tradi- ware Engineering Education - -A Construc- tional computer science program. tive Criticism."Proceedings of the The middle position occupied by the Sixth Texas Conference on Computing MIS students on a number of career Systems, November 1977, pp. 5B-7 - 511-13. orientation and achievement motivation measures may mean that they a. Weinberg, G. The Psychology of Computer will generally be successful in the Programming. New York: D. Van Nostrand liaison and managerial roles they are iVil. often placed in. Also MIS programs apparently attract a different type student than the computer science programs. As the number of MIS.programs increase the number of graduates seeking DP positions may increase at a faster pace than expected. The high salaries and eager on- campus industry recruiters would seem to be the most obvious reasonsfor the continuing increase in MIS and computer science enrollments, but that's not what these students are reporting. Personal interests are more important than salariesfor all these late '70s college students. With a continuing increase in college graduates and 'a variety of computer educational programs, perhaps the software race Will catch up with thehardware advances faster than thelatest predictions indicate. 12 NEM 1900

FORCE-FEEDING SPSS IN MARKET RESEARCH AND ANALYSIS AT HAMPTON INSTITUTE

Howard F. Wehrle, III School of Business Hampton Institute Hampton, Virginia 23668 (804) 727-5362

INTRODUCTION Hampton Institute (HI) is a small, Computers in Business Management (7), !West/stove, four-year historically which emphasises business applications black college that grants graduate degrees and problems with substantially less in education, nursing, and (in conjunction attention to detailed programming tech- with George Washington University) engi- niques. In addition, preliminary guidance neering. As of September 1979, the School from the Chairman of the Department of of Business included 15 faculty members Management and Marketing had emphasized and 792 students (the largest enrollment that FORTRAN is not necessarily the pre- of any element of the Institute) who referred language. Accordingly, minor presented 36 states, the District of preliminary study has addressed the possi- Columbia, and the Virgin Islands. The ble introduction of the C language (2) School of Business is actively pursuing since software compatible with the Harris preliminary actions toward accreditation PDP-11 (now biking installed at HI) is of the undergraduate program by the available. This guidance recognizes the American Assembly of Collegiate Schools fact that the goal of the School of of Business (AACSB), with the mid-range Business is not to train computer pro- goal of an accredited Master of Business grammers or potential data processing Administration (MBA) program. managers, but to train competent junior managers with a breed overview of busi- COMPUTER EDUCATION FOR BUSINESS AT ness. HAMPTON INSTITUTE Hardware available to the School of The focal point for computer education Business as of September 1979 included at HI is in the Division of Computer five keypunches, one terminal interactive Science, Department of Mathematics, which with the IBM 370/168 computer at The offers a major in computer science. COBOL College of William and Mary in Virginia is the principle vehicle for teaching, al- at Williamsburg, and an IBM System 3 at though two APL terminals are available. Hampton Institute. The- System 3 allows Xn the School of Business, a single batch processing of Statistical Package computer course, "Computer Concepts in for the Social Sciences (SPSS) programs. Business," has used FORTRAN. Until the 1978-79 school year the course war. taught THE CHALLENGE OP BUSINESS 428-01 by a full-time visiting professor, an On August 30, 1979, the author, primarily IBM employee. Upon his departure, the management rather than marketing oriented, author volunteered to teach both sections was offered the opportunity to teach, as of the course, continuing the soiewhat an overload, Business 428-01, "Market. restrictive program-writing orientation Research and Analysis." This opportunity using FORTRAN, which was 4stablished by was eagerly grasped, after only brief his predeceesor. cone/devotion. A conventional syllabus Beginning in September 1979, however, complemented the text end supported the the main thrust of the course was sub- concept of team determination, team lead- stantially redirected from a rather narrow er selection, and research topic recomeen- technical orientation to a broader mana- dation by the students. For the class of gerial orientation more appropriate to 59 students (34 juniors and 25 seniors) potential junior managers who are under- specific guidance was given orally at the graduate students in business. This re- same meeting: "This course is team- direction included a new text, O'Brien's oriented. Accordingly, by the next class

23 Business/Economics13

meeting please arrange yourselves in teams demonstrate sample SPSS capabilities sad of not less than five nor more than seven to challenge the students to define ana- members, at least one of whom must have logous application to their project. completed 'Computer Concepts in Business' For further reference, two copies of (to ensure a basic familiarity with the the 4PSS manual (5) were available: one mechanics of computer programming, key- chained to the counter in the computer punch operation, and program debugging). center where card decks are turned in for Select your teas leader, and provide the batch processing; the other, the instrue-. instructor a written list of the team torsa parsonel copy. (A third copy, members, designating the team leader, the recommended for purchase by the Institute individual with basic familiarity with library as a reserve reference, computer procedures, and a list of three was not yet availeble at this writing.) topics you propose for research. One of The instructor (along with his SPSS these topics, if deemed suitable, will be manual) was also available to team members approved by instructor. if none of the as a resource person/advisor. first three is considered suitable, the The projects moved forward. Host team will be required to submit a second classes were split between rather cursory slate of three topics." (This requirement lecture coverage of the assigned text and was never invoked, although the instructor substantial time for team work and coa- was asked by members of several teams, "Why gulation with the instructor as required. did you approve that topic? lt wasn't the Meny questions arose, most of which were one we really wanted; we just added it on answered in the Socratic manner or by to meet your requirement for three topics." :oscine reference to assigned coverage The response: "lt was the only one on in the text. "Be of good cheer and read your slate which had any 'meat', the others ahead." Needless to say, these techniques were lightweights which would not hey* roused el:beta:nisi complaint from some of presented any challenge to you.") the class. By the second class meeting, eight teams Sampling of the approved student topics were organized and moving out smartly; the is included in the appendix co this paper. remaining seven students wasted their first One common complaint concerning these four class meetings fighting the problem topics took the general form,"What does but none dropped the course. this have to do with market research and A sample milestone chart and work plan analysis?"The reply was standard: "The were distributed the first day of class; thrust of this course is precisely that teams were advised the next requirements of market research and analysis; while after approval of their research topic we are not attempting to ace the Nielsen were successive submission of an hypothesis, ratings, or ascertain the marketability work plan, milestone chart, and question- od Old Tennis Shoe Bourbon, you are exer- naire(s) to elicit data appropriate to cising the techniques that would be ap- determine whether the hypothesis could be plied in either of those two or a multi- supported or rejected. tude of other examples."

RESPONSE: CONDUCT OF THE COURSE STUDENT PROBLEMS One may wonder, at this point, what need One substantial inhibiting circumstance is there for computer support or what com- in anotherwise orderly progression of puter application exists?After topic ap- learning by doing was the delayed dis- proval specified questionnaire(s) would be tribution of the sample SPSS program by administered to a random sample of not less the instructor to each team. (He had his than one hundred nor more than five hun- own problems in debugging what should have dred respondents. been a relatively simple, straight-forward Each team was issued (and acknowledgement program, owing to his lack of familiarity made by signature of s student team member) with job control language (JCL) and the a folder containing selected extracts from evolutionary progression of SPSS from the SPSS manual describing the options Version 6.0, which he had last used six and statistics available for SPSS proce- years ago, to the current Version R.S.0) dures CONDESCRUT1VE, CROSSTABS, FRE= As the semester moved forward, and the 'Imam PEARSON CORR, and SCATTERGRAW milestone for submission of completed and a data binder containing an unexploded reports approached, the instructor received printout of a program operating on a successive feelers and ultimately almost sample of five with such data as last four frantic pleas to delay submission of the numbers of social security account number, report. Although time was not available age, height, weight, shoe site, color of to do other than edhere to the schedule, hair, and color of eyes. These data items, as a small encouragement the teams were admittedly remote form most business appli- advised that chioce of date and order of cation, were selected primarily both to presentation would be available to the

74 i4w 14 NECC 1980

teams in order of their report submission: a comprehensive review of elementary sta- the first team would choose first, second, tistics as an appendix to one chapter, third, position on any of the three days; the statistics course required in all the last team would have no choice. undergraduate sequences in business at Several teams complained, with apparent NI should be a prerequisite to business justification,about a severe imbalance 428. This requirement would eliminate of work; some team members were obviously the tendency of some students to call up not pulling their share of the workload. unnecessary statistical routines and fail Since it had been emphasized that the same to call up some useful routines. report grade would accrue to each team Third, SPSS is a valuable research tool member, concern was expressed about the which all undergraduate students in busi- apparent injustice of the workers sub- ness should learn to use. This limited sidizing the drones. So, peer ratings application in "Market Research and Ana- were required of all members of each team. lysis" at Eamptoa Institute is the first These were compiled, reviewed, and, after step in that direction: it is a required consultation with a colleague more ex- course for ell. undergraduate marketing perienced than the author in personnel majors. testing and mensuration, appropriately weighted beforetkacorporation in the final course grade. (In addition to the report, four graded exercises highlighting application of text material were administered during the term.) Immediately after submitting each report, it was comprehensively and severely reviewed-by the instructor, as if it were the first draft of a professional report. The following day, the report was returned to the team leader, end after his persual, was discussed at length with him and selected members of his teas. It was emphasized to each team that their oral presentation would provide an opportunity to recoup some of the cuts suffered on the written report. Several teams wanted to revise and re submit the written report, but this com- mendable response wss rejected because the lessons learned from the severely criticized original report might be dulled. The preparation of oral reports would be in- hibited, and such a requirement would place an inordinate burden upon the students at a time when preparation for final examinations should make major demands on their available time.

FINAL ASSESSMENT AND LESSONS LEARNED Although the author considers that business 428-01, "Market Reasearch and Analysis ", was successfully (albeit rather painfully, for some students) completed, he learned some lessons which should significantly improve his second and subsequent conduct of the course. First, completion of "Computer Concepts in Business" should be a prerequisite to Business 428. This requirement would tend to even the student workload, since each teem member would have some experience in keypunching program/data cards; and would be able to participate more fully in preparation of basic data for machine computation. Second, although Lehmann (2) provides 0 Business/Economics 15

APPENDIX: APPROVED RESEARCH TOPICS FOR REFERENCES BUSINESS 428 Subject 1. Alexander, Daniel E. and Messer, Andrew Team C. PORTRAN IV Pocket Handbook. New York: McGraw -Hill, 1972. 1 What effect does the Hampton Institute population have on gross sales of 2. Kernighan, Brian V. and Ritchie, Dennis stores in the Coliseum Mall? M. The C Programmina Language. Engle- wood Cliffs, New Jersey: Prentice- 2 Declaration of majors: Why we choose Hall Inc, 1978. the majors we do 3. Lehmann, Donald R. Market Research and 3 Food additives and preservatives Analysis. Homewood, IL* Richard D. Irwin, Inc, 1979. 4 Salt II Treaty 4. May, Phillip T. Programming Business 5 Gas rationing program in the area Applicatkons 4,12 FORTRAN. BOStOki aoughton-Mifflin Co., 1973. 6 Trade school is a good alternative to college 5. Hie, Norman H., et.al, Statistical, Package for the Social Sciences, 2nd 7 Hampton Institute's intended growth ed. New York: McGraw -Hill Book Co., 1975. 8 How future computerized purchase will affect consumers and industry 6. and Hull, C. Hadlai. SPSS Batch Release 7.0 Update Manual, 9 Social Security March 1977. Williamsburg, VA: Computer Center, The College of William and Mary in Virginia, 1977.

7. O'Brian, James A. Computers in Business Management: An Introduction, revised ed. Homewood, IL: Richard D. Irwin, Inc, 1979.

8. SPSS Pocket Guide Release 8. Chicago: SPSS, Inc, 1979.

26 16 NECC 1980

SHORT-RUN FORECASTING OP THE U.S. ECONOMY

William R. Bowman Economics Department U.S. Naval Academy Annapolis, Maryland (301-267-3156)

OBJECTIVE alternative fiscal and monetary proposals, The work of professional economists and by private individuals when evalua- most often used by government officials ting expected economic growth rates and and private businessmen lies within the inflation within defined sub-sectors of realm of macro-economic forecasting. the economy.The number of the more well This art, however: is rarely acquired by known large -scale models is limited to a students of economics due to the expense handful, however, due to the enormous of developing large-scale econometric cost of building and maintaining the models of the economy and the high level computer-based models. Some models, like of statistical knowledge presumed to be the Data Resources Inc. (DRI) model, have necessary in macro-economic modeling. as many as 200 stochastic equations and In response to this void, a research over 350 endogenods variables. seminar in economics at the :*.S. Naval As indicated above, these cost consid- Academy has been designed to offer under- erations also limit how undergraduate graduate students a chance to build schools teach this important aspect of inexpensive, simplified forecasting economics. If macro- economic forecasting. models. These models are based upon is taught, it is often dons by having the economic theories, or extensions of student merely manipulate previously theories, learned in previous courses. developed models. That is, they may plug They are derived with limited know- in a hypothetical value for a chosen ledge of few, but powerful, statistical policy variable (say the corporation techniques rarely encountered by under- income tax rats) and observe the model's graduates. As piurt of a class exercise, predicted effects on the aggregate econ- each student develops one sector of a omy. While this approach teaches stu- macro - economic model. These sectors then dents the sensitivity of economic sectors become integrated into a simplified model to public policy, it tells them nothing of Gross National Product (GNP) and about the assumptions behind the model of inflation (as measured by the GNP the structure of the equations used in implicit price deflator). the model.

BACKGROUND METHODOLOGY The art of forecasting the economy Re ression Analysis combines intuitive judgments with Titomethodology used for introducing computer-based statistical models of the student to the art of forecasting is aggregate spending behavior. The goal of basically one of learning-by-doing. The these models is. quite heroic, to say the student is first giypp an introduction to least; one must model the decisions of regression analysis.( The theory behind groups of millions of individuals in e the statistical procedure is discussed market place. These individual decisions. only with regard to the assumptions one when aggregated, determine the level. of must make when using regression analysis...... _._production,--as well as the aggregate (Formulae are not proven-since the stu- price 3evel of all goods and services dent is considered a user.) produced. Computers make it easy -- and fun -- The uses of macro-sccnomic forecasts for the student to do regression analysis have been widespread. They are used by from the very beginning of the class. government policymakers when evaluating Simple time-series practice data files

27 Business/Economics 17

are created and stored. These files are access. Much effortis_then taken to "accessed and analyzed with the regression derive a simple model whose explanatory package Time Series Processor (TSP) that variables have the expected sign and are was originally written at Harvard and statistically significant. (See the later modified at Dartmouth to run on a "estimated coefficient" and "t-statistic" time-sharing basis. for the explanatory variables in Table 1 The control statements used to execute of Appendix B.) the program are quickly and easily In addition, the student must work to learned. Students May print out their obtain the best over-all fit of the data data base in easily readable format, com- and minimize the degree of autocorrela- pute correlation matrices, and run pre- tion. (See the "adjusted R squared" and liminary regressions with simple control the "Dublin - Watson statistic" of Table 1 statements. (See Table1 of Appendix A.) in Appendix B.)This process usually More advanced techniques made necessary involves strong trade-offs between the by the problem of autocorrelated error latter two statistics and provides the terms can also be taught to the student student with a real world sense of fore- with the TSP package.Date transforma- casting problems rarely encountered by tions such as discrete lags, logarithms, economic majors. and first differences are easily com- After numerous structural changes of puted. Their effects on the predictive the model, the student selects that model power and autocorrelation can be observed which most closely achieves two objec- by comparing the model results Using tives. First, the model should have a transforms& data with tho3e of the pre- high degree of explanatory power over the liminary regressions. historical data period; second, its fore- More advanced transformations,,includ- cast errors should be minimized.The ing the Cochrane-Orcuttprocedure(21and latter is diagnosed by dividing the his- the Almon Polynomial Distributed Lag torical data into a sample period and a (MOM, are then presented and used. forecast test period. The model is then The latter transformations are usually used to generate predicted values of the considered to be beyond the realm of under- dependent variable that may be compared graduate students; however, the with the actual values. This process is learning-by-doing approach (made avail- especially important during established able by the TSP package) makes their use turning points of specific cycles for both possible and highly instructive. variables classified as "cyclical." (See Table 2 of Appendix A for the con- (See the actual, predicted, and forecast trol statements used for some of these values of a selected variable in Table 2 data transformations.) of Appendix B.) Model Design The Forecast With the statistical knowledge lincestructure of the model is acquired through doing regressions on the chosen, the student uses his best judgment practice data files, the student builds of the values of the exogenous variables a simple regression model (Ordinary Least during the forecast period.This period Squares (OLS) or Two Stage Least Squares is defined in the short-run, for example (TSLS)] for a selected part of the 1979:4 through 1900 :4, to minimize fore- larger macro model. The choice of cast errors. These values are selected explanatory variables is based upon based upon expectations of professional economic theories discussed earlier in economists and business leaders concerned the research paper. The quarterly data with macro-economic forecasting as base used to analyze each student's model reported in recent issuccof numerous is the Department of Commerce's data bank journals and magazines.(") used in its Bureau of Economic Analysis The student then plugs these expected (SEA) quarterly model. This data base values, or a range of values, into his consists of nearly 750 variables and model to produce "conditional ex ante transformations of these variables that forecasts." These forecasts may Si-- may be used as endogneous and exogenous altered if student expectations cast variables. Thus, it provides (at the doubt upon the likelihood of the model's time of this writing) the student with an results, i.e., the "judgmental forecasts." exhaustive source of-quarterly time It ie in this last-pilaffsthat the-Student series information for the period of comes to appreciate the limits of 1949:1 through 1979:3. computer-based modeling and the impor- The student selects independent vari- tance of pereonal, informed judgment, ables that are deemed appropriate for the common to all large-scale forecasting theory he has developed previously and models. creates mass storage data files for easy

.28 18 NECC 1900

Once each student's structural model ---anorforecast-have been completedi-the forecast of GNP and inflation is done by using a reduced fora model with the exogenous variables and predetermined endog- enous variables used in each student's own structural model. This step permits each student to discuss his research project with ether class members, while providing each with an awareness of the inter-relatedness of their work.

CONCLUSION By the end of the semester each student has become fully immersed in a highly technical research topic. Each has related previously learned economic theories to empirical model building using regression analysis on time series data. With the aid of the TSP software and the college's computer hardware, each student has acquired a sense of accomplishment rarely experienced at the undergraduate level. The blend of'economic theory and computer-based technical analysis often opens the eyes of undergraduate majors to a whole new world of excitement in learning.

FOOTNOTES (1) Cochrane, D. and G. Orcutt, "Application of Least-Squares Regressions to Relationships Containing Auto- Correlated Error Terms,* Journal of American Statistical AssoiiiT1757VO1.44 (1949), pp. 32-61. (See Table 3 of Appendix B for an example of the Cochrane-Orcutt adjustment factor.) (2) Almon, S., "The Distributed Lag Between Capital Appropriations and Expen- ditures,* Econometrica, Vol. 30 (1965), pp. 178-96. (See Table 4 of Appendix B for an example of the Almon Distributed Lag coefficient ) (3) The text used in the course is: Chisholm, R.,- and -G. Whitaker, Fore- casting Methods, Homewood: R. r-rrain, 1971. (4) Three sources of information are most often used: (1) Federal Reserve Bank publications (e.g. St. Louis's Review, New. York's Quarterly Review, and WIW%'s New England Review) and large commercial bank publications (e.g. Citibank's Mopthly Economic Letter, Chase Manhattan Bank's Business rti-alrf and International Finance, and Morgan ------Guarantyks-MOrin570iiranty-8urvey)1 (2) national business magazines (e.g. Business Week, Forbes, and Fortune): and (3) newspapers e.g. Nei-YEFEtines and the Nall Street Jourarit

29 Tools and Techniques for Instruction

HYPERTEXT - A GENERAL PURPOSE EDUCATIONAL COSUVTZH TOOL

Darrell L. Ward North Texas State University

Steve Bush The University of Texas Health Science Center at Dallas

The Hypertext computer system is described in this paper.Although its applications are varied, the major emphasis here is the use of this system in the educational environment. The major features as visible to the student and instructor are developed. For the instructor, these features include the use of Hypertext as an organization and lecture presentation tool. For the student, Hypertext provides a model of information that mirrors a library yet permits a computer - assisted instruction approach as information is perused.

ItiTRODUCITON

Hypertext, as an information organizing fa- and the student. Section 2 will describe the cility, wee first described in 19/0 (1).The model of information that Hypertext presents to major emphasis of this paper will be the use of its users.Section 3 will introduce the instruct- Hypertext in an educational setting.Hyper- or's use of Hypertext in creating lecture mater- text has been implemented at The University ials, proaenting information in the classroom in- of Texas Health Science Center at Dallas teracting with students, and maintaining current and is currently being used within the Medical materiels in the subject areas.Section 4 will Computer Science Department at that facility. demonstrate the student.s use of Hypertext.This The implementation details of this system section will describe the environment with respect will not be described, though it is appro- to one course, although general use by the student priate to describe the computing environ- could benefit the total educational p:ocess.. ment within which Hypertext now functions. Finally, section 5 wila summarize the ideas pre- Hypertext operates on a highly reliable, sented and describe areas of further research dual processor Tandem -16 minicomputer system. using the Hypertext tool. The Tandem-16 was deeigned to satisfy critical computing functions, thus the architec- MCOEL or THE HYPERTEXT ENVIRONMENT ture reflects the design criteria with a high degree of redundancy. The total sys- The basic unit of information in Hypertext is tem philosophy is to run non-stop and, in a pages the user is provided an environment of fact,ersingle hardware failure does not pages and relationships among those pages. Each crash thio-Itiretem or contaminate the data in page in Hypertext is owned by an individual from any way (1.1.: The implementation language the community of users and may be declared private, is TAL(Tandem mApplication Language),_ as .preventiktother users from reviewing its con- block-structured, IIGOL -like language. The tents. However, the philosophy of Hypertext is to Tandem-16 is capable of supporting page-mode provide an environment of information (haring, terminals, which permit user interactions to thus typically most pages are public, accessible occur on pages of data, much like pages of books. to the community of users. With the above as background, the remain- Each Hypertext user enters the system der of the paper will alarm the way Hypertext through a page designated as the top page or presents information to both the instructor entry page. When the user is identified to

19 3(j 20 NECC 1980

Hypertext and verified as a valid user, the top another page by inserting a "HYPERRIMP" to that page- of that user-is -immediately.presented to him. page number. Any number of pages may be-ietated The user is then in command mode and mays to a particular page by linking them in the above 1) edit information on the current page. described manner. The Hypertext system imple- 2) view another page via a menu selection for- ments the linking by providing the user a numeric mat available on the current page. selection menu of pages (numbered sequentially 3) view another page via an 4mplicit relation- from 1) which are accessible from the current ship between the current page and the "next" page. page, permitting the user to surely select the 4) view another page by explicitly naming the page of interest by depressing the appropriate page of concern. numeric key. 5) establish a new page and formalise a rela- The above environment of pages is illustrated tionship between the new page and the current page. in the following example representing one user's 6) establish a relationship between the cur- top page and its set of possible relationships rent page and some other already existing page. with other pages. (See figure 1 fora graphic The above operations are by no means exhaustive, model of the example.) In this example the top but identify some of the major functions that re- page provides for the selection of one of three late to the educational use of Hypertext. other pages. The other pages includes 1) a page containing "things to do" with no BALATTOVSHIPS AMONG PAGES In HYPERTEXT additional relationships. 2) a page containing the current classes Each page within the Hypertext system is giv- taught by the instructor (e.g. FORTRAN, COBOL, en a unique page number that is available for dis- etc.) and additional access to the rosters of play along with the informational contents of the each of those classes. page. This system-assigned number is an explicit method of estabiishing inter-page relationships. Any page can be created or altered to point to

4e0 sots (I)Wags to do

Urals to do (2)Orr.. neaten o prepare Mot vane * 3:00 p.a. mottos o trade rung:3 ocou (3) reurso kotttlelo o reid OM Article o wart on SIN ptoposai o moth on DINS lecture page 23 notes o call J. Smith tat* utak yagy SS

Covers GGGGG tS Class ootettels (1) FORUM (1) FORUM

(2) COWL (2) COSOL (3) Structured Frog. (3) Structured otos. (4) 06.S See Urns to tot PASOAN clan: negotiate peso 27 %($) Op. Sys. psis 66,

tOtalthS class rooter COSOL cleat' Teeter Structuted frog. close soscss 0Ed Illttfane Soo Jests Mitten Jones Jill Smith 40 Shirley 1011 Shirley lilt MOTOR Yet Jeff herd Jerry Knowles

Pent 96 Nab 73 4 Tess 34

Figure la. An example Hypertext organisation for an instructor

31 Tools and Techniques for instruction 21

TORUN eouctin (1) Introdoctlat (2) txpresslomo Asslinevnt state. (4) I/o ($) Logical Ii

(10) Nettles sublime.

pate 123

Introduction to FOAMY

text

Peie 124

1. .11LM Dope. eel.cher UNIVAC Which compeer *pee:- CTuelt, Mender the nor MC were Involved handed the development Irallershlp ofMop /* the devevelopom of of MIRAN? FORTNAlktin lecke, led Dachas led the Vay far the development ffietdevplepimac effect. (1) UNIVAC of TAMAN am le an linter, (2) ISM ..... (3) DEC

Ma00, pato111

... continuation of TOMO text ...

Page 03,

Figure lb.Hypertext example continued

3.C) 22 NECC 1980

page_containing_courses.already-develop-- -review the index, select- the - appropriate - -- ed in Hypertext and the capability to access the topic of the lecture and depress that key

index of each course via the menu selection pro- -review the pages of the lecture one at a . cess. Each course (only the FORTRAN cotirse is time, altering the contents of any pages as illustrated) will consist of pages linked together desired and selecting the next page by de- with the possibility of many routes through the pressing the RETURN key or function key 16 course materials depending on the user's responses. (the last page of the topic normally points to the index page) TRAVERSING PAGES IN HYPERTEXT -return to the entry page by going back 1 landmark Hypertext offers its users a variety of options for traversing its pages. This section As the instructor peruses the lecture contents, will not attempt to describe all the options but there is ample opportunity to thoroughly test out will try to convey the atmosphere offered by Hy- all branching situations (via the back page func- pertext. There are several aids available to the tion) and to alter any text that has changed or is users as they peruse Hypertext, including: incorrect.Also, the instructor can quite easily 1) the ability to retrace pages already construct additional materials and link those at visited. the appropriate point while traversing the course 2) the ability to place landmarks (either in materials. a temporary mode or permanent mode) on pages so that the user can directly return to a visited THE INSTRUCTOR ENVIRONMENT page of interest. 3) the ability to proceed forward through The previous section hinted at some of the pages. facilities available to individual instructors. The traversal of pages can be illustrated via the This section will explore those aspects related previous example (the instructor's Hypertext). directly to the teaching function and how Hyper- Consider a typical set of operations that an text can assist that function in most instances. instructor might wish to accomplish: Previous work has shown the delete utility advan- 1) add "grade COBOL exam" to the list of tages of incorporating the computer into the pre- things to do. paration and delivery of course materials (3). 2) drop "Ed Jones" from the roster of the The current Hypertext system significantly extends FORTRAN class. the previous work by providing the following 3) review the FORTRAN lecture material that additional functions: will be presented in class the next day. 1) the ability to traverse course materials We will assume the user has successfully logged in a very flexible manner(landmarks, backing up, onto the Hypertext system and is currently skipping, etc.). positioned at the top page. 2) the facility for embedding, within a page, a call to an external program thus providing for To accomplish an open ended system with respect to simulations, -select the "things to do page" by depressing demonstrations, etc. 1 3) a dynamic, easy to use organization tool -indicate the desire to edit the page (EDIT for creating and maintaining course materials. command) 4) a framework for combining materials to be -alter the page to reflect the added item presented in class with CAT materials to be taken (grade COBOL exam) by the student outside of the classroom. -back up 1 page (now positioned at the top The flexible approach toward visiting pages page) within the system allows the instructor to create course materials that suffice both for in-class To accomplish 2: presentation and for self-paced instruction. It -select the "Classes" page by depressing 2 permits the instructor to include additional -select the "FORTRAN class" Peg* by depress- material based on the response of the class. It ing I also allows the instructor to leave more detailed -indicate the desire to edit the page explanations, examples, and problems for the -alter the page by deleting the line contain- students to discover on their own outside of class. ing Ed Jones As as example of the above consider the pages -back up 1 landmark (the top page is an im? of information shown in figure 2, a short segment plicit landmark, thus we are now back to the of a FORTRAN course with branching Possibilities. top page In an'isi -Clips fxesentaiion; the*irstruCiOrOiii request, from the class, responses to the question To accomplish 3: contained in page 200. It may be apparent that -select the "Courses" page by depressing 3 the majority of the class understands the concept; -select the "FORTRAN course" page by depress- thus the instructor can select to disregard fel,. ing 1 Itup of the question and go directiy to the next block of materials to be presented, beginning at 33 Tools and Techniques 'for Instruction 23

-4311th-of-thr-folltivInn con a valid FORtean viesesslont (1) 23 (2) X4(4) (3) 2-1462

...4c 200

2.5 IS A ve114 0.1mess a Lou - It 2s a comment aftwt this is elpv. "Pg. 101-51 is valid even:440e le set page 14$ (1) for lovely an Assieneeot FORTRAN - see pose 145 the complete set of statement (1) fox the templets rut.,* set of expression roles

Mole 201 pap 20) pate 202

Logical Expressions

Vela

vase 204

Figure 2. Typical text in an instructional segment

page 204. However, there may very well be some Hypertext allows him to access information the students that do not understand the question and instructor has created in a friendly, reliable, interactions that transpired and do not stop the and flexible manner.Although the student does instructor to request additional information. not have the capability to alter this information, This segment of the population could turn to the he may get a hard copy of any pages that are CAt mode of Hypertext and review the question on reviewed. Of course, the instructor may permit their own. the student the ability to create pages. For Finally, from the instructor perspective, a example, consider figure 3 as a possible student highly reliable computing system is essential. view of Hypertext. The top page permits the So far, the experience with Hypertext has been student to organize each individual course of quite good: there have been no system crashes interest within Hypertext as well as his total during the approximately 60 hours of in-class educational environment. The student use of such presentations given to date. a facility, with easy access to a reliable computing resource, would indeed promote a creative STUDENT ENVIRONMENT and exciting environment.

Again, earlier work describing the class SUMWARY room use of computers and student benefits apply in full to the Hypertext system (3). Hypertext, as a tool to assist the presen- Briefly, these include: tation of information both in and out of class, ______1)_no.requirement for extensive note taking. has been presented.An-overview Of.the Hyper- 2) a clear, oonsistentprensentation medium. text environment, as well as its uses from both 3) an outside of class CAt facility for re- the instructor and student vantage point, has view of the in-class materials. been described. With Hypertext, as pointed out above, the ?unt- Use of the computer in the classroom has tions available to the instructor are significant- been slow to develop due to the lack of support- ly extended. The student profits as well, as ing software and hardware. A reliable computing system with a creative environment such as

34 24 NECC 1980

Top_page

Things to do

Current courses

Homework ensign. Wimp to du page

Acquaintances

Upcoming exam

Current courses page (00e pointer co each course) Upcoming cues p4$4 (this may alni2til pottererto study oec114om tor each exams) Homework saalgneenes Pate

Figure 3*Example Nypertoxt for a student

Hypertext should facilitate the use of computers 1977-1978* pp.21 -32. to assist the educational processes* Some areas* notably the reliability and interface* have been 4. R. Chong. "On-Line Large Screen Display extended from earlier work (3)1 however, there Systems for Computer Instruction." Proc. remain sone existing areas for future research. of ACM SIGCSE-SIGCSE Joint Symposium, The ability to provide a user-oriented graphic February 1976* pp. 189-191. facility embedded within a 'yet.; such as Hyper- text is an outstanding problem. Also, the ability 5. W. Trace."The Use of ATOPSS for Presenting to provide quality video display for a large Elementary Operating System Concepts." audience requires additional work. SIGCSE Eulleting, 711, February 1975, pp. 168-171. =MEN= 6. J. Rogers. "A Computerised Classroom for 1.T. S. Nelson. "No More Teachers's Dirty Looks." Instructor's Experimentation and Training."

_Cteputer.Decisions, September-_1970. - Commuters in,Educationi-0.-Lecarme and R. Lewis (eds.), IFIP -North Holland Publishing 2.Tandem Corporation.Programming Manual. Co.* 1975. Cupertino, Ca. Noveeber 1977.

3.D. L. Ward. "A Computerised Lecture Prepara- tion and Delivery System." Journal of Educa- tional Technology Systems, Vol. 6(1), 35 Tools and Techniques for Instruction 25

A DYNAMIC PROCESS IN TEACHING TECHNIQUES by Mall E. Moth, Ph.D. INFORMATICS INC. 311M Yuma Sheet Conga Path, Caffein* 31304 (213) 1117-9131

ABSTRACT training customers. Their work begins early in the design Programs to teach people how to use computer of the product; software documentation and their teaching products (hardware and software) often do not meet the techniquesare consideredintegralpartsof the needs of the diverse audience of users. The needs of one oduct package. group may vary considerably from the needs of another. A Duringthecreationof learningactivitiesat lag exists between-the hydra-like growth of the computer Informatics,the education development specialist is beldame and software industry, and the technology for involved in a set of relationships and interactions that can teaching customers. This paper describes a practical best be illustrated as a network. This network design Inaba technique which has proved successful at includes a systematicprocess for seeking relevant Informatics Inc. information to improve both course designs and teaching 'Teaching designs are based on identified procedures. techniques. Figure 1 represents this process of continuous This approach Is represented by a network of relationships interactions and decision making. and interactions. The process starts with recognizing the audience segments for whom the courses should be IDENTIFYING THE USERS developed, specifyingtheir learning objectives, and In the computer industry, hardware or software planning a design customized to user needs. Brainstorming products are manufactured to meet the needs of a wide for ideas and selecting the most relevant comes next. After spectrum of users. The first step in teaching techniques is the form of presentation is drafted, at least two dry runs to identify the audience segments of this spectrum the are conducted before representative audiences. Feedback people for whom the courses should be developed. For Is collected, analyzed, and evaluated; results are integrated most practical purposes, two major audience segments, into the evolving course. A field test at a customer site is the end users and the technical support personnel, can the next step; more feedback is collected, analyzed, and be defined: evaluated. Finally, the validated responses are integrated 1. The End User Audience.This segment of users into the courseware to produce a tested and useful usually includes managers, clerical personnel, educations! program and casual programmers. The end users are, in general, not a data processing - oriented group INTRODUCTION becausetheirprimary interestisintheir Helping customers learn how to use software particular job; data processing is only a means to products is an important element in the design of those some other end. products. Some customers need technical training to 2.The Technkwl Support Audience.This segment install, implement, and support a product at their facility; includes the support personnel, mainly data base others need to learn to use the product for reporting or administrators (DBAs), system designers, system processing data. Meeting those need!, Is the job of programmers, -sad application programmers. specialistsIntheProduct Depending on the level of complexity of the product, a Communications Group of Informatics Inc. plan should be developed for subdividing each audience The Product Communications Group Is a part of segment into smaller segments with users of similar Technical Product Support; it Includes technical writers, backgrounds and job requirements. editors, graphics designers, and education developers. These communications specialists work togetherto document software products and to design courses for

36 26 NECC 1980

THE DYNAMIC CYCLE OF EDUCATION DEVELOPMENT

Figure I

37 Tools and Techniques for Instruction 27

--IDENTIFYING USERS' LEARNING °enema review committee meeting, the education specialist reviser.- The second step is to identify the learning objectives the form and content of the evolving course and rebuilds it which reflect the needs of each audience segment. For the into a coherent design. two major groups of users, two sets of objectives are identified. COURSE MEDIA The end users need to know how the product can be The media used in teaching are generally: handouts, used to their benefit and how to obtain the information overhead projector tansparencies, the board, and the which answers job.related questions and solves their instructor's guide. This guide should be prepared carefully problems. The following set of learning objectives might to provide consistent instruction. Wall charts are also be developed to reflect end users' needs: importantcoursemedia;theyaredevelopedfor To obtain job-related information. continuous reference during class session. Examples of To acquire a basic level of skills essential to the wall charts used for IMS/VS-envirorament data structure use of the product. are shown in Figures 2 and 3. Figure 2 is intended for the To explore and identify the capabilities of the technical group who normally looks at a hierarchical product for solving problems. structure of logical statements from top to bottom. Fixate 3 is intended for the end users, presenting the same The technical audience needs to know how to solve more complex problems bow to keep the product concept in an approach that deals with building a running efficiently, maximizing the benefits to their relationship within a framework of reference, a readable organization. Tice learning objectives developed for the left-to-right flow. technical support personnel might be represented by the following: TIME AND PLACE CONSIDERATIONS To understandthedifficultconcepts and Time and place must be considered in course procedures which dcn1 withtheactivities development techniques. The duration of the course, when associated with the product. in the sales/installation cycle it will be taught, and where it To explain and, where necessary, customize the will be taught all affect the design of the course. The product to meet the end users' needs. education specialist has to ask some questions about timing: Is the class needed before the installation of the To identify the specific security requirements and product? If so, this may give the technical group an the constraints that can be imposed upon the user opportunity to become aware of what is expected during for efficiency or security purposes. the installation procedures. Should it be taught during installation? When doesthe end user need the PREPARING THE DESIGN PLAN information? The education developer also has to consider The educational courses are developed to satisfy the location: Where will the course be taught? Will students learning objectives which reflect the needs of the identified have access to terminals? Will actual data bases be used users. The procedures start with gathering information. for examples? The choice of customer site or a regional Tice education development specialist interviews a diverse office has impact in these areas. These decisions are made cross-section of specialized personnel; they contribute by productmanagement, marketing,andproduct ideas and techniques relevant to course development, communications after considering both technical and which might shape the product learning process. People's marketing needs. contributions are based on personal experiences in the field during interaction with similar users or on ideas from published sources, or they are gathered from colleagues. IMPROVING AND EVALUATING THE COURSE The neat procedure is analysis to find out how to fit The process of course development starts with laying together the collected ideas, how to look for the simplest out a preliminary desks resulting from selected ideas and and most relevant approaches to be used in presenting the committee discussion. To improve the course, dry runs are capabilities of the product to the users, and how to conducted before an audience representative of the group produce a design covering all of the important and for whom the courseware is being developed. These dry practical aspects of the product. Once the overall design runs help determine the effectiveness of the instruction, pima is established, skills needed to accomplish each which is measured by the feedback collected after each dry objective are specified and written down to form the basis run. Changes to the course are introduced, and the course for group discussion. In industry, they refer to this group is retested in another dry run before a new audience; more as the document specification review committee. The feedback is collected, evaluated, and compared to the committee members are representatives of the same results collected from earlier tests. This comparison groups who contributed ideas. They are called to review measures the degree of improvement. and comment on the design plan and the proposed specifications. As a result of the document specification

I ; 3 s SAMPLE USER DATA BASES

PLANT SKILL FLANT.ID. SKILL.CODE. FLANT.NAME SKILLNAME LPLANT FLANT.PHONE PLANT.REGION

...whl. PROO.CDOE. EMP.NO. pLANT.io' PROD.DESC EMP.NAME EMF MOD PROO.AMT EMP.SEX PRODAITY

I--. SALYEAR. ED YEAR' EIN.NO. SALYTD EDDEGREE ED SAL SALDED ED.SCHOOL

SUO.NAME. SUS.GRADE SUB

'DENOTES SEGMENT KEY FIELDS

Figure 2

39 Tools and Techniques for instruction 29

0 qb

BUILDING A FRAMEWORK (defining relationships)

PROM CODE PRODUCT PRCS DESC . PROD. NAT PROD. CRY

PLANT. ID Karr mow.NAAE SALYEAR PLANT. PHONE SALARYISALYTD PLANT. REGION SAL.DED

DAR NO OWLOYEE Di? NOE

YR ECUCARON ED.DEGFEE ED.SCHOCI.

SWECT 'MEN SUB. NAME SUB.ORAD SKILL PLANT SIP SKILL. CODE FLAME, (mew SU.. NAPE

Figure 3

40 30 NECC 1980

VALIDATING AND INTEGRATING THE 1. .USION INSTRUCTIONAL DESIGN are products and their instructional activities In industry, product communications group should not should be developed to work for the users' benefit. Users release the course to the field before a field test at a are encouraged to be a part of this dynamic process of customer's location. After teaching the course to the education development. Their feedback is needed in order intended audience at a customer site, the developer and to learn how the product and its learning process can be course instructor collect written and oral comments that made to work better. aid the developer in determining the degree of success in Product educ. ion development techniques and the achieving the identified objectives for the course. The field evaluation cycle have no time limit. They start with test itself indicates that the minimum evaluation effort in brainstorming but never end as long as the users are active producing a product learning activity is accomplished. in monitoring the product and its educational programs to Responses produced from the field test reflect the attitudes service their job-related needs. and feelings of the users to the course learning materials. Inthe process of validation, some of these responses are considered very useful to fieldinstructors and are integrated in the instructor's guide to assure that field instructors are aware of users' expectations. Other responses may lead to the elimination of misconceptions in the course presentation. These changes are made before the alai release of the hinting document; however, no attempt is made to customize the instructional activities to every individual customer's needs. Examples arc set to a typical and general application familiar to users. These examples can be modified, based on users' feedback, to become more realistic in the final touches for refining the courseware product.

SUMMARY In summary, the instructional design technique that produces a successful learning activity should follow these criteria: 1. The learning activity or class material is designed to achieve specific learning objectives. Mainly, these objectives are to impart the knowledge of the product capabilities to the users and to enable them to use the product efficiently. 2.The learningactivityis designed tooffer replicable instruction, that is, instruction that can be taught in the field setting by any of the field instructors. The instructor's manual serves as a guideline to assure consistent instruction. 3.The learningactivityisdesigned soits effectiveness can be tested and demonstrated. Typical problems, based on user needs, should be solvedinclass.Hands-onpractical application of what has been learned should also be provided.

4r Tools and Techniques for instruction 31

CONSIDERATIONS AND GUIDELINES FOR DEVELOPING BASIC SKILLS CURRICULUM FOR USE WITH MICRO- COMPUTER TECHNOLOGY

Robert M.. Caldwell Division of Educatior.:2 e.t4dies Southern Methodist University Dallas, Texas-76276 (214) 692-2347

The availahil. of low cost micro- then display questions to test the student's computer technology is creating a revolu- comprehension of that text.This sort of tion in education. Institutions of all of instruotion has its use but implies to types can now take advantage of the many most educators that the computer's power benefits offered by computer-based educa- lay only in its ability to present text tion at a cost that is easily affordable and ask questions.This form of instruc- for most. In addition, advanced micro- tion serves to reoreate the very worst of processor technologies interfaced with a what presently ocours in a traditional wide range of audio-visual devioes in just classroom OUR ignoring all other teach- a few months have increased the oapabili- ing strategies coat oan help developlearn- ties of micros to include color, graphics, ing styles and learner independenos(Garson, animation, and music and speech reproduc- 1980). tion. In short, microcomputers have made To complicate this problem further, available an extremely fldxible and power- some oompanies are now developing author- ful teaohing medium at a prioe that is ing systems which will allow classroom finally cost-effective for most users. teaohers, students and curriculum devel- In response to this growing interest opers to create their courseware through in the use of microcomputers in instruo- a process which utilizes templating and/or tion, several major publishers currently menu selection.To be sure, many edu- offer limited curricula in basic skills oators will use these processes to create for delivery on microcomputers. A num- exciting programs which utilize the ber of school districts such as the Dallas system's capabilities to its fullest. Independent School Distriot and those dis- Many others, on the other hand, will tricts affiliated with the Minnesota duplicate the type of programs mentioned Educational Computing Consortium also have above; they will need help, guidanoe and developed materials whioh cover a variety education about what microcomputers can of skill areas.Because of the signifi- do and how instruction can be presented cant expenditures associated with develop- to take full advantav of the unique ing these programs, however, most of them features of the new technology.They are limited in scope and employ a rather will need explicit guidelines that will narrow range of teaching strategies and help them devise ways to make contacts machine capahilities. In addition, few between the learner and the learning of the individuals ourrently engaged in experience more meaningful, more effi- instructional deveW "nct have had much cient and more productive.The purpose experience in using Highly interactive of this paper, therefore, is to present medium like the Mi4 umputer. As a re- specific guidelines for designing instruc- sult, much of what passes for courseware tional programs that will be delivered today neither helps to develop higher on microcomputers so that those programs level cognitive skills nor challenges will use the capaoity of the microcomputer learners to use higher order learning system in a way that will develop in strategies. Most lessons utilize a drill learners a range of cognitive skills and and practice format in which thecolputer help learners develop useful learning asks a question and requires the student strategies. to respond. Many so called tutorial programs are no improvement. They merely present segments of expository text and 32 NECC 1980

ties allow for precise diagnosis of GENERAL FEATURES OP PROGRAM DESIGN skills already mastered, remediation in One of the most important factors skill deficiencies, and exact evaluation inherent in programs delivered on com- of learner progress. puter-based systems is their ability to S. Progress should be measured in adapt instruction truly to the individ- terms of mastery of performance objec- ual needs of each learner.With this in tives. mind,then, programe of computer -based 6. Strategies for dia nosie and instruction should include the following 4 prescription should be used. e effic- general features (Caldwell and Rizzo, iency of a program is due primarily to the 1979): diagnostic inventory made of the ekills 1. Learner Control over the of each learner. This information can instructions iiWaia-re a feature of then be used to place them appropriately program design often ignored by instruc- within the curriculum and to direct learn- tional designers. Certainly there are ers to the instructional material most situations in which a presentation of Appropriate to their needs. instruction must be linear. Certain sub- 7. Programs should be, when possible, jects and concepts require it. In most multi - sensory in format, other cases, however, consideration should he given to allowing the learner SPECIP:T GUIDELINES POR INSTRUCTIONAL as much control over the learning DEVELOPMENT sequence as possible. Options should be Programs of computer-based education incorporated into the instructional have been developed in a variety of de- sequence which allow for review of pre- signs and formats. Some use drills arrang- vious frames; for decisions about the ed in stionds while others are built around type, difficulty level, and number of a series of tutorial leesons. Many even problems or exercises received; and for incorporate all or most of the features alternative branching routes that might mentioned above.Within these programs, lead-to the accomplishment of lesson however, are characteristics of instruc- objectives in less time. In short, tional design that heavily affect the students should be given the opportunity success of the instruction. The following to advance, review, and'exit lessons ie a description of some of the more com- except where such control defeats the mon characteristics of lesson design that purpose of the lesson. This ability of can contribute to instructional effec- learners to pace themselves provides a tiveness and some that can seriously de- degree of individualization not present tract from it. in purely linear programs. Creating Text and Graphic Displays 2. A system should be totally Many alternatives can be used to break individualized and offer highly adaptive the monotony of lines of text filling an and res onsive learning environments. entire screen: By a ow ng self- pacing adk. individ- 1. Use graphics to box important sen- ualized branching, learners are helped tences or paragraphs. Boxes made of lines to select the pathway through the or keyboard characters do nicely to alter eateries that is most appropriate for the visual display of text on a monitor. their needs. Reverse highlighting ae. color can also be 3. Programs should be modularized used .effectively to accentuate visual dis- and structured in coherent, hierarchia plays. In addition, microcomputers can be patterns. This type of organizational connected to graphics tablets which make pattern allows for great flexibility extremely interesting displays in almost in program implementation because any size or shape. curriculum materials specifically ad- Whatever is used, it is crucial to dress each necessary skill in a defined interrupt continuous lines of text on content area in a manner that provides screen so that the screen does not look for development of skills not mastered crowded and cause verbal overload in stu- or allows bypassing of skills which dents. Too much text can have the effect have already been mastered. This proof discouraging the learner, especially if cess can reduce student frustration he/she is a poor reader. and increase curriculum effectiveness. 2. Allow the student control over the 4. All skills to be mastered sequence of presenting text by breaking should be carefully stated in or- large portions of the text into discrete formance objectives. The accurac units or segments which the student can El-i7W.Biram is based on the spec fie call up in sequence by merely pressing the definition of the objective in terms space bar or some other key activated for of performance competencies.Activi- this purpose. This procedure serves two Tools and Techniques for instruction 33

purposes: it allows learners to read at a Figure 1 rote that is appropriste for them, and it PREDICATES breaks the reading task into small segmeats eo that the reader ie not overwhelmed by Which mould you like to study page after page of text on the screen.

3. Animations, graphics, cartoon char- 1. Predicate Nominative acters and other creative devices serve to create variety end interest in the display 2. Direct Object Which appears on the monitor. Used creatively, these capabilities of microcompu- 3. indirect Object ter systems can contribute greatly to effective instructional programs. 4. Predicate Adjective 4. Double spacetext material whenever poeeible to enhance the visual effect. 5. Review 5. Use color to enhance the display or One pro- to highlight' whenever possible. a. Linking verbs gram uses color-coded feedback to distinguish it from other text and to emphasize b. Action verbs key concepts. Color is also useful in providing prompts and to direct attention to various portlons of the screen. c. Adverbs Creating the Instructional Sequence 1. As a general rule, try to show Choose a number or letter learners rather than tell then.TU.over- use of exposition is the single biggest mistake instructional designers make. Competition in the games can be They feel as if they suet write a lecture effect. into each frame or prOline complex direc- against other students, the computer it- tions, instructions or explanations when self, time limits, or performance criteria set by other students (e.g. "the best score they can very simply use graphics or the " or "the beet ability of the computer to erase, rewrite, on this game to date is flash and even animate to make concepts time for this game ie 613". clear. (The presentation which accom- A word of cautiorli-appropriate here, panies this paper illustrates this point however. In en attempt to bring novelty to with k number of examples.) their programs, some designers have defeat- 2. Make lessons as interactive as ed their own goals. In one such leeeon poeeible. Force learners to Make students are encouraged to save a tiny man choices; help them make decisions by who is standing in a deep pit from a large providing them with options and altern- heavy stone which ie slowly rolling down a This atives.Simulation and dialog programs hill on the left side of the pit. are the beet instructional strategy for teat is accomplished hy solving $ Series With each promoting interaction, but it can be proof math problems successfully. vided through other means as well: correct response the man moves up the right a. Menus side of the pit toward freedom. If the Learners may choose from a variety learner fails to get the problem correct, of options within a leeeon or within a however, the stone rolls closer and loser program by making choices from menus. until it eventually falls into the pit and These menus allow the learner flexibil- squashes the poor fellow. In an observa- ity to pursue his/her interests or to con- tion of this sequence with several child- trol the sequence in which topics are pre- ren, a number of them chose to see the man sented. For example, s language leeeon get squashed with the full knowledge that Which deals with predicates might allow to do eo they must fail every single math students to access concepts which have no problem presented to them.This exercise, particular instructional sequence. Figure therefore, did little to develop math skills 1 illustrates Just such * menu. but did much to distract the learners from the true intent of the leeeon. b. stftticerfarmatie Learners SIMMUAIMI Oven a variety of c. Prompts activities and choice Of content. Tutor'- The power of computer-based instruc isle should be accompanied by creative tion resides in its ability to ehape learn- drills or instructional games that rein- er behavior toward learning outcomes in a force skills and intonation and enhance way not possible with most other media. An the interactive mature of the instruc- important factor in shaping hehavior ie the Designers who ignore the tion. These games and drilla can stimulate Use of prompts. sotiVation hy capitalizing on a novelty use of prompts often turn the instruction

44 34 NECC 1980

into a guessing game. For example, a number of math programs currently available present time consuming especially if learners are problems then simply respond to the learn- unfamiliar with the keyboard. This problem er's incorrect response with a "No, try can have the effect of distracting learners again."Thie type of feedback gives the from learning the concept being presented. learner no information about how he/she ie A related problem ie that many errors in doing; it only promotes guessing until the typing are judged by the computer to be incorrect responses, and therefore etudente correct answer is discovered. Prompts such as, "Too high, try again" or "Remember, are routed to remedial sequences or to seg- carry the ones," on the other hand, pro- ments of the lessons they have just been vide learners with guidance toward the through. Similarly, if typed responses are not correct answer. Without them, interaction becomes meaningless. More about prompts carefully cued, learners will often be for- will be said later in this discussion. ced into closed loops; that is, they try to d. Task Description find the correct answer by typing random responses. Learners are very cation misled about When this problem occurs learn- what it ie they are supposed to do or how ers have little idea about what is expected they are supposed to respold in a given of them and become frustrated trying to find the response that will advance them to the exercise because the instructional designer Thie common error in instruc- has confused the task for them.The follow- next frame. ing i4 a perfect example: In this lesson tional design has an extremely negative the learner is presented with the task of effect on learners. Some advantages can be found in the distinguishing complete and incomplete typed response. however, which are not pres- sentences: ent in forced choice reeponee modes (Cald- The store closed early. well, 1974): Type comp (for complete)or inc(for a. Learners seem to express more favorable attitudes toward writing and the use incomplete) of verbal language. The task here ie clearly discrimination b. Spelling skills and the ability to between complete and incomplete sentences, generate language improve. but it ie confused by asking the learner to c. Personalized feedback ie made poss- type "comp" or "inc."The instructional ible through direct entry of student names. sequence is marred not only by a poorly 2. As mentioned earlier, response by designed drill and practice but also by a direct entry should not be required of a poorly conceived response format. An im- student without an example or a practice proved version would simply ask the student exercise. Thie can prevent learners from to respond by typing a 1 for complete (or a making errors in the practice exercises. c) or a 2 for incomplete (or an i), thus These errors can become important, particu- focusing attention on the discrimination larly if one ie using response data to task rather than the typing task. branch learners or to do item analysis. Task confusion can also be lessened or 3. A number of other helpful suggestion* eliminated by providing learners with a can be found in the guide to developing ins- sample exercise before they are engaged in tructional software developed by the Minne- a drill or a quiz': sota Educational Computing Consortium (1980). Input and Response Some examples from that guide follow: a. Be consistent in the way questions 1. The method an instructional designer chooses to facilitate learner im- are asked and the required format for res- Do not code answers unless it ie put and reeponee ie largely a function of ponses. the type of learning outcome desired.Res- necessary. Request the responses of YES or ponse modes most commonly include typed res- NO rather than coding responsee such as 1 When processing these answers ponse, touch reeponee or light pen response. YES and 00NO. The first criterion for choosing one of these check the first character to determine if it modes is whether the response will help ie "Y" or "N". If it ie neither, erase the student's answer and re-ask the question. reach the objective of the lesson.One b. Ask questions while the information illustrationof this rule is in the example cited earlier in which etudente are needed to answer them ie still on the screen. Don't ask a question then change the screen asked to discriminate between complete and to ask another one leaving the choices for incomplete sentences. In this case a typ- If ed reeponee was inappropriate to the objec- response on the screen jut erased. the student must choose from a fixed eet of tive of the lesson. In fact, typed respons- options, then those options should remain on es in general seem prone to high error rates the screen throughout the quiz sequence. In because of the nature of typing tasks (Cald- well, 1974). Typed responses also become some cases it ie helpful to allow etudente to go back to formerly displayed frames to Tools and Techniques for Instruction 35

refresh their memories. This review ie this type of response can delight 'earners, accessed by means of the WNW key. its overall effect ie one of distracting c. Put options for a question in a them from the purpose of the lesson. COIMID. Do not imbed them in a sentence. Options having multiple words should be RWinforceftent coded with numbers or lettere. -Considerable space has been devoted in d. Answer checking after accepting'inthis paper to suggestions regarding responses to various types of imput. The section putdeserves specific consideration.Don't ask a question and expect a very specific dealing with prompts serves as a good ex- Generally, reinforcement should be answer. For example, if the question ie ample. "What is the capital of Minnesota?" some specific and directed at providing informa- learners might answer, "The capital ie St. tion that will help shape the learner's be- Paul." If the response judging is too rig- havior toward the desired learning outcome. id and is looking for just "St. Paul" this As an illustration, consider the following sample sequence: student's answer would be wrong. It ie therefore very important to do keyword ing to the word- - matches rather than a match on the whole Come) answer entered. (Here students are required to transform Also, when looking for correct answers, the word by retyping it with an -ing end- consider how many times to allow a student ing.) A correct response on the first to miss a question. A typical method used try would warrant positive reinforcement by may designers ie to ask a question a such ae "great, super, well done, excel- maximum of three times. This way, the lent."These and other reinforcing state- learner has the opportunity to be prompt- ment(' can be generated randomly from a ed twice. Any more than three tries seems list of 20 or 30 possible statements. In- to indicate that the student has not mas- correct responses on the other hand, tered the concept and it ie beet to move should follow a pattern of prompting which him/her on. However, it is a good idea should lead the student to the correct to ask the same question again later. answer. For example: 4. The number of questions presented First incorrect response: Come) during a formative evaluation ie the sub- giggling (drop the e).The student ie ject of considerable debate. Essentially, tiedlfrj reminding btm/her of the rule the number of questions presented in any under examination (drum the e). one exercise ie dependent upon the objec- Second incorrect response:Come, tive being assessed. As a general rule, comiing (come + ing). Here the learner however, five (5) ie a number that seems different response. In the workable for most learners. It ie a second prompt the computer system animates sufficient number to assess progress to- the a in come and it drops from the equa- ward mastery of the objective and requires tion then-f6W-ing moves into place. just enough time so that the exercise it- Third incorrect response: Come, self does not become tiresome for the gmipx. If on the third try the student learner. Five items ie a good number for still not type the correct response, generation of item pools also if one uses the correct answer should be given, and the guideline that three times the num- the student should then move on to the ber of items should be written as are next problem.This type of reinforcement presented in the exercise. pattern contributes to meaningfulness of 5. One must also be wary of the type of the material to be learned because it: information learners are permitted to in- a. Provides specific information put. For example, a common practice in that helps guide correct learner behavior many computer- ascieted lessons is to sek, toward achieving the desired outcome. "What ie your name?" or "What do your b. Reduces the frustration often friends call you?" The purpose of such a experienced by learners in CAI programs question, of course, is to use this infor- that simply provide a "no" response to an mation laterto provide personalized incorrect answer. A simple "no" ie un- feedback or to use the learner's name in satisfactory because it provides no the content of the presentation (e.g. specific feedback that helps the learner "MIMO to: Johnny"). This practice can some- to discover correct responses. Instead times have disastrous effects if the learners are forced to guess until the learner decides to get cute. One student, correct answer ie found (Caldwell, 1070. for example, was observed to enter "idiot" Reinforcement strategies depend to just such a question as those cited largely on the nature and type of learning above. All successive feedback after that being attempted.Reinforcement, however, resembled statements such ae, "No, Idiot, can be made more meaningful if it is per- try again" or "Good work, Idiot." While sonalized and specific to student

1 4V 36 NECC 1960

response sequences. (e.g. "You did well, Educational Computing Consortium, Allen, but would you like to review pre- February 15, 1980. fixes before going on?") In summary, the design of instructional programs should incorporate various teaching strategies to add variety to an overall curriculum and to address cognitive processes at all levels of that domain of learning. Also, programs should use to their best advantage the features of color, graphics, and. animation offered by microcomputer technology. Exciting new developments in speech and sound periphekals also exist and should be incorporated where relevant and appropriate. Also, learner control over the instructional sequence allows for individual pacing and better opportunities to achieve mastery through branching, diagnosis and remediation. This flexibility can add significantly to computer based programs if they are designed and implemented carefully. The scope of this presentation limits discussion of the many subtle variables which contribute to effective instructional design, but attempts to begin an organization of some of the features which seem to contribute to successful programs. It is hoped that it might stimulate other authors and designers to share their rationales and experience in designing high quality programs of computer-based education.

REFERENCE NOTES Caldwell, R.M. "The Effects of Selected Strategies for Teaching Reading to Non-literate Adult Learners Using Computer-Based Education." Paper presented at the annual meeting of the American Educational Research Association, San Francisco, April, 1979. . "Evaluation of a Program of Computer-Assisted Reading Instruc- tion for Semi-literate Adolescents." Paper presented at the annual meeting of the American Educational Research Association, Chicago, March, 1974.

REFERENCES Caldwell, R. M. and Peter J. Rizzo.. "A Computer-Based System of Reading Instruction for Adult Non-readers," AED8 Journal, Summer, 1979, 12,4. Gerson, f7-177TWe Case Against Multiple Choice," The Computing Teacher, February-March, 1980, 7,4. Minnesota Educational Computing Con- sortium User Services, A Guide to Developing Instructionanare for the Apple II Microcomputer, St. Maul, Minnesota:The Minnesota

4 7 Invited Session

RESEARCH ON MICROCOMPUTER USES IN EDUCATION

Chaired by David Kniefel New Jersey Educational Computing Network

THE AFFECTIVE AND COGNITIVE EFFECTS OF THE EFFECTS OF MICROCOMPUTING ON LARGE MICROONNPuTER-BASED SCIENCE EDUCATION CENTRALIZED TIMESHARING

Ronald E. Anderson, Daniel L. Rlassen, Kent T. Rehrberg, Minnesota Educational Thomas P. Hansen, Minnesota Educational Computing Consortium Computing Consortium

ABSTRACT ABSTRACT Microcomputers are used increasingly imiesota schools have already acquired to deliver and enhance science instruc- nearly 1,000 APPLE II and other micro- tion, but the impact of these new computers, even though they have had technologies has not been extensively access to a large central computer via studied.An experiment was designed to a statewide educational network. This investigate the effects of a brief CAI situation offers opportunity to research activity on student attitudes, beliefs, some important questions such as: (1) How and knowledge. Three hundred fifty do people justify their acquisition of high school students were tested and re- microcomputers? (2) How are micros used tested before and after taking a 20-30 in the schools? (3) Does microcomputer minute lesson on water pollution.The acquisition substitute for or foster time- instructional material was all delivered sharing use of a centralized facility? by an APPLE II microcomputer using high The MECC research serves as the basis for resolution graphics and some color vari- projection of the role of microcomputers ations. The findings support the claims and instructional computing in the short- that: (1) even a very brief CAI exposure ' term future. can be instructionally effective; (2) understanding about computers, i.e., DISCUSSANTS computer literacy, can be improved as a consequence of such CAI: (3) students Harold Peters, Associate Director, CONDUIT become attitudinally more positive about John Castellano Jr., Indiana University computers from microcomputer CAI: (4) graphic enhancements may not improve student responses; (5) system malfunctions may revise student conceptions of themselves as well as computers. The micro offers a vast potential. with some possibly questionable effects. Since it allows new kinds of person - computer relationships, we can not presume to know its ultimate impact on individuals.

3748 Computers in Humanistic Studies

CREATIV/TY THROUGH THE MICROCOMPUTER

George M. Bass, Jr., Ph.D. Assistant Professor School of Education College of William and Mary Williamsburg, Virginia 23185 804-253-4289

"Creativity is a battle tivity; and finally, to increase their against fixed attitudes." own creative abilities. Eleven students (Raudsepp, 1977, p.55) took .the course.

In many ways teaching a college course COURSE ACTIVITIES on creativity is very similar to teaching In order to distinguish this course students hdw to use a computer. Both sets from just another elective, I wanted to of students come with many preconceived use activities not typically found in notions about the subject. Usually these other courses they had taken.Entering attitudes have not been built upon direct national contests, participating in crea- experience with computers or creativity tive growth games, working on word puz- experiences but on hearsay and vague ex- zles and logic problems, reading and solv- pectations.This lack of experience has ing detective stories, producing poetry often led students to be fearful of their and written compositions, and using a own ability .o be fyeative or to master microcomputer were introduced throughout the computer. Yet a willingness to break the semester. (More traditional activi- free from theba frozen views and to see ties such as assigned readings, class ideas and problems from a new perspective presentations and projects, and a final is needed in both areas for the student position paper were also used to increase to succeed. the learning process.) During the fall semester of'1979, I Because the microcomputer is such a taught an advanced education course on new educational technology, it is an creativity.This course was designed to ideal way to break the chains of student be a seminar for students who had success- expectation and release any hidden poten- fully completed an introductory education- tial. Yet the fear of the unknown or the al psychology course. My overall goals exaggerated visions of computer power were to present students the necessary (such as HAL in 2001) were a real concern. background about creativity as it has In order to redarfhis possibility, I been studied by psychologists and educa- began the first class with TLC -- Tender tors; to introduce them to instructional Loving Computer. Other exercises using strategies which may help develop crea- the capabilities of our Apple II micro-

49 38 Computers in Humanistic Studies 39

computer were later interspersed through- overall structure but required the addi- out the course. tion of certain words and information from the user to compose a complete tctle of how MICROCOMPUTER ACTIVITIES -- the Apple II spent its summer. Besides OBJECTIVES AND RESULTS showing how unique paragraph-length sto- 1. Course Syllabus ries can be created using a microcomputer, Since this was the first time the this exercise laid the groundwork for a creativity course had ever been taught, future class discussion on form versus I felt it necessary to give students my content in creating something new. course objectives, assignments, and re- While the students' creative solutions quirements during the first session. to this assignment were indeed varied (a However, I wanted to communicate this picture scrapbook, a summer job advertise- basic information in a nontraditional ment, "first grader's" paper, fantasy way as mentioned earlier. I also wanted story), their response to the Apple II the students in the class to meet one story was enthusiastic.They worked as a another and to learn of any personal ob- group supplying the adjectives and other jectives each individual might have for words requested by the program.However, the class. To accomplish these objec- the first time through the story, they had tives, I wrote a program on the microcom- no idea how these words were going to be puter to interact with each student. used. The resulting narrative was a hu- After requesting the student's name and morous, if not entirely accurate, account some background information (interests, of what a microcomputer might do over the reasons for taking the course, etc.) and summer on a deserted college campus.The engaging him in chit chat, the program second time through the program the stu- presented relevant information about the dents chose their words more carefully to course and then introduced the student fit into the structure of the story they to the capabilities of the microcompu- remembered.While this second experience ter. Using a variety of programs cur- resulted in a new version of the tale, in rently available for the Apple II, the some ways it was not as entertaining as students were presented with computer the first version. In resulting class graphics, text manipulation, and game discussions the notion of appearance ver- playing experiences, e.g., Rock/Scissors/ sus substance, "how said" versuswhat Paper, Dragon's Maze, Star Trek. saidilith,are role of the creator and his The results of this experience were audience in giving meaning to a creation quite positive. The students had never were all related to this computer/human- seen a microcomputer before and were generated story. surprised with the variety of things it 3. The Apple II Pops could do. As each student took a turn, The apple II microcomputer can make the rest of the class crowded around the music-like sounds through the Apple's color monitor to watch the action. They built-in speaker. A number of commercial got to know each other through the typed programs are available to create these conversation with the Apple II.They songs with minimal programming. Using the also got to see many of the programs FORTE Music Interpreter by Rainbow Comput- available to them outside of class. (For irs Inc., I showed the students how to two students this introductory session program and save songs they composed using led to a semester-long journey with Star the Apple's speaker. I wanted the stu- Trek after almost every class!)This dents to construct a nonverbal creation, computerized syllabus generated interest to become more comfortable with using the and set the tone for an unusual class. microcomputer as a tool, and to learn an 2. "How I Spent My Summer Vacation" introductory programming approach. Assignment At the class concert during which each One of the most common assignments aspiring composer introduced his creative that teachers give when students return effort, it became readily apparent that my to school in the fall is to write a theme objectives were achieved. Although the on how the student spent his summer vaca- variety of songs (from birdlike melodies tion. In order to get the students in to atonal experimentations to K-Tel music the class to begin thinking of novel ways advertisements) again showed the diversity to solve problems or accomplish tasks. I in the class, all students completed the gave the class this typical task, but assignment and expressed their growing asked them to solve it in a creative way. ease with the Apple.They commented on When they next came to class with their the computer as most forgiving, but also creative solutions, I shared with them most demanding. Mistakes were easy to one solution using the microcomputer. I oorrect by retyping the note or line, but programmed a story which contained an putting commands in the correct syntax was

50 40 NEOC 1980

required for the program to work. In ad- and simulations should also warm-up stu- .dition, the students expressed an appre- dents and increase their anticipation. ciation of each other's efforts regardless Stage 2 burns this initial interest into of previous musical experience. deepening expectation. Song development 4. Aphorism Construction is one example of the yellow; information Adapting an Aphorism Generator pub- processing patterns which Torrance sug- lisheAby J.D. Robertson (Creative Com- gests. Stage 3 simply tries to take crea- utin , August 1979), I programmed the tive thinking and keep it going. Seeing pple to construct aphorisms such as the various capabilities of a microcompu- "Alcohol is the liver of stress" through ter in this course should permit students random combinations of lists of the three a better understanding of future .techno- key nouns.The class was asked to genera logical applications in their own lives. ate sayings using the format " is the Although psychologists have perhaps of ." as well. Pour stints were focused more on the instructional demands chosen to do this individually, Is of of creativity and problem solving courses two and three students were chosen to do than other subjects, the psychological it collectively, and the remaining two principles underlying such instructional students worked as a team to select those planning and design van be generaliz..d to aphorisms generated by the computer which other college subjects as well. Treffin- they thought most creative. After each ger and Huber (1975) have emphasized that individual or team had chosen their best the content of any course can be analyzed two sayings, these aphorisms were read to according to an instructional systems ap- the rest of the class without identifying proach. Such an approach focuses on iden- the origin. Each class member separately tifying specific instructional objectives; ranked the three best aphorisms; these diagnosing characteristics of entering rankings were then tallied and the high- learners; developing sequential, learning est rated sayings identified. My objec- hierarchies and teaching strategies to tive was to stimulate a discussion of accomplish the objectives; and assessing individual, group, and random efforts in performance to determine learner achieve- creating original products. ment. Following such a basic instruction- Although the results were certainly al model will allow any teacher to esti- not definitive (one pair had written both mate the value of particular educational of the top ranked sayings), they did lead technologies in a chosen course. to alively debate on the difference be- The specific issues for incorporating tween random replacement and meaningful the computAr into this current educational combinations. The idea of "creativity is technology have been widely discussed dur- in the eye of the beholder" was also reining the past decade. Usually the chief forced by this exercise. The writer and criticisms against such a move have re- the reader put meaning in aphorisms &M- volved around cost and teacher resistance other creative products. (Trow, 1977). Yet these concerns have mostly been associated with computer- IMPLICATIONS assisted instruction using a large time- The use of these microcomputer activi- sharing system. Recently, there has been ties added greatly to the overall achieve- a shift from CAI programs aimed at indi- ment of the course goals.One reason for vidualized teaching of a single subject this success was the close fit between the matter toward computer-managed instruction course topic and the capabilities of the in which the system directs the entire in- computer. Indeed, creativity with its structional process (Splittgerber, 1979). "battle against fixed attitudes" stresses Even with this move toward CM/, there a concern with new images and experiences. still remains an implementation problem Torrance (1979), a major figure in crea- into the public schools. Economic, tech- tivity research, has even recommended a nological, and instructional design con- three-stage instructional model to en- siderations have kept the potential bene- hance incubation and creative thinking.. fits from being realized. Activities using a microcomputer can be But a revolution is in the making: implemented at each stage in his model. With the recent rise in accessibility of Stage 1 is aimed at heightening anticipa- microcomputers, the cost argument has been tion; it tries to motivate learners to greatly deflated. With the groWth in relate each learning task to meaningful microcomputer design and capabilities, experiences. The course syllabus and many of the technological concerns about summer story certainly attracted the stu- access, memory storage, and program lan- dents' attention and stimulated their guages have also been met.With the grow- curiosity and imagination. Other compu- ing commitment to instructional' design ter activities such as role playing games relevant to CAI and CMI, better software Computers in Humanistic Studies 41

is becoming available. Nevertheless, it REFERENCES teacher resistance is not overcome, all these changes will probably be for nought. Eisele, J.E. "Classroom Use,of Microcom- The experiences detailed in this paper puters."Educational Technology. reveal one possible strategy to introduce October 1979, P. 13-15. the applicability and benefits of microcomputers.By incorporating course con- Raudsepp, E. Creative Growth Games.New tent with microcomputer capabilities, York: Jove Publications, 1977. students can gradually come to appreciate the vast educational potential in this Splittgerber, F.L. "Computer-based In- new technology. With such a focus on struction: A Revolution in the computer-enriched instruction, the teach- Making?"Educational Technology. er is less likely to fear displacement January 1979, p. 20-26. from the teaching/learning process. The emphases of CEI experiences are on stimu- Torrance, E.P. "An Instructional Model lating class discussions and group inter- for Enhancing Incubation." Journal actions, not taking over the teacher's of Creative Behavior. 15199-117--- role. As Eisele (1979) has pointed out, P. 23-35. the possible uses of microcomputers in the classroom are many: as tools for Treffinger, D.J. and Huber, J.R. "Design- creative problem solving, games and simu- ing Instruction in Creative Problem lations, drill and practice; testing and Solving." Journal of Creative Behav- test construction.By taking advantage ior. 1975, 9, p. 260-266. of these uses to enrich classroom activities, today*s educator can be accountable Trow, W.C. "Educational Technology and and productive without feeling deperson- the Computer."Educational Technology. alized or replaceable. December 1977, p. 18-21. And isn*t that a reasonable way to win the battle against restrictive fixed attitudes?

SUMMARY AND CONCLUSIONS Since many students and teachers are hesitant to use the computer because they are unfamiliar with what it can -- and cannot do, it is imperative that ways be developed to introduce gradually a more realistic appraisal of the computer's capabilities. If these benefits are provided through computer-enriched activities which do not reduce the main role of the teacher, more acceptance of such activities should be forthcoming. With the advent of microcomputer systems almost any small college can afford the computing power that was available only to large institutions just a decade ago. An example of such an application in a _OPllege course on creativity was provided to illustrate the feasibility of such a CEI approach.

52 42 NECC 1980

GIVING ADVICE WITH Acateina

James W. Gets= Department of Philosophy University of Notre Dane Notte Dame, Indiana 46556 (219) 283-6471

Ovet the last thtee years, my colleague Paul heaven, or required same form of superhuman intel- MOWN* and I have been at wotk on an interactive ligence for its discovery. Very little attention computet ptogtam called EMIL which we use to help is payed in science education to allowing students teach our courses in formal logic.Since June of to appreciate the thinking processes which go into 19/9, we have been devoting our efforts (thanks to the analysis and mention of scientific problems funding from The National Science Foundition) to in a real setting. There is a tendency to obscure implementing a computerised "copilot" for EMIL the very human process of fumbling around, of try- which the students can call on to obtain advice on ing out strategies, of assessing failures, and of how to tackle problems that they find too diffi- cresting better lines of attack, which are all cult to handle on their own. patt of the scientists' daily life. A course in The methods we are using to provide out stu- logic gives students the opportunity to refine dents with advice are easy to Implement and quite their skills et problem solving in an environment effective. They represent an approach to comput - where the difficulty of the problems they conftont *tired education that doesn't fit neatly into the can easily be adjusted to their growing abilities. standard categories (i.e. recotd keeping, multi- In the standard sort of course, students' abili- ple choice CAI, testing, games, simulations, ties at finding proofs vary widely, so that those etc.), but one which has the potential for wide- who do not have an initial knack are severely or. spread application. The goal of thin kind of wilted. Even when strategies for proof - finding tutoring ptogran is to prompt the student to e re carefully discussed in class, some students in- develop and use creative strategies solve problems variably complain that they can't do a new problem which do not necessarily have any one on their own in spite of "understanding" the lec- correc answer.The advice-giving program is tures. Pert of the problem for these students is Socratic; it asks the students leading questions that they cannot convert a verbal explanation of concerning the problem before them, questions techniques. into s flexible tool for dealing with a which help them analyze and resolve their diffi- new situation. culties. With a bit of tutoring, most students with thee. A typical course in formal logic requires difficulties improve rapidly.If students think students to find formal proofs. Students are out loud while attempting a proof, a gentle nudge presented with a sot of formulas of logic and lure and there often leads to success.If they asked to apply certain logical rules to them to don't understand the rules, or simply haven't both- derive some othce formula.The derivation which ered to learn them, guiding then through a proof or they are asked to construct consists of a ss two tends to straighten things out fairly quickly queues of formulas each of which follows from and to improve their confidence and activation. previous members of the sequence by one of the Just as in teaching most skills, the effective rules, and which ends with the formula.they are strategies involve letting the student perform the to prove. This kind of learning is important task under guidance; lecturing on the proper pro- not only in that it provides the foundation for cedure and telling students to go home and do like» the mastery of the basic concepts of. logic, but wise is relatively ineffective. also because it gives the students the opportun- Of course there are good teems why tutoring is ity to learn some of the practical problems and not widely used in an introductory coots* in logic. strategies involved in creative thinking, parti- These classes are usually quite large, so tutoring cularly in creative thinking characteristic of simply-takes too much of the tes6er's UM*. Not mathematics and the sciences. only that, grading exercises in proof-finding is Giving students practice in the creative so- tedious, so teachers tend to give students rale.. lution of formal problems particularly import- tively few exercises that require then to create a ant in science education.Too often scientific proof. Sven if students can learn on their own, knowledge is presented as if it is descended from they simply don't get enough practice to develop

53 Computers in Humanistic Studies 43

any skill, unless they catch on right off the bat. of logic.Third, the program lets students tam Very often, the teacher relies on exercises that lines st the bottom of the proof Which they hope require a single answer, such as those that ask to derive later so that they can work the proof the student to fill in the justifications for the backwards if they like. The reason we allow, and lines of a proof that is already completed.This in fact encourage working backwards, is that ef- process does familiarise students with the rules, fective proof-finding strategies require an ana- but it gives then no practice in the art of find- lysis not only of the formulas already derived, ing a proof. but of the formula to be proven as yell. Often the Computers make it posigible to simulate the tu- proof - finding problem can be considerably simpli- toring situation. Students can enter their proofs fied by using the goal formula as a guide-post for at the terminal, and the computer can be program- determining what the steps previous to it oust med to see whether each line of the student's so- look like in the completed proof. Our program lution follows from previous lines and to describe allows students to record the results of using the difficulty if anything goes wrong. When the such strategies right it the terminal, instead of student gets lost the computer can make sugges- submitting a polished product to the computer. The tions about how to proceed. fourth advantage of our program is the main topic This use of computers in education is particu- of this paper. Since September of 1979 EMIL has larly interesting because it departs radically been giving students good advice about how to from the multiple- choice foneshich has become al- solve problems that they ere unable to do on their most paradigmatic of computerised teaching materi- own.In this way it is providing a good portion als. A proof-checking program doss not requite of whet can be offered by s human logic tutor. the student to cowl tv ady preselected answer, but There are three main strategies for designing to find a solution by any of a potentially infin- computer program that can offer advice on proof ite somber of lines of attack. In a sense, the finding.The first is simply to store a completed program does not demand an answer, but *imply pro- version of the proof that the student is working vides an ongoing check of the student's progress on and to store a list of comments that ere in- in achieving a result. It does not require a set tended to help a student who asks for aid on com- response so such as provides tool that studenti puting a particular line.If the calmest the stn can use in their own way to develops skill. dent :eceives proves isahelpful, the student can Compared to multiple-choice programs, proof- oak to see the next line of the stored proof, or checking progtams make heavy demands on the com- indeed any number of lines, up to and including puter, the teacher, and the student. The coeput- the entire proof. - -- or must interpret the student's Name at the This hint strategy requires that a completed terminal, determine whether they ere correct, and proof must be stored in the computer, along with respond in an intelligent way. The student and appropriate comments, for every problem students teacher must familiarise themselves with the pro- will work on. It also presupposes that there is cedures for operating the computer erosion, and likely to be only one reasonable sequence of steps suet put up with the inconveniences caused by hav- that leads to the conclusion.If studentsp- ing to use a computer which is generally overbur- proach a problem in an unusual way, there may not dened already, sad which occasionally malfunctions. be enough similarity between their proof and the They gust also put up with the inevitable mistakes stored proof for the computer to be of any help. a programmer sakes in designing the logic teaching finally, it presupposes a top -to- bottom pattern of system. And yet, if we are ever to develop com- proof construction. But frequently the very next puter teaching syetems that provide students with Jeeps in a proof will not reveal a strategy that twig' for learning, rather than merely with ongo- Leeds to success; suck strategies must rather be ing multiple choice exeminatians, we must over- explained with reference to what happens much come these difficulties.Working out effective later in the proof. This sort of hint routine strategies for proof - checking programs can pave does not help students appreciate the global str- thew for developing less authoritarian styles ategies which require knowledge not just of whets of cceputerised education in other areas. the proof has been, but also of *ere it Lavine, Our logic tutoring program, called EMIL, has and these are generally the most useful strategies. several advantages over other programs of the Mother technique is to write a program that eau* kind. First, there are a large number of allows the computer to generate a solution to the logic textbooks, each with its own version of the student's problem and to recognise standard *Stil- rule of logic. EMIL is the only program that ettoes during the course of that solution. This lets teachers supply the program with the set of strategy eliminates the need for storing a proof, rules to be used with their textbook, instead of with commentary, for each problem to be attempted, forcing them to use the book which goes with a since the computer generates its own solutions. set of rules written into the program. Second, But this strategy runs the risk of generating E MIL is extremely gentle with the student's input, strange proofs, which students are unlikely to to- and generally repairs typing mistakes rather than capitulate. Also, each formulation of the rules complaining about them. This le important because of logic will require its own custom-tailored pro- our students ere, for the most part, unfamiliar gram for generating proofs. Furthermore, the pro- both with the typewriter keyboard and the notation gram to generate comments must be very carefully

5.4 44 NECC 1960

written to avoid misleading advice. Worst of all, (See Figure I.)Each record contains the text of this strategy still does not help students to see a question followed by a lint of accepted answers, global strategies; like the stored-proof strategy, each followed by a number which indicates which this strategy uses a top-to-bottoet approach to record to Sump to in case the student responds with proof construction and confines itself to giving that answer. The last item in each record begins advice about the very nett line of the proof. begins with a "*" (which indicates that there.are Another difficulty with both of these approach- no more accepted answers) and contains text which es to the design for an advice-giver is that the is printed in case the student does not respond programs doss not attempt to construct advice on with one of the accepted answers. Most of the the basis of whatezer progress the student may questions we ask are answered with "yes" or "no", have already made on the ;nobles. The failure to but we found the use of other sorts of answers build advice *round the students partial successes more conVe2:ent for certain questions. The text tends to discourage invention of novel, yet pron. of the advice to he given is simply stored in the loins, partial solutions, to devalue the students' question file foil. ad by "*."This "*" indicates own creat...te abilities, and to lower their self- to the program that this "question" has no accept- confidence. It dampens the students' engagement ed answers, and ro the program should stop the ad- In the problem-solving process while reinforcing vice giving process after printing it. them for stereotyped solutions. We have built a number of improvements into The third approach to the design for an advice- this simple program. First, the sequence of the giver, the one we have adopted, overcomes these questions should vary depending on how such the problems by paying ..are attention to the techni- student has learned and how difficult the problem ques actually used by human logic tutors. One of is. Our first solution to this problem has been the main things a human tutor should do is prov- to assign each problem that the student is us. It ide students with effective problem solving tools on a level number and to use this number to roav for analyzing the situation they are in and for the question-asking program to separate quests... breaking the problem into simpler subproblems to trees for each level we have defined which the same tools can be applied all over The second enhancement is motivated by the fact again. An effective tutor does not give the that we want to mention items in our questions, Os.-r solution, or evenrpiedet-of-ft;bhi-tiotead-pro- that change during the execution of the program, vides an apprenticeship in 'AO art of asking the for example, the last line number the student has relevant questions, the r avers of which will finished In the proof, or the name of the rule lead the student to sae ow the problem can be that he intends to work with.Obviously the text broken down into moreenageable parts. Questions of the questions in the file cannot mention spe- like "Can you apply "-Us rule to lines you have cific line numbers or rule names. Our solution already derived?" lard "What rule could be used to is to introduce variables, or fillins, to our derive s formul.. 4.4 this shape?" when presented question text, in a coherent sequence are very effective in help- FIGURE I: SAMPLE RECORDS FROM A QUESTION FILE ing students .sevelop strategies which they can 1. 'CAN YOU APPLY MP TO ANY PROVEN LINES' '7' 2 learn rt. vs. effectively in a wide variety of 'N' 3' *ANSWER YES OR proof - fading problems. 2. 'APPLY MP TO THESE LINEC"*. "e actual program that we use to give advice 3. 'WHAT IS THE NAM CONNECTIVE or YOUR GOAL .us writs... by me in about four hours. The lm- FORMULA?' '4' 4 'V' S t->1 6 **PLEASE MOWER placentation was ao easy because the advice pro- 4, V OR 41 gram does little more than ask students a lead- that are then replaced with the cottesponding ing question and thee branch to s new question specific information Suet before the question is on the basis of their answer. Eventually, the printed at the terminal. We have adopted a con- program tuns out of questions to ask, and sped. vention that words beginning with "4" are vari- fie advice is given the basis of the informa- ables, and so, for example, a line of advice on tion prov44 ,- in the st4ant's answers to the our question file might read like this: 'OU questions. se questicao can be thought of as SHOULD APPLY 4kULE TO LINE 6GNUM'.This directs being structured in a Use, with the path along the program to fill iz the specific information the branches being determined by the students' about the rule name and line number so that the armors, and the advice for each situation be- student sees, for example, YOU SHOULD APPLY GOUT lug located st the tip of each branch. TO LINE S. Since programming our advice-giver was so It may surprise you to learn that although our simple, the main focus of our attention has advice-giving program was running with these two been on the creation of a file of questions enhancements in September 1979, we were which have real pedagogical merit.Since the working on a central portion of the advt. ing questions axe not written into the structure of program in January 1980. That was because ea still our program, modifying the question tree in re- had to program the most important improvements the sponse to4atwe learn about effective advice development of subroutines which can answer all has been a painless process which dose not re- the questions which are posed to students by the quire any programming expertise. advice-giver, and comment on any errors in the Our question file bee a very simple format. students' responses. Though students are gamer- Computers in Humanistic Studies 45

ally quite accurate in their responses to ques- wrong with their car, or whether to itemize their 'tions posed by our advice-giver, they occasionally deductions.All it takes is a simple program to make eine:Nuts that can result in their receiving run the questions and a question file that is bad advice. But informing students of their er- carefully constructed to reflect the best strate- rors is not the only reason for giving the comput- gies that people actually use to solve the kinds er the ability to monitor the correctness of the of problems which are at issue.Depending on the students' responses. Once students run the advice- context of its use, sone or all of the enhance- giver a number of times, they become bored at hav- ments to the basic program that we have developed ing to answer a number of pointless questions. The could be used. questions become pointless not because they aren't It is worth pointing out exactly how our ad- needed in analysing proof - finding priblems in gen- vice-giving progra differs from the standard eral, but because the Student is r- l aware that approach to CAI using the multiple-choice format. a particular portion of the me' is not needed The differences are not particularly striking from for the problem being dealt with aen the com- the programmer's point of view. In both cases, puter is capable of answering qua time itself, we the program is designed to ask questions and to can decide which questions, at mi. :h levels of select new questions on the basis of the students' difficulty, should be printed at the terminal, and mowers. The advice-giving programs generally re- which we should let the computer answer for itself quire a more elaborate branching structure, and by examining the proof the student is working on. they may differ in being unable to evaluate the Experienced students may resent being asked any students' response.But the important differences questions at all and nay prefer an advice-giver are the ones that are obvious to the educator, for that merely prints a specific piece of advice. they have to do with the educational purposes of However, we believe that for met students who the program. need the advice-giver in the first place, posing Multiple choice CAI attempts to get the student the relevant sequence of questions is such more to memorise the correct answers to a certain kind valuable to their learning problem-solving skills of question. The stress is almost entirely on en- than is their obtaining advice. suring that the student lutes certain facts. In At this writing we have a version of EMIL that the case of advice-giving programs, the answers answers for itself all the'questioes we pose save are not pert of whet is being taught. If anyone, and we have a method of indicating in our thing, it is the questions we would like the stu- question file which questions are to be asked dent to master. By exposing students, over and under which circumstances. He *till need to do a over again, to a sequence of questions that have lot more research on how obtrusive the advice- been proven effective in problem analysis, the giver ought to be as a function of the students' student learns to develop efficient strategies progress and cognitive style.However, the main that can be used over a wide range of problems of advantage of our program is that we have complete a similar kind. Furthermore, the whole process flexibility over the circumstances under which the of learning to adopt principles of problem analy- program types out the questions. sis and decomposition is a valuable exercise of There is a final reason for programming the problem solving skills that can be applied to computer so that it can answer all the questions. virtually any domain where creative thinking is When this is done the program can traverse the required. question tree on its own and come up with the Although advice-giving programs may not look relevant advice. Once advice is available, the very different from multiple choice courseware to programs can follow it to construct proofs on its tt% programmer, they have radically different edu- own. Judging from extensive paper and pencil cational goals, the most important of which is the tests, our advice tree turns out to be highly development of problem solving ability. Given the effective (though not totally effective) in solv- simplicity of the programming effort as compared ing logic proof-finding problems. As a result, to games and simulations, advice-giving programs it is capable of solving for itself the vast are particularly attractive for any educator in- majority of problems we give our studs s. This terested in developing the student's creativity. provides us with an 'sporran tool for .mproving our program. By running a large number of problems through our advice-giver, we can determine the circumstanced under which it is unable to do a proof, and than use that information to crests e more sophist -.1ted version of our question data files. Our approach to giving computerised advice has a wide range of applications. It can be used, for example, to help college students with their physics homework, or with tracking down the identity of unknowns in qualitative chemistry, to help medical students learn diagnosis, or even to help people to determine what is Ir

46 NECC 1980

FROM A THEORY OF READING TO PRACTICE VIA THE COMPUTER

Dale M. Johnson* Center for Educational Research & Evaluation The University of Tulsa Tulsa, Oklahoma 74104

R. Scott Baldwin University of Miami Coral Gables, Florida

INTRODUCTION Barry & Stevenson, 1975: Fang, 1968; Mar- The purpose of the present paper is to ten, 1978, and Walker & Boillot, 1979). describe a project which uses the unique Recently, reading specialists have and high-speed capabilities of the com- posited that fluent reading develops as an puter to match adolescents with books integrated process rather than as a loose which the students are capable of read- collection of reading subskills.Fluent ing and which they enjoy reading. The reading entails much more than the sum of ultimate mission of the project is to specific skills. Such skills provide a increase the amount students read, which necessary but insufficient foundation for in turn would enable them to become__ the-development of nature reading-(Gooda better readers. -Enhanced reading ability man, 1976: Smith, 1971). Profound in its and attitude would then motivate students implications, the basic idea reduces to a to read mores thus, the probability of a simple formula--young people need to prac- cycle of life-long reading habits for the tice reading in order to become good read- individual would be increased (Mathewson, ers (Allington, 1977: Daniels, 1971: Fader 1976). McNeil, 1968; and Squire, 1973). How- Traditionally, reading has been viewed ever, students who are in most need of as a hierarchical skills arrangement. reading practice tend to read the least. This conceptualization has been reflected The results are that the reading deficit in reading instructional programs which accrues with age, ultimately resulting in incorporate highly structured approaches secondary students who can neither read on to word recognition and comprehension a level congruent with their abilities nor skills. The underlying assumption has enjoy reading. been that students who are thoroughly grounded in word attack skills, phonics, RATIONALE sight vocabulary, word analysis, spelling, The basis for the project described in and other reading subskills would use this paper is that students do not read these skills in reading, thereby becoming partly because they fail to find reading able readers. materials which ar- comprehensible and Computer assistance in building read- interesting to them. However, a variety ing subskills is not unusual. Illustra- of books do exist that span the abilities tive projects in CMI include the Wiscon- and interests of virtually all students. sin System of Instructional Management Therefore, a sound argument can be ad- (WIS-SIM), the Instructional Management vanced that the problem is not one of System (IMS), the Interactive Training availability but one of accessibility. System (ITS), and the Stanford Project On the basis of current trends in read- (Splittgerber, 1979). Numerous CAI pro- ing, it could be reasonably hypothesized jects in reading have been implemented that if various student characteristics since the mid-1960s that capitalize on and preferences could be matched with both drill and practice and tutorial corresponaing book characteristics, the modes (Atkinson, 1968; Atkinson, Fletcher, probability of the student increasing his Chetin, & Stauffer, 1970: Atkinson & Paul- or her interaction with print would be son, 1970; Madachy & Miller, 1976: and enhanced. Further, the increased reading Morrison, 1979). Computers have also would result in a better likelihood of the been used to establish the "readability* student becoming an improved reader and levels of print material (Barry, 1979; developing more favorable reading habits 57 Computers in Humanistic Studies 47

and attitudes. Therefore, a major task linguistic readability, etc., will com- becomes one of maximizing the match be- bine in a predictable manner to determine tween student characteristics and perti- just how interesting a particular book nent print characteristics 3v. order to will be to a particular reader (Imam & select a book that the student can read Patyk, 1967; Smith & Johnson, 1972). and will enjoy. Gilliland (1972) pre- Thus, it becomes clear that the quality sents this task as follows: of the match between a reader and a book Readability is primarily concerned is a function of numerous personal traits with a basic problem familiar to and textual characteristics: however, the all people who choose books for literature does provide a comprehensive their own use, or who choose books list of the most important variables. for others to use. This is the problem of matching. On the one THE BOOKMATCH SYSTEM hand, there is a collection of In keeping with the goal of getting the individuals with given interests most suitable books to students, the most and reading skills. On the other pertinent variables were incorporated in- hand, there is a range of books to algorithms which were subsequently and other reading materials, coded for computer processing. First, differing widely in content, individual student preferences regarding style, and complexity.The ex- print characteristics are considered. tent to which the books can be Figure 1 shows the variables -on which read with profit will be deter- students can express their personal mined largely by the way in sentiments. which the two sides are matched... Three of the characteristics relate to (p. 12). the characters portrayed in the book, one The time-consuming task of matching relates to the setting or location of the students and books has typically relied story, and then 08_tvical interest areas on published bibliographies; -human recall, are considered. With respect to the "in- and trial and error. Such techniques are terest areas," the student simply obviously limited.Finding books which cated how well he or she likes each par- contain all or most of a desired set of ticular area from he list of 88 topics characteristics is a complicated task, in on a three choice response format coded fact, it is virtually impossible if the as follows: (Y) Yes, very much, (S) set of characteristics is large and the Sometimes. It's OK., (N) No, I don't. Of resources for finding them are limited the OS interest areas, 20 are human drama to human intervention techniques. Simpson themes (ecology, drugs, personal beliefs, and Soares (1965), Jongsma (1972), and loneliness, etc.). Stanek (1975) have found that librarians Corresponding to each connecting line and teachers, regardless of their in- show in Figure 1, an algorithm was devel- tentions, fail to accurately consider oped and coded into a FORTRAN subroutine the reading interests of young people for a main program. However, Figure 2 when purchasing or recommending books. displays additional algorithms for refin- Research over the past decade or so ing the match between students and books. provides clues as to what variables arc Figure 2 shows three student traits that important for matching considerations. represent somewhat different constructs Robinson and Weintraub (1973) and Squire from the student variables (preferences) (1973) have confirmed the importance of shown in Figure 1. reading interests of students when The student traits which are linked to selecting reading material. Further, the various book variables are reading abil- reader's age, grade level, socioeconomic ity, attitude toward reading, and grade background, sex, and cognitive reading level. For example, three algorithms use ability will tend to influence reading a measure of student reading ability to- preferences (Ashly, 1970; Carlsen, 1967: gether with: (1) linguistic difficulty Jungeblut & Coleman, 1965; Hansen, 1973; of the book,(2) when the story appeal Scharf, 1973: and Soares & Simpson, 1967). begins, and (3) the length of the book. Book characteristics which tend to be Conceptually, better readers can more influential include amounts of dialogue, readily comprehend more difficult text, concreteness of language, type of narra- they can comfortably tolerate more intro- tion, and the degree of action and con- ductory background prior to the beginning flict provided (Carlsen, 1967; Jungeblut of the story appeal, and they will tend to & Coleman, 1965; and Simpson & Soares, cope better with longer books than poorer 1965). The reader's age, grade, sex, readers. ethnic background, personal history, Students' attitudes toward reading are hobbies, and the book's length, theme, also used (Figure 2) to help discriminate

5V 48 NECC 1980

among books. The algorithms use a meas- information on books written for early ured attitude score based on the follow- adolescent readers (grades 5 - 9).New ing heuristics. Students having better volumes are periodically being added to attitudes toward reading, as opposed to the data base which presently contains those with poor attitudes, will be able about 5,000 volumes established over a to cope with longer books, will not re- three-year period. Initially, pertinent quire early story appeal, and will rely book variables were identified through a less on physical action as a prerequisite review of literature in the field of read- for enjoying a book. Finally, the grade ing. The original pool of variables was level, since it is associated with chron- verified by a committee of public school ological age and maturity, is also com- reading teachers and university reading pared to the linguistic difficulty (which specialists.Two of the original varia- is measured in grade equivalents) of the bles were subsequently omitted from the books and to the length of the books. list due to variations in printing styles Generally, students in lower grades will across separate editions of many of the not enjoy longer and/or acre difficult books. These two variables were "number books. of pictures" (or non-print figures) and A FORTRAN main program was written for "size of type." the Xerox Sigma 6 system which includes Second, a master list containing the subroutines incorporating the schemes titles of books to be included was com- shown in Figure 1 and Figure 2. The func- piled from published recommendations from tion of the program is to receive individ- widely recognized sources including the ual student variables as input and match American Library Association, Library of them with each of the approximately 5,000 Congress, School Library Journal, American books in the data base. Each variable of Guidance Services, International Reading each book is compared with the respective Association, and the National Council of student variables and differential Teachers of English. weights are assigned according to the de- Reviewers were tutined to evaluate gree of match and the relative importance characteristics of books and to prepare of the variable. The weights are con- annotations. These reviewers consisted of verted to points and summed, resulting in reading professionals witn masters degrees each book being scored for each individual and with either adolescent reading or student. The books with the highest num- library experience. The book evaluations ber of total points are listed as a per- were checked for reliability and were sonalized ,bibliography for each student. further validated by a reading specialist. The functional operation of the match- The annotations were refined by a profes- ing process (referred to as Bookmatch) is sional editor. Book data were then coded shown in Figure 3. As can be seen, the and stored for computer match of book first function is to extract a subset of characteristics with student traits and books which are accessible to the student preferences. from the book data base. In most cases, this subset is the school's library hold- SUMMARY ings which have been furnished by the Based.on the notion that practice in librarian. Second, each variable of each reading will be facilitated when readers book is compared with either the student's and appropriate literature are brought preferences and/or traits. The algorithms together, a project called Bookmatch is accumulate a point total (score) for each presently being implemented. In order to book. Finally, the books receiving the match reader characteristics an' pre- highest score (most accumulated points) ferences with textual characteristics of are printed. This output includes the books, a medium with capabilities for vast author(s), title, type of book (fiction, data storage, retrieval, and rapid arith- nonfiction, biography, etc.), and a brief metic comparison operations was required. annotation of the book. The computer was the logical choice to Each student receives a printed list handle the task. During the Bookmatch of his or aer own personalized bibliog- processing, books are scored according to raphy. The bibliography includes the top the degree to which they favorably compare 20 books and must contain some nonfiction to student preferences and characteristics and biographies along with fiction. On the basic of these scores, the computer Summary reports for teachers, librarians, provides a printout with a personalized and administrators are also provided. annotated bibliography containing the most comprehensible and interesting books that DATA BASE are accessible to the individual student. Of significance to the success of the system is the data base which contains 59 Computers in Humenistic Studies 49

BOOK VARIABLES STUDENT VARIABLES (PREFERENCES)

4SEX (SEX OP MAIN)1( PETERENC) CHARACTER(S)

ETHNIC ETHNICITY OP REFEREN IN CHARA

)(AGE AGE OF MAIN REPERENC) CHARACTERS) )1(

(SETTING4) SETTING OF PREFER= THE STORY

INTEREST AREAS OP 2VE BOOK

Figure 1. Schematic fpr matching algorithms of student preferences and book variables.

6Q 50 NECC 1000

SEQUENCE or (INDIRECT) STORY ----lit

GRADE)4 PHYSICAL LEVEL LENGTH OF ( BOOK

ATTITUDE AMOUNT of PHYSICAL READING ACTION

CHARACTER ISTICS

.( BOOK CHARACTERISTI S

Figure 2. Schematic network for matching algorithms of student traits (characteristics) and book variables.

61 Computers in Humanistic Studies 51

new Lunn sown=

SELY.CT BOOK LIBRARY DATA HOLDINGS BASH

INPUT STORE STUNS? LIBRARY'S MOLDINGS .--IVARIABLES

4 INPUT XaBOOR MATCH ARIFIRENOSS (sm.1)

CUMULUS BOOK TOTAL SCORE

SORT BOOKS BY POINTS (SCORES)

OUTPUT' INGS AND BEST BOOMS 4

(ANOTHERsTunetakftroPOR tisX'? i( STUDENT ./

Figure 3. Functional Schematic for BOOM= operation. 52 NECC 1980

REFERENCES Newark, Delaware: International Read- Allington, R.L. "If They Don't Read Much, ing Association, 1976. How They Ever Gonna Get Good?" Journal of Reading, 1977, 21, 57-61. Hansen, H.S. "The Home Literary Environ- ment--A Follow -up Report." Elementary Ashly, L.P. "Children's Reading Interests English, 1973, 50, 978. and Individualized Reading."Elementary English, 1970, 47, 1088-96. Jongsma, B.A. "The Difficulty of Child- ren's Books: Librarians Judgments Atkinson, R.C. "Computerized Instruction Versus Formula Estimates." Elementary and the Learning Process."American English, 1972, 49, 20-6. posys.%ologisf. 1968, 23, 225-357-- Jungeblut, A. & Coleman, J.H. "Reading Atkinson, R.C., Fletcher, J.D., Chetin, Content That Interests Seventh, Eighth, J.C. & Stauffer, C.M. Instruction in and Ninth Grade Students." The Journal Initial ReadinUnder Computer Control: of Educational Research, 196758, 392- tanfor Project. TechnicaTW47513Ft T51. 131.3117ard, Calif.: 'natl.. 'or Mathematical Sciences in the Sc !Carton, H.A. " G M Program Rates Level of Sciences, 1970. Text Difficulty." Computerworld, 1978, (May), 9. Atkinson, R.C. & Paulson, J.A. An A roach to the Psychology of Instruc- Madachy, J. & Miller D.J.The Use of CAI t on. TcahlEil Report 157.3TIEE7A, in the Lan ua e Pr ram at GaTigunt. Calif.: Institute for Mathematical Santa Bar ara: Assoc atran73Eai Studies in the Social Sciences, 1970. Development of Computer -Based Instruc- tional Systems, January, 1976. Barry; a:c.- "Computerized Readability Levels--Their Need and Use." Journal Mathewson, G.C. "The Function of Attitude of Educational Data Processing770757 in the Reading Process."In H. Singer and R.B. Ruddell (eds.), Theoretical Models and Processes of Reading (2nd Barry, J.C. & Stevenson, T. "Using a M=Newarrirei,Dwares International Computer to Calculate the Dale -Chall Reading Association, 1976, 665-76. Formula."Journal of Reading, 1975, 19, 218-22. Morrison, F. TICCIT. Dynamic Phoenix, 1976 (July), 45-6 Carlsen, G.R. Books and the Teen-a e Reader. New YENT-Iirper & Row, 967. Robinson, H.M. & Weintzaub, S. "Research 'Related to Children's Interests and to Daniels, S. How 2 Gerbils 20 Goldfish Developmental Values of Reading." 200 Games IINTIFhem Library Trends, 1973, 22, 81-108. HewE12-1.7a. .iihTIORATaTTheWar-- ail-stetPress, 1971. Scharf, A.G. "Who Likes What in High School."Journal of Reading, 1973, 16, Emans, R. & Patyk, G. "Who Do High School 604-7. Students Read?" Journal of Reading, 1967, 10, 300-4. Simpson, R.H. & Soares, A. "Best-and- Least-Liked Short Stories in Junior High Fader, D.N. & McNeil. Hooked on Books: School."English Journal, 1965, 54, Prram and Proof. Nis7-173Fks G.P. 108-11. Putnam 11-13nr-f568. Smith, F. Understanding Readin . New Pang, I.E. "By Computer:?leach's Read- York:Holt, Rinehart an W nston, Inc., ing Ease Score and a Syllable Counter." 1971. Behavioral Science, 1968, 13., 249-51. Smith, J.R. & Johnson F.D. "The Popular- Gilliland, J. Readability. London: ity of Children's Fiction as a Function University of London Press, 1972. of Reading Ease and Related Factors." The Journal of Educational Research, Goodman, X.S. "Reading: A Psycholin- 1727437-7977467-----7-- guistic Guessing Game." In Singer and R.B. Ruddell (eds.).Theoretical Soares, A.T. & Simpson, R.H. "Interest in Models and Processes of Reading. Recreational Reading in Junior High Computers in Humanistic Studies 53

School Students." Journal of Readig, 1967, 11, 14-21.

Splittgerber, P.L. *Computer-Based In- struction: A Resolution in the Making?* Educational Technology, 1979, 19, 20-5.

Squire, J.R. "What Does Research Rev;a1 About Attitudes Toward Reading?" In R.A. Meade and R.C. Smith (eds.), Literature for Adolescents: Selection an -thiC -Columbus, Ohio: Charles E. Brirra, 1973.

Stanek, L.W. *Real People, Real Books: About YA Readers."ToE of the News, 1975, 31, 417-27.

Walker, N. & Boillot, M. °A Computerized Reading Level Analysis." Educational Technology, 1979, 19, 47-9.

*Note Dale Johnson will present paper.

64 54 NEW 1960

NON - HARMONY: A VITAL ELEMENT OF EAR-TRAINING IN MUSIC CAI Join C. Groom- Thornton and Antoinette Tracy Corbet School of Music North Texas State University Denton, TX 76203 (817) 788-2791 ext. 269

INTRODUCTION MD BACKGROUND and context. Although each pattern is Computer-assisted instruction (CAI) rather specific, melodic and rhythmic in ear-training has become an important context can vary somewhat and harmonic assistant to classroom teaching in the context can vary gz:atly. For this first two years of college-level music reason, the necessity of drill and prac- theory. The student's aural skills (which tice of such variance presents a logis- are needed to deal with the three basic tics problem in the classroom but is elements of music -- harmony, melody, and greatly helped by random access and tran- rhythm) can be greatly reinforced with position available with CAI.__ -such individual instruction. Part of the success of this reinforcement is that METHOD while these elements are integral parts The ear-training CAI system at North of a complex multi-dimensional subject, Texas State University employs the Auto- they can be separated from each other to matic Music System (ANUS) designed by a great extent, so that classroom teaching Professor Dan W. Scott of the Computer (and CAI lessons) can concentrate on Sciences Department. The ANUS system virtually one element at a time. In consists of a Motorola 6800 microprocessor particular, this system allows the less with a CRT terminal and specialized music advanced student to concentrate his hardware (MUSOR). The microprocessor efforts upon developing his understanding translates the musical score and accesses of one broad subject at a time. As his the MUSOR, which is the actual sound- skills and confidence grow, he is better producing apparatus engineered by capable of identifying these elements in Professor Scott. The ANUS system is various combinations, and eventually of connected to an BP 2000-F Timeshared BASIC understanding them in their musical computer, which provides additional file context. storage capabilities. This system uses a As soon as the student has developed flexible score language for notation input some skill in each of these three areas and offers a variety of timbre and articu- and has begun to understand the integrated latzon possibilities. Of primary impor- whole, it is necessary to introduce the tance is the facility of rapid playback concept of "non-harmony."As the name which is a substantial advantage over implies, this element lies outside the other computer-based music systems in its realm of the established harmonic strucT cost range. tare but relies on melodic and rhythmic The non-harmonic tone lessons were context to differentiate among the various constructed using a simple melodic and forms that it takes. Despite its rather rhythmic four-voice chorale style. This negative title, non-harmony is a very simplicity was not due to any system positive part of music an aspect of limitations, as the system is capable of the concept of variety which helps to rapid and complex articulations of rhythm, delineate style. In practice, non- a wide range in melody, and up to seven harmony takes the form of short melodic simultaneous voices for harmony. Rather, and rhythmic patterns which augment the the simplicity,was'maintained to limit harmonic context. These patterns fall the number of variables of elements other into approximately ten categories for than those being tested and to give a which terminology is fairly standard. fairly normal musical context. For each Each term refers to a type of non-harmonic example heard, the basic context was a tone which defines its own special pattern

65 Computers in Humanistic Studies 55

five - to seven-chord progression of four by the examples (which were carefully voices in quarter-notes. To this strictly written to provide maximum exposure to harmonic setting, one or more non harmonic the variations encountered in music), the tones were added in eighth-note motion. summary lessons gradually grew in number Since there were ten categories to be of examples since they demonstrate a covered, the following factors determined variety of combinations of these elements. the order in which the terms were pre- Of course, this variety is further sented. Group I consisted of the five expanded by random selection and trans- non-harmonics that normally occur off the position. beat or in a rhythmically weak position A standard format for the student's as compared to the note of resolution answer was kept for all of the lessons. which comes directly after. Group II Because of the complex situation presented consisted of the four non-harmonics that by non-harmony, the student's response occur on the beat or in a rhythmically could not be reduced below three elements: strong position, and usually displace or the rhythmic, melodic, and non-harmonic delay a note of the harmony. The last contexts. (This sequence best repre- category normally has no such rhythmic sented the student's understanding of the definition. Due to the rhythmic contexts, factors involved, as determined by the Group I is considered easier to hear than author's experience in teaching.) Group II. Also, in actual practice in For the first element, the student music literature, the patterns in Group I types a number from one to six, locating generally occur more often than those in the beat with which the non-harmonic is Group II. associated. (Although many examples are Within Group I, the five unaccented seven chords long, a basic precept of non- non-harmonics (with their standard abbre- harmony is that it must finally resolve viations) were presented in this order: into harmony. Therefore, non-harmony 1. _unaccented passing tone (OPT) could not,be.associated with the seventh 2. neighboring tone, upper and lower beat.) The second element of the stu- LN) also sometimes called dent's answer is a capital "S" or "8* "auxiliaries" which indicates that the non-harmonic 3. anticipation (ANT) occurred in the soprano or bass melodic 4. escape tone or echapee (ET) voice. (The choice was limited to the S. changing tones (CT) upper and lower voices to afford a measure Within Group I/ the four accented non- of variety while avoiding the unnecessary harmonics were given this order: complication of determining locations in 6. accented passing tone (APT) inside voices when these voices can be 7. suspension (SUS) extremely close.) The third and last 8. retardation (BET) element of the student's answer is the 9. appoggiatura (APP) abbreviation that identifies the non- The final example was: harmonic itself from among the ten pos- 10. pedal tone (PBD) sible labels.Even though this part of The ir'arnal order of the two groups was the answer would be obvious in any of the chosen to present the most-often heard single-element lessons, the act of typing categories first, in ordtr tarstmrtthe the abbreviation reinforces the student's student with the most familiar patterns.The aural impression and, establishes a pattern only exception to this practice occurred of response which does not need to be with categories 8 and 9. The less-often altered for the summary lessons. All heard retardation (18) was placed before elements of the answer were separated by the appoggiatura (19) in order to present commas, even when the example involved it directly after the suspension, on which more than one non - harmonic response. it is patterned. Alsothis order was One other factor, a chord tone, was selected because Jummary lessons are used added to a few random examples within in various places between individual pre- each lesson. A chord tone may resemble sentations, and a t meaty to show simi- a non-harmonic tone in rhythmic and melo- laritieu and differences between the sus- dic placement, but it is part of the har- pension and retardation would logically mony (as opposed to non-harmony), and its come before proceeding to the next (and pattern of usage does not necessarily less - related) non-harmonic. parallel that of a non - harmonic tone. The general need for both immediate This factor was added to insure that the and cumulative summaries of such a large student would truly listen for non-harmony number of elements resulted in seven such and not just attempt to label anything lessons being inserted into the sequence that moved! These chord tones do not already discussed.While each of the ten require any response from the student single-element lessons seemed well served other than to understand that they were

CC 56 NECC 1980

notes which were not part of the non- C: CHORD PROGRESSIONS WITH UNACCENTED harmonic answer. C: PASSINGATONES H: FIRST ANSWER IS THE NUMBER OF THE RESULTS HI CHORD AFTER WHICH THE NON-HARMONIC In final sequence, there are seven- H: TONE:SOUNDS. teen lessons, delineated as follows: H: SECOND ANSWER IS "S" OR "B" FOR Nl. UPT - Unaccented passing tones H: SOPRANO OR BASS VOICE WHERE THE NON - occur most often of any of the non- H: HARMONIC TONE OCCURS. harmonics and are located equally H s THIRD ANSWER IS THE ABBREVIATION (UPT) well in soprano or bass. H: FOR THE NON-HARMONIC. N2. UN, LN - Upper and lower neigh- Hs TYPE LETTERS IN CAPITALS (WITH SHIFT bors occur mainly in soprano, as H s KEY). reflected Immthe examples. H: (SOME MOVING EIGHTH-NOTES MAY BE CHORD N3. ANT - Anticipations occur almost Hs TONES.) exclusively in the soprano and The *H" statements are "hints" which the usually very close to the end of student may branch back to at any time to the progression. obtain reminders about valid answers and N4. Immediate summary of Lessons 1, format. 2, and 3 (the most-often heard T: $N, HERE'S A LESSON TO HELP YOU LEARN unaccented non-harmonics). T: NON-HARMONIC TONES. N5. ET - Escape tones occur mainly Ts THESE WILL BE UNACCENTED PASSING-TONES. in the soprano voice. P: 2 N6. CT - Changing tones involve a The "T" statements are texts which the definitive four-note pattern and student sees, and the "$N" addresses the occur only occasionally in the bass. student by his own name. The HP: 2" is N7. Immediate summary of Lessons a two-second pause (with the text still on 5 and 6 (somewhat similar patterns). the screen) to allow the student to absorb N8. Cumulative summary of all unac- the information. cented non-harmonics (Lessons 1 CS: through 7). Ss 2 N9. APT - Accented passing tones, Ts I'LL PLAY 5 TO 7 CHORDS. like unaccented ones, occur in both T: YOU TYPE: WHICH CHORD (1,2,3,4,5 OR 0 soprano and bass. T: 6) THE NON-HARMONIC OCCURS AFTER, N10. Summary of APT and Urf (Lessons Ts WHETHER THE NON-HARMONIC IS IN SOPRANO 1 and 9). These two patterns occur OR BASS (S OR B) IN CAPITALS. quite often and are alike except T: THE ABBREVIATION FOR THE NON-HARMONIC for rhythmic placement. Ts (UPT) IN CAPITALS. S - Suspensions occur in both Ts TYPE A COMMA (,) AFTER EACH PART OF voices in many cases. Ts YOUR ANSWER. N12. RET - Retardations parallel Ts FOR INSTANCE:3,B,UPT suspensions in every way except Ts MEANS THE NON-HARMONIC OCCURRED AFTER that they resolve up instead of . Ts THE THIRD CHORD, IN THE BASS VOICE, down. T: AND IT WAS AN UNACCENTED PASSING- N13. Immediate summary of the ac- Ts TONE. cented non-harmonics covered thus Ps 10 far (Lessons 9, 11, and 12). The"CS" then clears the screen so that the N14. APP Appoggiaturas may hays following text can appear as a stationary slightly varied melodic patterns group without rotating off the screen. and most often occur in the soprano. "S: 2" double-spaces the text of added M15. Cumulative summary of all ac- visual impact. The text contains specific cented non-harmonics (Lessons 9, instruction on the anwer format and gives 11, 12, 13, and 14). a hypothetical student response with N16. Tap - Pedal tones are the least explanation. Again, there is a pause of often used, but are the most obvious ten seconds for comprehension. non-harmonics to identify, occurring Ts HERE'S AN EXAMPLE (WITH ANSWER), $N -- mainly in the bass. T: $R 4CD 3C 2E GE N17. Final cumulative summary of all Ts 4EE of the non-harmonics (This lesson T: 4G 3BD 2D GQ contains the most examples). T: 4E 3CU 2G CUQ Ts 4F 3C 2F ADO SAMPLE T: 4G 3GD 20 BE The following example shows Lesson Nl Ts CUE (UP Lth comments on specific features. Ts 4GD 38 2F DO Cs L, 1/13/80, JCGT & ATC Ts 4CU 3G 2E CH 67 Computers in Humanistic Studies 57

Ti ? responses are kept and made available to P: 20 the classroom teacher for evaluation and Ti eS 813 advising. T: YOUR ANSWER: 5,S,UPT -- RIGHT, $N ? pi 10 SUMMARY AND CONCLUSIONS The next section plays a sample example Music poses interesting instructional for the student (lines 145 - 205 process problems because it shares with other the sound) and then gives the correct fields (for example, chemistry and math) answer with a pause for comprehension. a well-defined vocabulary for musical Ti REMEMBER: elements. However, musical teaching must Ti TYPE A "," BETWEEN THE PARTS OF YOUR include showing the relationships among Ti ANSWER. these elements. Non-harmony is an essen- Ti TYPE A "7" TO RE-HEAR THE MUSIC. tial component of these relationships, Ts TYPE " / /HINT" FOR A HINT ABOUT and thus forms a model for the solution Ti ANSWERS. to the problem of teaching.relationships Ti (YOU MAY WANTTO USE PAPER AND PENCIL in many fields. Ts TO JOT DOWN ANSWERS BEFORE YOU TYPE T: THEM IN.) SELECTED REFERENCES P: 10 Apel, Willi. The Harvard Dictionary of The student then sees the final reminders Music. HarvainfaVerexty Press, on basic procedures for dealing with the Ca ridge, 1969. musical examples. Bales, W. Kenton and Joan C. Groom. Ti LET'S GET STARTED, SR, "AMUSs A Score Language for Comput- Vi 123456SBUPT, er-Assisted Applications in Music." 0: 10CRNIAP4,453 ). Presented at the 1979 Association The actual examples are then announced for Educational Data Systems Conven- to the student."V"' gives-the valid tion3*Detroit, Michigii. characters for the student's answer. For Hamilton, Richard L., and Dan W. Scott. every lesson the first eight characters "A New Approach to Computer-Assisted are the same; the numbers 1 through 6 Instruction in Music Theory." are the possible answers to the first Presented at the 1977 Conference on question of beat (rhythmic placement); Computers in the Undergraduate Cur- and "s" and "B" are the possible answers riculum, Denver, Colorado. for voice (melodic) placement. The last Hofstetter, Fred T. "Computer-Based characters label the non-harmonic and Recognition of Perceptual Patterns range from a single one- to three- letter in Harmonic Dictation Exercises." grouping (for the single-element lessons) Presented at the 1978 Association to the complete list of all non-harmonic for the Development of Computer possibilities for N17. In this sample Instruction Systems Conference, lesson of Ni, only the single abbrevia- Dallas, Texas. tion of VET is valid. Finally, "Q" Killers, Rosemary N., W. Kenton Bales, formats the questions. In this case Richard L. Hamilton and Dan W. Scott. there are 10 possible examples from "ANUS:The Computer in Music which the program makes a random selec- Instruction."Presented at the 1979 tion and a direct comparison. The stu- Texas Music Educator's Association dent completes the lesson when he Conference, Fortlorth, Texas. answers four out of five times correctly, Ray, Douglas, and Rosemary N. Killam. And he is allowed three tries,befobe the "Melodic Perception Development and correct answer is given by the computer. Measurement Through CAI."Published Ten examples then follow which are in the proceedings of the 1979 similar to the sample example of lines National Educational Computing 145 through 190. Conference, Iowa City, Iowa.

DISCUSSION The seventeen lessons of this non- harmony project are all presented in parallel format to establish a consistent pattern of thought and approach and to minimize confusion as the lessons gradually become more complex. Each element is introduced in a lesson devoted solely to its problems and contexts, and summary lessons gradually combine the elements in a logical order. Data on students'

68 Computer Literacy

A CASE FOR INFORMATION LITERACY Dr. S. D. Schiaming ISM Corporation 10401 Pernwood Road Satbeeda, MD 20034 301/897-2090

INTRODUCTION a statistical point of view, they will Although digital computers have represent s small minority of the popu- been present in educational institu- lation interacting with future informa- tions for over 25 years, only a small tion systems. it is the end user major- percentage of students have been ex- ity for which computer literacy is posed to them. As the cost of com- probed in this paper. The result of puting continues to decline, it is be- this exercise is a recommendation for coming financially feasible to intro- the creation of a course entitled "An duce the "majority of students to com- Introductory Course in information puting and data processing. Literacy" for the majority of under- Dr. Andrew R. Molnar (1) is an.ert1- graduate students who are pursuing a cultte spokesman for the expanded inte- major other than computing. gration of the computer luto American education. Citing our shift from an TECHNOLOGY DIRECTIONS industrial society to a knowledge -based The 1980s will conceivably see an society together with the international order of magnitude growth in computing challenge to our technological leader- power. Responsible for tnis growth ship, he makes a compelling case for will be the continued decline in cast computer literacy. Dr. 4elner is not of electronic components together with alone to his views. The te:: "computer the elastic demand for data processing literacy" is being increasingly used services. The concept of the computer and discussed at educational confer- rill rhease as processing capability is ences and in the educational litera- dist.- uted into many devices. Today's' ture. hobb t with a microprocessor at home The objective of Ott paper is to cepa... of communicating with a distant explore the future content of computer large computer is a beginning example literacy education.Bose' upon an of distributed processing. One of the assessment of current computer /data reasonsfor this trend is that the processing education, trends in tech- cost of computing has dropped below the nology, ma a projection of the future cost of anuaicstion. Thus it becomes inforastioe environment, a set of three desirabl ;o satisfy local data pro- future skill categories is defined. cessing kegnirements at the source Computer /Data Processing Pro-, while communicating only that data fessionals (2) - the computer required by other sites. Communication experts costs will also continue te, decline . Inforaatioa System Profession- as newer forms of broadhead communi- als (3,4) - the integrators of cation, such as satellite transmission, the computer and business pro- become commonplace. Again, as in ege CerfeS computer industry, change will he drama- The End Users - the information tic in the communications industry. system customer There can be little doubt that informa- The first two categories are, of tion technology growth will continue to course, key areas requiring educational he significant through the latter part planning and investment. However, from of this century.

NOTE: The views expressed in this paper are those of the author and do not necessarily represent the views of the international Snail:ass Machines Corporation. 69 58 Computer Literacy 59

The key implication with regard to tigation. Nevertheless, the nation has educational curricula is that not only a much higher demand for personnel, will computers change, but the informa- such as IS graduates, who have a combin- tion envlrossent in which they operate ation of technical and organizational will change; the way data elements are skills than for computer scienc t. grad- moved, stored, and used will be altered uate with 'solely' technical skills." as wall as the way they are processed. With the previous discussion as a An analogy to transportation aye- basis, certain conclusions can be made teas can be drawn *ere. Vehicles, road- regarding future educational require- ways, and parking facilities are compo- ments of the three categories of stu- nents of a transportation system just dents previously defined. as computers, communications, data, and applications are components of an infer- FUTURE EDUCATIONAL REQUIREMENTS nation system. Vehicles are not Vocational and Computer Science studied in isolation by urban planners, Professionals - The demand for and correspondingly computers should this type of professional to- not be studied in isolation from other gether with the curriculum components of the information system. needed to prepare him or her appears to be relatively well CURRENT COMPUTER/DATA PROCESSING understood and served by the EDUCATION educational community. For the purpose of this discussion, . Information System Profes- it is convenient to think of three sionals - As reflected by the clssses of students that correspond to Nunamaker study, the information the skill categories presented in the system concept has been slower introduction. to emerge. However, due to the . Computer Majors - tho t prepar- efforts of the ACM (3,4) and the ing for careers in da,a process- nation's business schools (42 of ing. the 52 satisfactory IS under- . Information Systems Majors - graduate programs were compo- those preparing for careers in nents of business or management managemenc with a major in infor- colleges) (6), the problem is mation systems. understood and undoubtedly steps . Other - those with career objec- will be taken to improve the tives other than computing or quality and quantity of informa- data processing . . . the future tion system graduates. end users. . All Others (the future end Dr. J. Baubles's most recent inven- users) - It is this group for tory of computing in American education which current computer literacy (5) provides insight into the relative appears to be inadequate. They distribution of these classes: in ex- will take their places in market- cess of 901 of the students fell into ing, production and administra- the "other" category while 301 of the tion with a cursory knowledge of institutions offered courses in the the computer but not the informa- first two cstegoriss. From these station system in which they func- tistics, it would sees the computer tion. A hard-earned lesson in major is receiving sore than his fair the computer indubtry is that to share of reseurca$. major cause of information sys- The status of information systems tems fatless has been that they (IS) education is addressed in a recent didn't do what the users ex- article (6) reporting preliminary 'find- pected The end users must rep- Legs of the ACM Curriculum Committee resent their needs as input to for Information Systems. Prof. Jay F. the information system design Nuns-alter, chairman of the committee, process. .id: "The IS field integrates eye- The following course description .efts analysts, statistics, management addresses the minimum elements of an science, accounting, economics, fin- information literacy offering. ance, marketing, production and computer and communication teahnolo- AN INTRODUCTORY COURSE IN INFORMATION gime . "The U.S. has nearly five LITERACY computer science departments for every There are four key components in IS department according to a recent this information literacy proposal: study not connected with the ACM Laves- computers, application programming,

7u 60 NECC 1980

data, and communications.The computer . Industry Structure portion could be satisfied with the - Common carrier services traditional "Introduction to Computing/ . Switched Data Processing" course currently avail- . Leased able at many institutions. There is no . Value added - packet need to elaborate here on the contents switched of this type of course. . Satellite . Other Application Programming. - Tariffs Coding-in a high-level language is . Information Networks not sufficient to understand the role - Combining count:Ideation° of application programming. In fact, and data processing for the generalist, it may be of second- - Network architectures ary importance to an awareness of the . Centralized application requirements/preliminary . Distributed design process. The following topics . Design Tradeoffs should be addressed: - Nigh volume - batch . Requirements Study Interactive - Documenting information flow - Public vs private in an organization It should be noted that each of . Prz:tminary Application System these four topics is a complex special- Dep4Ln ty in its own right. No pretense is - Transaction definitions made that skill in any of the topics - Screen formats would be developed through the course, . Generation of a Software but rather a conceptual knowledge of Specification the elements in an information system . Design of an Application would be gained. Also of equal impor- Software Module tance, the role of the end user in an - Flowcharts informtion system design would be estab- RIPO diagrams lished. . Coding of an Application A two - semester sequence would be 'Software Module desirable; however, a one - semester sur- - Coding vey course could also be of value. The - Testing business school, because of its prece- - Debugging dence in information systems curricula, would be a logical choice to ofer this Data introductory course. Much as the treasurer is respon- In conclusion, it is hoped that sible for the money resource in -an this presentation stimulates thinking organization, the business community in the educational community with has come to realize the value of their regard to itsroil in preparing atm data records andthe need to assign dents to coexist with information tech- management responsibility for their nology in the 21st century. safekeeping and use. An infornstic literacy course should include the NOTES following subjects. 1) Andrew R. Molnar, "The Next Great . The Concept of Data as a Re- Crisis in American Education: Com- source puter Literacy," ARDS Journal, Asso- . Data Characteristics cief'-.1 for EducationalMISys- - Physical relationships tems, Volume 12, No. 1, Fall 1978, Logical relationships p. 21. . Storage Alternatives (2) " Curriculum '781 Recommendations . The Role of Data Base Adminis- for the Undergraduate Program in tration Computer Science," Communications - Data dictionary concepts of the ACM, Vol. 22, No. 3, March 1149. Communications (3) "Curriculum Recommendations for The fourth and final component of Undergraduate Programs in Informa- the information literacy offering, com- tion Systems," Communications of munications, should impart an awareness the ACM, Vol, 16, No. 12, December of the regulatory, cost, and technolo- 7V73. gical aspects of the links in a contem- (4) R. L. Ashenhurst, ed., "Curriculum porary information system, Recommendations for Oteduate Pro- Computer Literacy 61

fessional Programs in Information Systems," 4 retort of the ACM Curri- culum Committee on Camputer Educa- tion for Managemen, 2972. (5) J. Hamblen, "Computer Education in Nigher Education -- _Status, Alter- natives and Needs," AFIPS, :978. (6) "ACM Cites Dearth of DP Programs," Computerworld, November 5, 19)9.

7o 82 NECC 1980

A BYTE OF BASIC

Judith A. Hopper

Arapahoe County District 6 Grant Jdnior High Littleton, Colorado 80122 (303) 795-2560

INTRODUCTION The average U. S. citizen has not the class, to be one semester long (eighteen foggiest'idea of how computers work and weeks), met five days a week for fifty- how pervasive their influence actually is. five minutes. The Apple II is a 32K Consequently, he has no idea of what to system with a Centronics printer. Xt was do when a computer makes a mistake; he has felt a hard copy would be a benefit in no idea of how to vote on local, state, ur that the students would have something national issues involving computers (e.g., tangible to take home to show to their the establishment of a national data parents. bank)s he is, in short, culturally dis-_ The class was begun by acquainting the advantaged. students with computers in generals what It is therefore essential that our is a computer system, what are its parts, educational system be modified in such a and what does each part do. Using their way that every student (i.e., every pro- knowledge and expanding what they knew, spective citizen) become acquainted with the class discussed various types of the nature of computers and the current input/output devices as well as particular and potential roles which they play in usage, i.e,, in business, industry, our society. It is probably too late to science, and research. do much about adults, but it would be In order to get the students on the disastrous to neglect the next genera- computer as quickly as possible, flow- tion. charting and algorithms were tae next - Committee on Computer Education, topics presented. As it was important for Conference Board of the Mathe- the students to be able to break down a matical Sciences problem into its component parts and For today's children, understanding analyze each stzp of the solution, flow- computer fundamentals is one very charting was introduced as a pictorial important factor in building an informed representation of this pzocess. citizenry for the future. Students who Simple programming was introduced next are computer literate will have .setter to acquaint students with the computer. career opportunities and less career Feeling comfortable with the computer and shock. They will be better able to cope learning the on-off procedures in using in a world with rapidly moving and ever- the computer were the main objectives. changing technology. Thus, in this era The introductory programming was simple to called the Age of Information, schools insure each student initial success at the can no longer delay in bringing the terminal. computer into their curricula. Egg cartons were then used to build personal computers for each student. In A BYTE OP BASIC beginning to write more difficult programs Seeing the need to get some kind of the students had to put their program program with computers started in her through their "EGG 12" system first to be school district, the author of this paper sure it was doing what they had intended. wrote a proposal to pilot a computer Once it was demonstrated that it would run literacy and BASIC programming class at on their computer, they could run it on the junior high level. The district pur- the Apple.Using this technique forced chased an Apple II microcomputer and them to evaluate their own %ark and make things were off and running. corrections as needed. A class of fifteen eighth and ninth As the class progressed through the graders was chosen for the project. The semester, more and complex programming

73 Computer Literacy 63

was introduced. Most all of the common class. They discussed various jobs BASIC statements were used (LET, IF-THEN, directly related to using the computer, TAB, GO TO, REM, FOR-NEXT, PRINT, etc.) the training required, the responsibili- and also some library functions (SOB, ABS, ties that go with the jobs, and the avail- INT, RND). In addition, as the machine ability of openings in the field. being used has color graphic and string It was hoped that with the completion variable capabilities, these were soon of this course the students would have a added to the repertoire of the student working background knowledge of program» programmers. ming, that all fears of a computer would Interspersed among the programming have been extinguished, and that they lessons, other topics were presented. In would be aware that computers have great answer to the question of what does the potential and capabilities but also limit- computer do with the program once it has ations. If the development of the been typed in at the terminal, machine computer continues at the rapid pace at language and the binary system were dis- which it is now moving, the students' cussed. Changing numbers back and forth world will truly be a computerized world. from base two to base ten, and adding, Any knowledge of the field they can gain subtracting, and multiplying in base two now will serse to make their lives that were some typical class activities.A much easier upon completion of their comparison was also made of a program education. written first in BASIC, then in.assembler, and finally in machine language. FINDINGS *he concept of language interaction This first class was an enthusiastic resulted in the introduction of the group of fifteen students.They were origins of BASIC and the history of com- excited about working with the computer puting. Beginning with Stonehenge, the and could hardly wait for some hands-on abacus, and the slide rule, the class experience. This enthusiasm continued proceeded through Pascal, Babbage, .hroughout the semester. Each time the Hollerith, Aiken, Hopper, von Neumann, students wrote a program, they immediately and others. This discussion of the wanted to test their programming abilities, development of the computer and its gaining a great deal of satisfaction in changing faces served to illustrate watching the computer do what they had where we are in terms of both current intended. computing capabilities and conjectures But while there was enthusiasm, there about the future. was also some apprehension among the The class enjoyed learning to use students. One of the primary goals of the various computing devices, the abacus class was to overcome this fear.The being of particular interest to them. students need to be comfortable and feel The slide rule and calculator were also at home with the terminal. Providing them introduced as well as a ten-year old some initial success with computers was a Hewlett-Packard.2000 series computer, way of achieving this goal. which required marking Hollerith cards The biggest problem for the class was with a pencil. As the students were the availability of only a single terminal. exposed to these devices, they learned With fifteen students writing two and both the limitations and the capabilities three programs a week it was difficult to of each. schedule time so they each could use the An important part of any literacy computer at least twice a week. Some came class is not only learning what a computer in before v-hool, others during lunch and can and cannot do but also keeping up with after school, as well as during the full some of the literature available on com- hour being used during class. It is hoped puters. Throughout the course the this problem will be alleviated next students had to read a number of articles semester with the acquisition of four from various periodicals acquainting them additional microcomputers. with what was going on in the world of Another problem encountered was the lack computing. They had to bring current of typing skills in the students. Having articles in from magazines and newspapers to use the one-finger-hunt-and-peck method and write reports.The vocabulary of consumed a lot of time when they:were computer -ese was carefully ()is:111890d so reedy-to type in their programs.As the that as complete an understanding as r.4.ass progressed through the semester they possible was available. Any neY develop became better and faster at getting their meats of computer usage were also pointed programs in, but it remained a time- out and discussed. consuming problem throughout. Field trips were made to observe the Now that the first class is over, the use of computers outside the olassro,m, students are asking for a second class at and several people working in t:o com- a higher level.They are very interested puting field were invited 'do speak to the in the many facets of the machine and

74 84 NECC 1 980

would like to pursue its capabilities. tutoring individuals, or any other func- Unfortunately a second course in computing tions -- limited only by the imagination. is not currently available at the junior Promtaming in higher level languages high level (it is at the high school), but as well as practice in developing software it is hoped that these students can be would be offered at the senior high level. used as resource personnel.With their The proposal also recommends the computer ability to program and run demonstrations, be used as a tool in higher level math they may be called on by teachers wanting courses, in science and social science information or simulations. By using the classes, and in the business department computer in other classrooms, it can for data processing. berme an integral part of the entire As staff members at each of the schools, school curriculum. in addition to the students, would need exposure to the machines, the proposal FUTURE OUTLOOK also provides for inservices and actual The computer programming and literacy classes for faculty members not familiar class at Grant Junior High was a pilot with computers. %t is hoped that teachers class for the district. Before jumping in various disciplines would then be in with both feet the district felt it interested in the computers and want to best to try an initial program, evaluate use them as a demonstration tool in their that one, and go on from there. classrooms. Concurrently with piloting this class, The school board has just recently visitations to other school districts accepted this proposal for a two-year with a "computer curriculum already in full - implementation of microcomputers into the swing were made. As at Grant, there schools in the district. The hardware appeared to be good student interest, but requirements are not merely cost-effective also considerable frustration over the but also offer the opportunity to improve inadequate number of terminals available. the quality of education.Where once this Then, with recommendations from visit- type of technology was used by a few ations, attendance at several conferences, privileged persons, today it is becoming and research in the area of computer ubiquitous.A degree of flexibility has education, a proposal was formulated for also been built into the proposal so the implementation of microcomputers into changes or enhancements to improve the the school district. The proposal calls program can be made. for one junior high and one senior high in Education is missing an opportunity if the district to be fully equipped to begin not a responsibility if it fails to pro- the implementation. It is hoped that vide its students with the necessary these two schools working together would background to move into the "information be able to iron out any problems that age." We need to start now to provide the would arise so the full district-wide public with the understanding of this implementation would progress smoothly. information science and technology. As The proposal for the computer imple- J.C.R. Licklider put it "People must mentation specifies definite curriculum master the technology or be mastered by recommendations. At the elementary level it." the gifted and talented program already is doing some programming.The proposal EQUIPMENT LIST calls for some subtle exposure of the High Schools: other elementary students -- at least to Apple II 48K acquaint them with computers so that Video their fears are eliminated. Double disk At the junior high level it is planned RP Oscillator that all students at some time during Firmware Card their thtee-year career take a computer Radio Shack 4C! literacy course (possibly nine weeks). 32K addition Structured within the class would be a Double disk basic introduction to computing -- what 15 Radio Shack 4K Level II the computer can and cannot do. In Master slave addition to the literacy course a class Interfacing connectors in computer programming would be offered Printers where needed for those students who would like to Junior High Schools: pursue the programming aspect. Also it Apple II 48K is hoped other disciplines (math, science, Double disk social studies, and English) would want to Printer use the computer as a demonstration tool 8 Radio Shack 4K Level II for simulati Is, data processing for Master slave experimental r classroom recruits, Computer Literacy es

A COMPUTER WORKSHOP FOR ELEMENTARY AND SECONDARY TEACHERS Herbert L. Dershem John T. Whittle Hope College Holland, Michigan 49423 (616) 392-5111

INTRODUCTION resources for ideas, activities, and Although computers have been used by programs related to the classroom use of secondary teachers for a long time, only 'computers. recently has the microcomputer made it 3) The teacher will know the BASIC economically feasible for the elementary language well enough'to write simple pro- teacher to use the computer in the class- grams, to introduce the language to stu- room. IA addition, new technological de- dents, and to read any BASIC program and velopments have made it possible for make minor modifications to it. secondary teachers outside the fields of 4) The teacher will have sufficient mathematics and science to use the compu- understanding of tI.e way a computer works ter as a classroom tool. Recent reports to explain it to students. and recommendations (1,2) have emphasized 5) The teacher will know the techniques these facts and indicated the need to and approaches most frequently used in make teachers aware of the potential the instructional computer applications and computer possesses as a tool in the have experience in their use. classroom. 6) The teacher will have Cosigned and This paper describes a two-week work- implemented at least one computer activity shop which was offerod by the authors in for his or her classroom and be capable the summer of 1979 to provide teachers of developing others. with knowledge of computers as they are 7) The teacher will know the types of applicable to the teacheks' classes. computer equipment available for classroom use and be aware of advantages and ORIGIN AND OBJECTIVES or THE WORKSHOP disadvantages of each. Through working with students and The workshop was intended for teachers teachers at both the elementary and who had no previous experience with compu- secondary level, the authors became ters and who wished to explore ways in convinced that the computer is a valuable which they could improve their teaching educatipnal tool. But it was apparent by use of the computer. that the computer could be used effectively only if the classroom teacher was WORKSHOP FORMAT aware of its potential. We believed that The workshop was divided into two once the teacher learned to use the com- parts, a laboratory and a lecture, each of puter, he or she would be able to use it which met for one hour and fifteen min- in creative and innovative ways in the utes every day for two weeks. Upon com- classroom. The biggest step was over- pletion of the course, the teachers re- coming the teachers' fear and awe so that ceived two hours of college credit. they could implement their own classroom The BASIC language was taught in the computer applications or adapt those of laboratory portion of the course. Actual- others to their own needs. ly,"two laboratory sessions were held each With this in mind, we designed a two- day, one before and one after the lecture, week workshop with the following objec- half of the participants attending each tives: session. Ten Radio Shack Level II 16K 1) The teaoner will be sufficiently TICS -80 computer systems, loaned to the familiar with the operation of a computer college by Radio Shack,were set up in the to instruct others in its use. laboratory and each participant was seated 2) The teacher will know where to find in front of a unit. The textbook, "Using

76 66 NECC 1980

BASIC "by Julien Hennefeld (3), was chosen Both laboratories and the lecture were Wcause it contains many sample BASIC pro- held in the morning. The room containing grams to illustrate concepts. A set of the computers was also left open in the these programs was placed on cassettes afternoons so the participants could re- for each laboratory; then during the lab- turn and work on their projects. Although oratory each participant would load a pro- this afternoon work was not required, many gram from cassette. Next, the instructor teachers did take advantage of this oppor- would talk about the concept illustrated, tunity. in addition, several had TRS-80 have the participants run the program, systems at home or at their schools which and usually ask them to make modifications they used in the afternoons or evenings. to see the effect. The topics covered in the laboratory during the first week were RESULTS as follows: Twenty educators from three local school districts attended the workshop. Day 1 - How to use the computer Of these, twelve were secondary teachers, LET statements four were elementary teachers, and four PRINT statements were administrators. All four administra- Day 2 - GO TO statements tors attended the first week only. Also, iF-THEN statements some of the teachers who had previous Day 3 - INPUT statements computer experience took the second week FOR-NEXT statements only. Day 4 - Subscripted variables A brief description of the projects Grapiics developed by the teachers is found in the Day 5 - String variables Appendix. Both the instructors and the PRINT@ statements participants were amazed at what has been accomplished in a two-week period. An During the second week the laboratory enthusiasm for computer use was generated time was used for each participant to by the workshop so that participants went design and implement a useful classroom back to their schools and pushed for the computer activity. The instructors as- purchase of a computer for their classroom. sisted the participants in the design A further indication of the success of and programming of the activities. Addi- the workshop is the demand for it to be tional features of the BASIC language and repeated this summer. This demand is com- advanced programming techniques were cov- i%g from teachers who saw the effect the ered as they were needed. On the final workshop had on last year's participants. day of the workshop each of the partici- As a result, the same workshop will be pants presented the results of his or her offered twice this summer as well as once project to the entire group. during the academic year. Hope College has In the lecture portion of the workshop, now established its own microcomputer lab- the participants were divided into ele- oratory, so the computers will not be bor- mentary and secondary educators. The top- rowed systems. ics of the lectures for each group follow: The only modification to the workshop .which we plan is not to allow participants Day 1 - How Computers Work to register for just one week. Those who Day 2 - Techniques and Approaches in took only the first week missed the im- the Classroom portant experience of putting together Day 3 - Elementary - More techniques their own software. it was also difficult and approaches to adjust the workshop to those who came Secondary - Survey of Resources in during the second week. As a result, Day 4 - Elementary - Survey of Re- the workshop will not serve the needs of sources those teachers with some previous computer Secondary - Problem Solving experience. Principals interest has been expressed is a work- Day S - Computer Literacy shop specifically for teachers who are Day 6 - Design Considerations for already using computers. We are consider- instructional Hardware ing offering such a workshop, which would Day 7 - Experience with instructional cover advanced programming and software Software design techniques. in this workshop we Day 8 - Survey of Computer Equipment would invite each teacher to bring along a Day 9 - Elementary - More experience student so that they could participate as with instructional software a teacher-student team. From our observa- Secondary - How to Teach tions, much of the software developed is Computing actually done by such teacher-student Day 10 - Presentation of Projects teams with the teacher doing the design Computer Literacy 67

and the student the programming. options exist in this field and the education necessary for pursuing each option., REFERENCES 1) Milner, S. "An Analysis of Computer Test generator - This test generation Education Needs for K-12 Teachers.' Na- program simplifies the high school math tional Educational Computing _Conference teacher's job of creating examinations by PromeedNg1773WCity, 1979. randomly choosing problems which have 2) Taylor, R., Poirot,'.7., Powell, J., parameters that may also be randomly se- andtHamblen, J. "Computing Competencies lected. for School Teachers: A Preliminary Pro- jection for All but the Teacher of Com- Carrying drill - This program drills the puting." National Educational Computing student on carrying skills in multiplica- Conference PromeedailtriBirCity, 1979. tion by presenting randomly generated pro- J. EitasBASIC: An Intro- blems. If the student responds incorrect- duction to Computer Prograiragi. Winn ly the program carries out the multipli- WiW7Ntlaidt, Inc., 1978. cation, carefully specifying each carry value along the way. APPENDIX. PROJECTS COMPLETED BY PARTICI- -. PANTS IN THE COMPUTER WORKSHOP Mortgage payoff - The student can use this program to examine the effects of Physics experiment simulation - A falling varying the parameters on mortgage pay- body experiment is simulated by the cam- offs. pater. The student is asked to provide the appropriate formulas to calculate expected results.

Geometry drill - A drill and practice exercise is conducted using geometric terminology.

Spanish drill - A drill and practice on Spanish grammar and vocabulary is conducted entirely in the Spanish language.

Ordering - This program is intended for use at the elementary level. The student takes a list of items and places them in a described order. Three implemented orderings are alphabetical, fractions, and decimals.

Parts of Speech - The student is given a word list and after picking a word from this list, leads the computer to identifying the word by answering the computer's questions about its part of speech.

Word house. - This program is intended for students learning English as a second language. The student is required to place each of a list of words into its proper category and word house.

Presidential drill - This program drills students on presidents and their terms of office. Golf statistics - This data collection and analysis program simplifies the paper work of the golf coach.

Career counseling aid - A student inter - sated in a career in aocounting can sit down with this program and find out what 68 NECC 1980

MICROCOMPUTERS AND COMPUTER LITERACY; A CASE STUDY Robert J. Ellison Hamilton College C).inton, NY 13323 315-859-4138

I win discuss the role of the micro. plemented by both the mathematics andbiol- computer in computer education at Hamilton ogy departments for their own special ap- College. The focus will be on the kirds plications. Some of the materials will of applications I have found useful and on also appear in the development of a file the methodology that was used to develop maintenance system in the advanced comptter the systems on the microcomputer. My ini- science course on data structures. The tial interest in the microcomputer has limited professional support for academic been its use as an aid to teach computer programming in a small college practically literacy. I will discuss the development forces us to write generalpurposesoftwear. of an interac ive statistical package that is the core df a three week introduction THE ENVIRONMENT to computing and programming. The package Hamilton College is a liberal arts in- permits elementary programming and file stitution with an enrollment of 1600 stu- management to be introduced eithout an ext dents and a faculty of about 135. Over 600 tensive introduction to programming. students a year will take courses which While the microcomputer can be an ex- use the computer. The computer science of- cellent instructional aid, it quite often ferings consist of a two-course programis inconvenient for program development. ming sequence (6) and two advanced courses Most of our applications could not be de- which emphasize file structures and topics velope4 on a larger machine, however, in data base organization. The first pro- since we are very dependent on programming gramming course is taken by about one third the screen and the graphics displays. I of the students before they graduate. The will discuss our experience with the UCSD academic computing is done using a remote PASCAL(TM) system which is now available batch entry to the IBM 370/168 at Cornell on a number of different computers. It University. We will shortly be able to use has significantly aided the design and the interactive system at Cornell also. management of the project. The system's The advanced computer offerings are taught extensive use of prompt lines makes it easy using PASCAL and microcomputers. for the novice student to use. r am exploring the possibility of ap- It was my goal to develop systems plications which do not require the full which could be easily modified and expanded. capability of the IBM/370 or which can The microcomputer is not large enough to take advantage of the simpler operating support a universal package; furthermore, environment of the microcomputer. While the more complete systems usually also the Cornell system supports a quite adequate bring with them a more complicated control number of statistical packages and progimmr structure. Since our goal is to introduce ming languages, for the novice tLe system computing to an audience with little if any is still too complex to be easily used. A computer experience, we wanted a simple and number of applications involving elemen- often specialized user interface. But if tary program instruction or simple statis- we are not going to offer a fairly com- tical analysis might be better served on a plete package for the micro, then we smaller machine or one specially configured should design a system which can be for the use. easily modiried. The statistical package Learning the control language for the I will discuss was written using a library use of a computer system is a special prob- of procedures which can be integrated into lem for those non-computing courses that a custom system. The package has been sup- want to explore the use of the computer as 79 Computer Literacy 69

a tool. The focus is too often on the de- step of the process. SAS also supports tails of the system rather than the prob- limited programming in PL/1. Inmost cases lem to be solved. A well designed micro- the program code is applied to each record computer system might be of service here of the file. It seemed that a package such also. as SAS could serve to introduce the con- We have selected the TERAK microcom- cept of a file, program variables, and puter as our primary unit. The CPU is a elementary programming. Using files as the Digital Equipment LSI-11. The TERAK was primary way to pass data is ideally suited chosen for its graphics capability, a 320 for a microcomputer with limited memory. by 240 raster image, and the wide range of The use of SAS presented some prob- software available. Operating systems for lems for our short course. Our computing the TERAK include DEC's RT-11, the UCSD on the Cornell system is primarily done PASCAL(TM) operating system, which we use by remote batch entry. Although I preferred most often, and the Cornell Program Synthe- to use an interactive system for this sizer (13), which I will discuss below. course, the IBM control language for the The TERAK can also be used as a terminal interactive system was too complex for to Cornell, and with the use of a communi- our audience. It also seemed difficult to cations package text files can be trans- introduce in such a short time the syntax ferred between the two computers. Programs and semantics of even an elementary part exist to convert files between the RT-11 of PL /l. and the UCSD disk formats. Ihavedeveloped My solution to the above problem was software to transfer the graphics and motivated by the availability of aprograra- character buffers to a Printronix which ming system for the LSI-11 (13) called the has a plot capability. Cornell Program Synthesizer. The system includes an editor for the language PL/CS GOALS 4 (S), which gives a template for each major Our most immediate need was foresys- construction. For example, a single key- tem to support a new offering for computer board command generates the template literacy in our winter term. Hamilton follows a 4-1-4 calendar. It was my intent to If (condition) then offer a short course which couldhelpmeet (statement) the demand for computer literacy. Since our first programming course is designed Statements are checked as entered. The for a general audience, I did not feel the system is designed for teaching program- need to teach programming. On the other ming and includes elaborate editing and hand it seemed important to introduce stu- tracing commands. I wanted to include some dents to the use of the computer to store of these features to assist in entering information and hence to the concept of a elementary programs for the statistical file. system. I felt it was quite important to involve the student more than as a passive THE STATISTICAL PACKAGE observer of the computer. Ourgreatestprob- The system supports real, integer, lem was finding programs and applications and character variables. The core of the which involved the student with more than system is an elementary file management selecting an option from a menu. Iampar- system that includes modules to open and ticularly interested in examples which in- close file, get or put the next record, volved the maintenance of files on the com- and retrieve or store individual variables puter. An immediate application was the use in a record. Each statistical file consists of a text editor and text formatter. The of a data file and a directory file that UCSD screen editor is very easy to use. I lists the variable names and types. The had also written the formatter (9)inPASCAL. file management system is stored in the I added user-defined macros with the out- system library and can be called from any put directed to the printer, disk file, or PASCAL program. The system is designed for screen. While the text editor could be used instructional use, not to support large- to discuss files and updating of informa- scale statistical analysis. It can serve tion, I wanted a more complex example. as a tutorial for the use of a more com- The use of a statistical package plete, flexible, and faster system such as seemed a possibility. The choice was in SAS. The system consists of the following part motivated by use of the SAS system general programs. (12) which makes extensive use of both permanent and temporary files. A SAS run con- Create: sists of a number of procedures to list, he record and directory files are sort, and modify the file. Most procedures created from the text files made by the create a file to pass data to the next editor.

f 70 NECC 1980

4.4

List: It not only complicated the design of the The directory and the data are listed interpreter for the language but also on either the screen or the printer. Vari- created the possibility of infinite loops. able headings are automatically printed. A single record may take several lines. Input /output

Sort: Output (file name The file may be ordered by any of the * variables in ascending or descending order. The current record is written to the Multiple sort keys can be specified. It may selected file. When the file is opened also be used as a procedure in other pro- the user has the option of specifying grams. which variables will be written. The record has all the variables in the input file as well as any new variables intro- This module supports an elementary duced by the program. Currently two output programming language to permit modification files are permitted. The program prompts of the file. The statements are prompted for an output file if none is specified. and errors are noted on entry. The program If an expression or variable is not is saved on the disk for documentation. correctly entered,a list of the variables Currently, an individual statement can be and their types can be requested before edited as it is entered but cannot be the correction is made. The program is changed after it is accepted. I will not automatically indented to reflect the include elaborate editing, since most of nesting of conditionals. the programs are rather short and are easily The following is a sample program. entered again. The variables QU/ZI, QUIZ2, FINAL, and The statements supported and the tem- MIDTERM are on the input file. The vari- plate provided include; able TOTAL is created by this program. The lower case entries were supplied by Assignment: the program.

(variable name) = (expression): TOTAL = 0 ; if NOT(QUIZI MISSING) then do; New variables are declared simply.by TOTAL = TOTAL + QUIZI entering the name. The variable type is end; automatically set by the type of expression. if NOT(QUIZ2 MISSING) .aen do; TOTAL = TOTAL + QUIZ2 ; Conditionals: end; TOTAL = TOTAL + FINAL + MIDTERM ;

If (condition) then do: output DEMO ; (statements) end; The system was used for the first time this winter. Our program editing was Select; not as elaborate as apparently was needed, when (condition) do; but the overall student response was quite (statements) positive. To the student, the basic pro- end: gram unit is the file which for a general when (condition) d0; course on computer literacy was appropri- (statements) ate, The limited programming also helps end; explain how a computer workst We currently plan to use the same system with our first. programming course to introduce files and otherwise do; interactive computer systems. A variant of (statements) the system will be used in the first year end; biology sequence. end; The system is easily expanded. New programs are written in PASCAL with the The SELECT statement has been recently PASCAL input/output functions replaced by added to the IBM version of PL/I. Only the those in the new file management system. statements associated with the first true It is particularly easy to have a student condition will be executed. The OTHERWISE develop an independent PASCAL program for block may be empty. The SELECT was used a specific file and then replace the input/ instead of an IF..ELSE construction as the output functions to generate a general former seemed to be less confusing. I purpose program. The system, especially chose not to implement the GO TO statement. the file management section, has been an 8j Computer Literacy 71

excellent source of examples for the advanced A procedure GETDIR whose purpose is to computer science students. read the directory could be declared with the following heading: DEVELOPMENT I would like to discuss the way we Procedure GETDIR( VAR D; DIRECTORY); have managed the development of this system and the use of UCSD implementation of If it is necessary to add more informa- PASCAL. A small college often does not have tion to our description, such as the num- the p ogreaming support available fox the ber ok missing values, then we only need development of large systems. The prolif- add a line such as eration of a number of specialized packages would only compound the maintenance missing ; integer; of these programs. While the initial effort is greater, I have tried todevelopgeneral- to the original definition. That change purpose programs which can serve in other is then reflected in all procedures which applications, for example, the file man- use that data type. agement system which supports both the The UCSD system has included some statistical package and general file main- non-standard features. I have tried to tenance programs for an advanced computer avoid the use of them when possible, but science course on file structures. The their definition of a unit has been quite graphics library, which I have not discussed, valuable. A unit is a separately compiled is used by mathematics, biology, and physics. library of subprocedures. It can also in- It is clearly desirable to reuse soft- clude type definitions. I keep the system ware, but keeping each of the various sub- type definitions in the library. Those systems supplied with the latest revisions types are then copied into each program can btl difficult to manage on a small sys- which uses that unit. Changes in the com- tem. Programming in PASCAL and the UCSD mon definitions are then reflected in all implementation of it have both assisted in programs which use that module. It has this task. A vital feature of PASCAL is been an excellent tool to maintain consis- the ability to define new data types. tency in student written programs. PASCAL includes the simple variable types The procedures and definitions can integer, real,and character. But for the be included in the program by the simple kinds of systems programming involved in command this project, we needed more elaborate structures. For example, we had to main- uses {unit name }; tain a directory for each file which maintains the variables and their type. The The statistical system has a unit devoted deini+ion of such a directory had to be to file management, another to parsing and consistent throughout the system. In PASCAL evaluating expressions, and a third unit we could define a directory in several to provide run-time support. In addition steps. we have other units with graphics procedures or random number generators. The use Const {compile time constant} of units does not affect portability cf Namelength = 10: the system. The source code of the procedures could be inserted instead of the Type uses statement.

memestr . packed array(1..Namelength) SUMMARY of char; {string} We are continually surprised by the Vardescript m record ease with which major changes can be made { describes file variable } to the system whose length now exceeds Name : namostr; 5000 lines. The use of PASCAL and the UCSD Position integer; unit has played a major role. It has been Typeofvars (realtype,integertype, very easy to develop the system incremen- chartype) ; tally and to quickly train students to Length; integer { length in bytes) assist in the programming. PASCAL is well Max ; real; suited to the kind of systems programming Min real; involved in this project. end We have been pleased with the student reaction to the statistical system. The We can then declare the type directory as combination of that system along with the PL/CS programming system has provided a directory m array(1..maxvars] quick introduction to programming and the of vardescript; use of files for our course on computers

82 72 NECC 1980

and society. We are planning to develop 13. Teitelbaum, R.T. "The Cornell Program tutorial materials so the system can be Synthesizer; A Microcomputer Implemen- used by other courses. Further evaluation tation of PL/CS." Technical Report 79- of the system will have to wait until it 370, Department of Computer Science, has been uzed by other disciplines. We Cornell University 1979 have been successful in having other faculty use the modules to develop their own 14. Van Loan, Charles. hComputer Science specialized software. We are now evaluating rnd the Liberal Arts Student." Techni- what additional modules should be placed cal Report 79-376, Department of Com- in common units. puter Science, Cornell University 1979 REFERENCES

1. Austing, R.H., Barnes, B.H., Bonnette, D.T., Engel, G.L., and Stokes, G.S. "Curriculum '711; Recommendations for the Undergraduate Program in Computei Science--A Report on the ACM Curriculum Committee on Computer Science." Comm. ACM 22, 3(March 1979), i47-L66

2. Austing, R.H., and Engel, G.L. "A Com- puter Science Course for Small Colleges." Comm. ACM 16, 3(March 1973), 139-147

3. Committee on the Undergraduate Program in Mathematics of MAA. A compendium of CUPM recommendations. /IAA 11 (1975), 52S-570

4. Conway, R. Primer on Disci lined Pro- gramming UsIRFE/ag.-Iritt top9:1-

5. Conway, R. and Constable, R. "PL /CS -- A Disciplined Subset of PL/I." Techni- cal Report 76-293, Department of Compu- tar Potence, Cornell 1976

6. Ellison, Robert J. "A Programming Sequence for a Liberal At College." Proceedings of the 1980 SIGCSE Sympo- sium on Cdbeuter Science Education

7. Grogono, Peter. Pro rammin i. Pascal. Addison-Wesley,

S. Jensen, Kathleen and Wirth, Siklaus. Pascal User Manual and Report. 2nd arfan:SFr_rWViiEreg 1974

9. Kernighan, Brian and Plauger, P.J. Software Tools. Addison-Wesley 1976

10. Lopez, A.A., Raymond, R., and Tardiff, R. "A Survey of Computer Science Offer- ings in Small Liberal Arts Colleges." Comm. ACM 20, 12(Dec 1977), 902-906

1.1. Nevison, J.M. "Computing in the Liberal Arts College.`' Science 194 (Oct. 1976), 396-402

'12. SAS Users Guide 1979 Edition. SAS 17WIENZIWIT 3 InVited Session

PERSONAL COMPUT/NG: AN ADVENTURE OP THE MIND.

Paul Hazen Applied Physics Laboratory Johns Hopkins University

ABSTRACT firgrants from the IEEE Computer Society, The Johns Hopkins University, Radio Shack, and other agencies, the International Instructional TV Cooperative, source of instructional TV materials to all educational TV networks nation-wide and internationally, has finished and is marketiol the implementation of a six- course national educational TV series aimed at the pre-college level in the area of personal computing and computer literacy. The name of the project is "Personal Computing:An Adventure of the Mind.* The objectives of this new series are to illustrate the uses of perscAal computing, to demonstrate the interface of humans and machines, to identify the fundamentals of communication in personal computing, and to motivate students to be Annovative in their own applications of personal computing. Since the personal computer is viewed by many as a mind multiplier, a furthcr objective of this educational TV ser.las is to greatly increase the number of minds that can be multiplied by taking personal computing to millions of children in classrooms across the country. Education and informational programs are closeli allied in that both attempt to communicate facts, concepts, and ideas. Both need to be designed with specific objectives in mind. Some of the objectives to be discussed.e both attitudinal and informational in nature; that is, they deal with feelings as well as facts. The underlyi:q thrust through-

out is that . . . LEARNING CAN BE FUN!

8.1 73 invited Sessions

EDUCATIONAL COMPUTING: EMT, PRESENT, FUTURE

Ronald W. Collins Dept. of Chemistry Jefferson Science Bldg. Eastern Michigan University Ypsilanti, MI 48197 (313) 487-0106

ABSTRACT During the past ton years considerable for most educational uses, the need for effort has been devoted to optimizing the prograMining time and exper 4.ca will still role of computers in education.A vari- be high. Furthermore, the several gener- ety of tutorial CAI programs have been ations of changes in hardware have not written, large data bases of questions yet led to pignificant improvements in for computer-generated exams have been computer-basd pedagogy.Educators amassed, sophisticated software for simply must give more consideration to graphics has been developed, numerous the question of how the computer can data reduction and simulation programs reshape the way and atyle it hich we have been written, and the use of on-line teach. To date, most instructional com- classroom computing has been studied. puting has simply becn appended to the In addition, the emergence of minicom- traditional pedagogical framewc.k, thus puters and microprocessors has lowered encouraging the self-fulfilling prophecy the cost of computing. Nevertheless, that the final impact will be minimal. the overall impact of'the computer on The future? By 1990 (or 2000 at the education has been minimal. One of the latest), there should be a number of new major deterrents has been the poor trans- and outstanding examples of true computer - portability of programs from one computer based/- derived /- oriented projects in to another. As a result, educators are education, rather than a mere continua- often required to develop their own tion of the current computer-augmented, software for use in courses; however, yet traditional approach. many instructors have neither the time nor the experti:,e to do so. Will the new stand -alone home computer systems improve this situation? Possibly, but

74 MAW Someone 75

THE OPEN UNIVERSITY

Prank Lovis The Open University Milton Keynes, England ME7 6AA (0908) 653371

William Dorn Mathematics Dept. Universik of Denver Delver, CO 80210 (303) 753-3529

ABSTRACT --TH-TTE2 Britain's Open University will The present paper will concentrate on: replace the two second-level computing (1) Why the DP bias was considered courses which it has been running since essential and how it shows-up in the 1973. There will be wily one new course course. since we ha's come tc realize that running (2) The content and presentation of two second-level courses is unnecessary the first four main themes listed above. and extravagant. The new course is ex- In this connection the systems analysis pected to maintain the current enrollment and design package bears specIsl mention, o. 700 students per year. as it is being developed in collaboration Production of the new course is well with the O K.'s National Computing Centre underway, and in this paper its content and will involve the student in a practical project. and presentation will be compares with those of its predecessors. Tile new (3) The specification and future imple- course was designed after careful consid- mentation of OUSBASIC, a computing language ation of the views of both students and designed to enable the student to write faculty members from which emerged the and run structured programs in the unique- - one urgent and clear demand that the and awkward -- distance teaching environment of The Open University. course should relate closely to the The paper will be illustrated with two commercial world of data processing. The main themes of the course are excerpts, of ten to fifteon minutes duratic practical computing, data structures, each, from films of two of the television files and file processtAg, systems programs- -The Introduction to Files and "ile Processing and Systems Analysis and analysis,and design, and the social impact of computing. Introductions to Design. operating systems, data bases, and distributed computing are also included. The course components comprise written materials, computer programming activities, 16 television programs, three teaching audio-cassettes, and six broadcast radio programs.

c*c Science and Engineering

DEMOGRAPHIC TECHNIQUES IN ECOLOGY: COMPUTER-ENHANCED LEARNING

A. John Getz, Jr. Department or Zoology Ohio Wesleyan University Delaware, Ohio 43015 (614) 369-4431

INTRODUCTION death rates (1 ,).and age specific birth Judging from the availability of CON- rates (mx) must be known and transcribed DUIT programs on ecological topics (5 of into appropriate format. For survivorship 8 in biology in the Summer 1979 issue of data, this means knowing the proportion Pipeline), there may well be more com- of individuals surviving at the start of puter use in courses related to ecology each age interval; and for fertility than in courses in other areas of biolo- data, this means recording the number of gy. For instance, valuable CONDUIT pro- female offspring per female in each age grams exist that aid students in learning inter.-al. From these data, numerous about population growth using the logis- descriptive functions can be readily tic equation and related models, inter- calculated by formulae of varying degrees specfic competition using the Lotka- of complexit (see Mertz, 1970 or Krebs, Volterra equations, and energy flow 1978 for a fuller description of these through the trophic levels of various quantities). These fUnctions or .uanti- ecosystems. Additional areas in ecology ties include the following. are appropriate for computer-enhanced learning. In particular, the use of the (1) The gross reproductive rate, G.R.R., computer to aid students in under- is a hypothetical quantity that indi- standing life tables is described here. cates the multiplication rate per genera- Such work not only helps students under- tion if females suffered no mortality stand these complex tables, but also per- prior to completing reproduction at the mits them to quickly and easily get a rates specified in W fertility tables. grasp of the consequences, in terms of bosh population growth rates and age dis- G.R.R. =Emx tributions, of alternative reproductive x=0 strategies. (2) The net reproductive rate, Ro, is the SIGNIFICANCE OP LIFE TABLES actual multiplication rate per genera- In many ways, a survivorship and fertili- tion if the population follows both the ty life table represents the ultimate mortality and fertility schedules in the synthesis of data on the life history life table. characteristics of a given population or species of organisms. Both age specific

8 '7 76 Science and Engineering 7

no i!101M ments for the computer. Co:18149ring first xmO the arguments against hand caltulation, there is the unfortunate, but nonetheless (3) The mean length of a generation, 0, real truth that many undergraduate stu- is given by: dents are terrified by the sorts of summation signs and exponents intrinsic to lxmxx several of the demographic formulae. At 0 a best, such students may laboriously plug Ro through their calculations and then finish with absolutely no confidence in (4) The innate capacity for increase in the accuracy of their results. At worst, the particular environmental conditions some of these students may not even be for which the table was written, rm, is able to perform the calculations. Clear- a factor that gives instantaneous growth ly, neither of these situons is con- rates. The calculation of rm must be ducive to students' learning about done by trial and error in the formula: ecology. On the other hand, for under- -rmx graduates capable of manually performing L. e lx.s.x 1. the calculations required, such an exer- x 0 cise rapidly becomes a numbtaglr mechani- cal process that, too, is not conducive (5) The finite rate of increase,X,..is to their learning about ecology. the factor by which population size changes por age interval in the life As for the arguments for use of She com- table, e.g., per year if the life table puter, three points.oan be made. First, is in years. for students ever more used to punching numbers into either calculators or computers, one additional application of /' a 0 1111 the computer generates very little anxiety. Math phobias can stay submerged. (6) The stable age distribution is the Second, the fast and accurate results proportion of organisms that would be in provided by the computer maintain both each age category if the population con- the interest and confidence ox the stu- tinued to grow indefinitely according to dent. Moreover, if the student wishes to the schedules in the life table. The pro- test himself and his ability to manually portion of organisms in each age category arrive at the same answers, the computer x to x+1 is given by output does provide such a check. Final- ly, because the tedious parts of thi Ilx exercise are done by a tireless computer, es the student is free to concentrate on x E "4 interpreting the biological significance im0 of his findings. This is precisely the concentration that most ecology profes- where I, like x, is a subscript indicat- sors presumably desire, and I have found ing age. . ) useful in the present instance.

While the data that are used to generate DESCRIPTION AND APPLICATIONS OF LIFET a life table in the first place are quite With the above points in mind and two intuitively understandable, e.g., the years' experience teaching about life number of female offspring produced by an icables without the aid of a computer, I average six year old female, the derived wrote a short program, LIFET, in BASIC terms, especially rm, are much less easi- for our PDP 11/TO at Ohio Wesleyan Uni- ly grasped. Because a thorough conception versity. LIPET would be suitable for use of ris desirable in conorete terms on other similar RSTS/E time-sharing (suck as numbersof offspring and num- systems or on minicomputers, and a pro- bers of seasons of reproduction) before a gram listing is available on 7equest. In student starts working with logistic brief, the program asks the student to growth models, a thorough familiarity enter the age specific survivorship and with all the various life table parameters fertility data for the population in is highly important. question and then proceeds with calculations of the first three of the six RATIONALE :( 'ONPUlER USE quantities defined above. The innate The reasons using the computer in life 'capacity for increase is calculated by table analyst flan be divided into argu- trial and error with active participation ments against h,nd caloulation and argu- by the student in the trials. When the

tti 8 (6,, 78 NECC 1980 ts

student .is satisfied with the value of There are at least three sorts of appli- rm that he has attained, the program cation of this proc:1Lm. First, and most uses that value to aalculate the Finite simply, is rarely giving the students rate of increase. Finally, the student some real data from a natural population is given the option of obtaining a (e.g., from Vinegar, 1975 or Medica and stable age distribution or not. A sample Turner, 1976). The students can organize program run is given in Figure 1. the data into life table format for RUN LIFET DO YOU WANT AN EXPLANATION OF THIS PROGRAM (YES OR NO)? YES

THIS PROGRAM CALCULATES LIFE TABLES AND DEMOGRAPHIC PARAMETERS. THE ONLY INFORMhTION NEEDED TO MAKE THESE CALCULATIONS IS THE AGE-SPECIFIC SCHEDULE OF DEATHS AND BIRTHS WHICH YOU SUPPLY AS 1(x) and m(x) VALUES WHERE: 1(x) = PROPORTION OF POPULATION ALIVE AT THE START OF AGE INTERVAL x; AND m(x) = AVERAGE NUMBER OF BIRTHS TO EACH FEMALE ALIVE AT AGE x

TO RUN THIS PROGRAM, MERELY ENTER THE MAXIMUM AGE TO WHICH FEMALES SURVIVE (RECALL THAT LIFE TABLES ARE FOR FEMALES ONLY) AND THEN FOR EACH AGE CATEGORY ENTER THE 1(x) and m(x) VALUES SEPARATED BY A COMMA. THESE DATA CONSTITUTE ALL THE INPUT NECESSARY TO MAKE DEMOGRAPHIC CALCULATIONS USING THE FORMULAE IN KREBS (1978, pp 160-170). SEE THIS REFERENCE FOR DEFINITIONS AND EQUATIONS. ONLY THE.EXACT r CANNOT BE CALCULATED WITHOUT ADDITIONAL INPUT ON YOUR PART. THIS IS BECAUSE THE EQUATION MUST BE SOLVED BY TRIAL AND ERROR, i.e., BY SUBSTITUTING ONE VALUE FOR r IN THE EQUATION -rx SUM e 1(x) m(x) = 1 AND THEN SEE/NO HOW CLOSE THE SUM COMES TO ACTUALLY BEING EQUAL TO I. FOR /JUR FIRST TRIAL, USE THE VALUE FOR r APPROXIMATED BY Rc AND G. IF ALL REPRODUCTION TOOK PLACE IN THE SAME YEAR, THIS WILL ALSO BE THE EXACT r. IF THE SUM YOU GET IS GREATER THAN 1, INCREASE YOUR EST/MATE OF r. IF THE SUM IS LESS THAN 1, DECREASE YOUR ESTIMATE OF r. INITIALLY MAKE FAIRLY LARGE CHANGES IN r AND ONLY LATER MAKE MORE SUBTLE CHANGES AS YOU ADJUST THE SUM TO 1 +1-0.005. ONCE YOU ARE SATISFIED WITH YOUR VALUE FOR r, STOP ADJUSTING IT AND THE PROGRAM WILL CONTINUE TO MAKE OTHER CALCULATIONS USING THIS FINAL VALUE OF r ON WHICH YOU HAVE DECIDED.

WHAT IS THE LAST AGE AT WHICH SOME FEMALES ARE ALIVE? 3 FOR X = 0 WHAT ARE THE OBSERVED VALUES FOR 1(x),m(x)? 1.0,0 FOR X = 1 WHAT ARE THE OBSERVED VALUES FOR 1(x),m(x)? .9,2 FOR X = 2 WHAT ARE THE OBSERVED VALUES FOR 1(x),m(x)? .7,1 FOR X 22 3 WHAT ARE THE OBSERVED VALUES FOR 1(x),m(x)? .5,1 THE GROSS REPRODUCTIVE RATE, G.R.R. = 4 THE NET REPRODUCTIVE RATE, Ro= 3 THE GENERATION TIME, 0= 1.56667 AS APPROXIMATED BY Ro and 0, r=.701242

THE PROGRAMWILL NOW CALCULATE AN EXACT r by TRIAL AND ERROR. WHAT IS THEVALUE FOR r THAT YOU WISH TO TRY? .7 SOLVING THEEQUATION WITH r= .7 GIVES AN ANSWER OF 1.1277 DO YOU WISHTO TRY AGAIN WITH ANOTHER VALUE FOR r(YES OR NO)? YES WHAT IS THEVALUE FOR r THAT YOU WISH TO TRY? .8 SOLVING THEEQUATION WITH rom 8 GIVES AN ANSWER OF .995479 DO YOU WISHTO TRY AGAIN WITH ANOTHER VALUE FOR r(YES OR NO)? YES WHAT IS THEVALUE FOR r THAT YOU WISH TO TRY? .79 SOLVING THEEQUATION WITH r= .79 GIVES AM ANSWER OF 1.00784

Figure 1. Sample output from LIFET Science and Engineering 79

Figure 1 continued DO YOU WISH TO TRY AGAIN WITH ANOTHER VALUE FOR r(YES OR NO)? YES WHAT IS THE VALUE FOR r THAT YOU WISH TO TRY? .795 SOLVING THE EQUATION WITH r=795 GIVES AN ANSWER OF 1.00164 DO YOU WISH TO TRY ',GAIN WITh ANOTHER VALUE FOR r(YES OR NO)? NO THE FIRM RATE OF INCREASE, LAMBDA= 2.21444 THE EXACT INSTANTANEOUS RATE OP INCREASE FOR THE POPULATION, ra .795

DO YOU WISH TO CALCULATE THE STABLE AGE DISTRIBUTION FOR THIS POPULATION? YES

GIVEN A STABLE AGE DISTRIBUTION, TIE PROPORTION OP ORGANISMS IN EACH AGE CATEGORY, C(X), WOULD BE: Of 0 ) = .626875 C( 1 ) = .254776 c( 2 ) m .C89485 c( 3 ) = .288641E-1

x 1(x) m(x) 1(x)m(x) 0 1 0 0 1 .9 2 1.8 2 .7 1 .7 3 .5 1 .5 121° 1.56667 GREs= 4 Rots3

r= .795 lambda= 2.21444

DO YOU WANT TO CALCULATE ANOTHER LIFE TABLE (YES OR NO)? NO

themselves if desired, or the data can holding their own and others are decreas- be presented as a table from the start. ing in size. With analyses completed, Either way, this students have the ex- students can either write about or dis- citement of finding out for themselves cuss hypothetical alterations in repro- whether various populations in nature ductive strategy that could improve the are growing or shrinking and at what lot of the marginally surviving popula- rates. If population growth is occur- tions. ring and data are also available on numbers of individuals alive in each age A third and final example of an applica- class, the students can also investigate tion for this program is to give students whether or not such population growth free reign to devise an optimal repro- has been going on at the given rates for ductive strategy in each of several a number of years. If it has, the ob- habitats tot a totally LypothOical served proportions of individuals in organism. A few guidelines need to be each of the age classes should be the laid down at the outset, and then stu- same as the proportions given by the dents can be set free to uoe their LIFET calculations of the stable age creativity. For example, one needs to distribution. Students seem to enjoy specify how many kilocalories total a checking this assumption of 1:Se table female has to produce eggs in her life- methodology. time, the range in sizes (caloric contents) of eggs to be permitted, and the A second application is the use of data harshness of the various habitat' in not just from one population of a terms of probabilities of the survival species, but from multiple populations of offspring to age 1 from each possible of a single species. The wcrk by Tinkle egg size Zn each habitat. Survivorship and Ballinger (1972) on intraspecific thereafter an either be specified or comparative demography of a lizard is :eft to the students' own devices. The especially appropriate. By analyzing the generation of an optimal strategy, or age specific death rates and fertility even possible strategies, for each en- rates using LIFET, students can enhance vtronment then "bequires that students their understanding of the population generate a number of life tables. Suc- consequences for real al.lmals of vari- cessful completion of such an exercise, ations !n life history strategies. These in my experience, results in an excel- particulan data, in fact, indicate that lent appreciation of the merit° and some of the natural populations are just drawbacks of iteroparity and semelparity

9 u 80 NECC 1980

in various situations. 6

SUMMARY AND CONCLUSIONS For the past two years, I have used an interactive computer program, LIFE?, as a means of enhancing my instruction of demographic techniques in an undergraduate course on animal ecology. I have found that I have been able to cover more examples and nuances of life history studies since I initiated the program than,I,was able to before. By obviating the time-consuming busy work of life table calculations, I can realistically expect my students to accomplish far more sophisticated and extended problems than were possible before I wrote LIVET. In my opinion, students who have had the benefit of using this program have gained a far greater understanding of life tables and demography than did my earlier students. Certainly, they have had more opportunities to do so in a challenging way. What's more, they enjoy it. This last point, in and of itself, can be a strong recommendation.

REFERENCES Krebs, C. J. 1978. Ecology ; 'The Experi- mental Analysis of Dietributionmnd Abundance. 2nd ed. New York: Harper & Row. 670 p. Medica, P. A., and F. B. Turner. 1976. "Reproduction by Uta stansburiana (Reptilia, LacertITTa, Iguanidae) in Southern Nevada." J. Hernetol. 10: 123-128. Mertz, D. B. 1970. Notes on methods used in life-history studies. In: Readin e in Ecology and Ecological Gene. es, J. H. Connell, D. B. Mertz, and W. W. Murdoch (eds). New York: harper & Row. PP. 4-17, Tinkle, D. W., and R. E. Ballinger. 1972. "Seelo orus undulatus: A Study of the Intrespeci ic Comparative Demography of a Lizard." Ecology 53:570-584- linegar, M. B. 1975. "Demography of the Striped Plateau Lizard, Sceloperus virgatut." Ecology 56:172-182. Science and Engineering 81

MICROCOMPUTERS AS LABORATORY INSTRUMENTS: TWO APPLICATIONS IN NEUROBIOLOGY.

Richard F. Olivo Department of Biological Sciences Smith College Northampton, Massachusetts 01063

for that incorporates a versatile editor, cost of has The low microcomputers a disassembler, and a pseudoassembler brought them into the same price range as that permits writing programs in mnemonics ordinary laboratory instruments, and their rather than op codes. A full symbolic as ability to collectand store data makes sembler, which 1 used for writing our pro them an extremely welcome to a addition grams, is optional, as is an SK BASIC laboratory. At Smith College, we have de (both of these are supplied as readonly veloped two applications for using micro memories that plug into designated sockets computers in neurobiology. Pne application on the AIM board). The assembler and BASIC is suited toadvanced st,Jents and re sockets will also accept 2716 erasable, searchers and involves the use of a micro programmable readonly memories (EPROM;). computer and a digitizing tablet tomoos Once a program is debugged,it can be lo Orephotographic data; the other is suit aded into an EPROM to be installed in the able for routine use in aneurophysiology assembler socket, permitting a student to courseand involves an analog/digital in run a program by typing "",°the monitor terface for collecting and displaying call to the assembler. The student does transient data. In describing these two not need to be anaccomplished computer applications;1 shall emphasize a number user) the programs are highly interactive of aspects that 1 believe are of general and provide abundant prompts for entry of interest in introducing microcomputers data and for menu choices. into undergraduate laboratories. In addition to the AIM, its enclosure, and a power supply, two other pieces of HARDWARE: THE AIM 65 hardware are necessary. The AIM'. 4K bytes 1 chose Rockwell'sAIM-65 microcom of onboard memory are not sufficient for puter for laboratory use. Like its smaller extensive digitization, so that asupple cousins, the KIM and SYM, AIM-65 is a re mentary memory board is desirable.1 chose latively inexpensive singleboard computer the Memory Plus, an SK board from The Com that is based on the 6502microprocessor. puterist (P.O. Box 3,S. Chelmsford, MA The AIM includes an input/output interface 01824), which has the additional feature we With two 8bit parallel ports, which of an EPROM programming circuit. Other useto connect the computer to laboratory manufacturers also make A1M /KIM compatible instruments, plus two timers and an inter-,. SR and 16K memory boards. The other major rupt register. The AIMalso has a full piece of hardware required is an kaebOard; a 20character alphanumeric dis analog/digital interface. Commercial A/D plays and a thermal printer.I consider it boards that caninterface to the AIM's an advantage that the AIM does not use a input/output ports are available, but the video monitor, since video would make the ones that I am aware of are too slow for system less compact and more expensive; digitization at the rates needed (about 10 the ATM's oneline display is adequate for kHz) in a neurobiology lab. Consequently, prompts and date, end the outpOt ports we built our own analog interface, which I providea means (with a digitaltoanalog shall describe later. At present we have converter) of plotting data at high reso two AlMs at work. One is a prototype sys !Utters on an oscilloscope or chart record tem with 12K of memorAnd a single analog er. TheAIM'. printer further provides input/output channel; its cost is about eachgroup ofstudents with inexpensive $1000, and we are seeking funds to buy and hard copy. build six more such systems for routine The AIM is also relatively convenient class use. The other AIM has 4K ofmemory to program. It has an extensive (8K) moni and is devoted exclusively to taking data 82 NECC 1980

from a HiPad digitising tablet (Houston skid to read or generatehexadecimal Instruments). That ssetoe's cost, includ numbers. Since oh* 6502 microprocessor can ing the digitising tablet, is about *1400. be set to do BCD arithmetic, it made sense For ,comparison, these prices are of the to start, compute, and end in BCD rather same order of magnitude asa laboratory than to convert to or from hex. oscilloscope or a chart recorder.

APPLICATION It BCD DATA FORMAT IN/111FACINS TO A MIMING TABLET 1 shall first describe our system for cool OTOMMKth collectingdata using Niro/ digitising oascommatesus ummood tablet. The program has been through sev aloudWard 1 1 1 1 1 1 1 x lxixlxtotem eral stages of design and is now in heavy :emu 0I 0 Isot kw 2n0 OYU regular use. It illustratesseveral as X kith 2o1 SO are SD 310 BYTE X Allil 4th SO LSO AA OrM pactsof interactive laboratory computing Y kth 0 1 0 1Op 1430 sir srm thot I believe r of worst importance. Y kith tad SD the SO NO OrM Nourephssiologiststypically msisura Y Aith thlt SD LED Oh 10111 theelectrical changes in nerve or muscle cells in response to stimuli. The stimuli and responses are displayed on an oscillo DATA TIMM scope screen and era usuallyphotographed on 35-me film by an automatic camera. The photograph* serve as primary data for sub 11476 empnt analysisof thetraces and for illustratiogs for publications. Analysis raquiros measuring the amplitudes and time OM SINK courses of the events, which is usually den by projecting the film in an enlarger onto graph war and (boiler, microcom puters) to/lovely counting 'soar's. A di gitising tblot under the graph paper is huge improvement* it takes data utmeti Mt= 1. cells, *poets up the *nise!' by afactor of at least ten, and is intrinsically mars accurate than hand analysis. For those who have never seen a small digitizing, tablet. In its BCD format, $0, HiPd tablet it consists of a flat surface (ours is 11 outputs seven bytes of data for each digi inches suer*)in which wires are imm/ tised point. The contents of these seven des), plus a cross hair cursor that Is bytesare shown in Fig. 1, which also placed on the surface and returns the X,V shows the timing of the data. Note that ceardintos of its position. In using a Opt,1 is a control word. bytes 2,2, and digitisingtablet. two problems most be 4 contain the Xcoordinate. andbytes 5. solved: interfacing the microcomputer to 6. and 7 containthe Vcoordinate. The the tablet, and creating an efficient pro timing diagram shows that the BCD strobe ves for recordists and analysing data. I line goes high aseach byte is output. shall describethe interfacingaspects Thus. a subroutine to read data into the first. microcomputer musttest for the presence The electronics in theHiPad tablet of the strobe pulse, then read the current output data in Home' formats. any of byte, store it, end repeat these opera which meg be suppressed by jumpers atthe tions until seven bytes have been read. I output connector. Different strobe lines connected theBCD strobe to control line aro available for each of the formats. so CAL and the light data lines to the A1M's that a microcomputer can be made to attend Port A. In the 6502 system, input/output to one format and to ignore the others. 1 ports and associatedregistersare ros/ chose to take data in binaryceded decimal fromandwritten to as memory locations, (BCD) form rather than in straightbinary so that operations on the strobe pulse and (hexadecimal), the alternativeparallel data resemble reading data from memory. format. The main reason for this choice is The full subroutineto take adata that the final output had to be in decimal point is shown in Figura 2. When the so form since hexadecimal data are incempro brotine is called, themicroprocessor heNtlible tomost biologists. I believe enters a threeinstruction loop in which *at an important principle casts here: a it remains until the strobe pulse occurs. studentor researcher should never be In this loop. it checks the interrupt flag

9 3 Science and Engineering 83

register 4SFR, part of the 6522 part of the program then begins; in this input/output chip) to see if line CAI is section, a series of points is taken for high. It then reads Port A. checks for the each frame and stored in tables in memory. tpected control word, and if it is found. Figure3 shows the print-out made during proceeds to read and store theremaining the early part of running the program+ and sir bytes of data. Once the sir X,Y bytes Figure 4 shows pert of the tables that are have been stored, they ererepacked to printed after all frames have been meas- placethesignofeach coordinate in ured. separate byte for more efficient handling. Several aspects of this program are of In the overall program. the OETPT subrou- general interest. First, every measurement tine is followed by another subroutine that the user must make is preceded by an that testswhether the point just taken explicit prompt on the All's display. lies in menu area of the digitizing ta- Prompts make the program easy touseand blet, if so+ the program jumps to the op- are always good practice in interactive propriet* section, otherwise the point is computing. Second, the program canbe treated as a legitimate data point. re-enteredwithout going through the ini- tializationroutine; existingdataare thereby retained in memory unless they are GET POINT Pt NIPAO AXPT DATE: 11/2049 deliberately erased. Rosyre-entry makes the program more forgiving of errors, such 01 OEIPT 054CC9 LOA IFR as accidental escape to the monitor, and CV ORRENTT(VA) V 0090 AND $02 y it also makes it possible to print tables OFAS DEO OETPT 410LF7(VA). or re-enter scale factors at 130024 LDA PORTA SCALEAS/DIV. 44290 intermediate A900 AND WO WAIOIV: 1000 stages in themeasurements. Re-entry is 1053 CAP OFFO NVIDIV! 0021 accomplished by two means. Thefirst was 1C93 916 OEIPT *603EI mentioned above: the left edge of the di- 1CC9 LOX 000 PRMIE 5100 lOS R1 CLOCK 1A:30 11/28/79 gitizing tablet is reserved for a menu. 0503 STROO1 IDASE..... 110000 LOA IFR and each point token is tested to see if 40020NA MO946O ARO 0,42 49000AS it falls in the menu area. Themenu is 000 STA001 VOASE 40034NV availablewhenever the program is espect- DE1045 LDS PORTA 40004 AS 1020 ST*XVOATA.X ing a data point, which in practice is 00 INX DOER IPEAK 43040NA most of the time. Theother method of CPX 0$04 00136 MS 2e2915 ONE STR091 re.-entry uses the three user-defined keys 00F4 VPERK 40094NV 50220 NS on the ASH's keyboard. During the initial- itEPACK DATA VORIS 44047.20NV 1900 LDV 04 40444 MS ization routine, these keys are programmed iSNIFT 4 PITS to initiate jumps to certain sections of 0306 SNFOAT 11,4 LSO XVINITA ...... wm 30C9 RON XVOATIS1 FRAME 0031 IDO A1 WO ROO 3V5A10 42 CLOCK 21.3011/25,79 0.111.0 05A9 LSD XVDRIA43 EXPT DATE. 11/21/79 EXPT OATE. 11/20/79 INC ROR XVOATIS4 IOASE 10000 NA 740 4 DORXVOATx,0010.33 40014 AS AP OEV VOASE 44031.60 NV 041 OLDS.ONE SMFDAT 46004 MS FAOVANN TAMA19115 FRO VPNV V/205 TINE INS IFEAK 43140 NO 030 34.00 3440 0136 030 $94.00 0444 !CO 416MS 01 31.60 1140 0134 071 095.60 0440 2138 P! A. VPEAK 400595 40 NV ME 31.40 MO 004 032 007.40 0514 2230 40220 NS VAALF 4040 0 MY 444 os I0412 10.02 FRS WAY 11401 IPAS FRO VARY WOKS TIME

04 20,40 3440 0154 034 007.40 8660 2400 PIONS 3. 033 30.00 3340 0134 033 003.40 OM 2511

I043 10.02 FRO VOW TANA TAMS FRO VIINV Vi2AS TINE

027 30.00 2000 0134 037 090.00 1492 2430 In addition to solving theseaspects 03$ 20.40 $000 0134 0$2.40 1244 2230 of interfacing, it also is necessary to Wive an efficient program for data collec- 10.01 I0.01 tion. Hy assembly-language program goner- FRO VONV IPAA IPAS FROVARY VIPS TINO otos almost 2K bytes of object code, a 043 22.00 1E00 1143 043 090.40 0700 3310 substantial amount. On initial entry. var- 044 20 40 3940 010 044 093.20 0472 3409 047 32 00 MO 0134 047 a6 00 153$ 3304 ious constants and tables ore set equal to zero, after which the date of the etperi- ment and several scale factorsare re- FIRM 11. quested by prompts to the user. The main

94 84 NECC 1980

the programThus, the user can redirect nals are observed on oscilloscopes, and a the execution of the program fairly con recurrentdifficulty in the past has been veniently and without losing data already that the transient signalsaredifficult entered. for students to measure or even observe. Another aspect of general interest is Storage oscilloscopesare helpful, but that for each frame measured, the program they are expensive and they do not provide stores data in a temporary buffer until hard copg. Routine photography is slow. the last measurement forthat frame is awkward, and expensive, particularly if taken. The data are then transferred auto Polaroid film is used. Microcomputers, on mtically to the tables in memory. The the otherhand, offer the possibil.ty of point of the temporary storage is to allow digitizing signals and playing them back theuserto escape and remeasure a frame for analysis. Playback can be repetitive without contaminating the final tables if and fast for observation on an oscillo an error in measurement is suspected. Any scope. or it can be oneshot and slow for one who hes used program that doesnot writing out on a chart recorder. ln addi permit correction at the time of entry tion, in many experiments signals are re knows the frustration that accompanies corded from single nerve cells rather than being forced to continue in-a process one from groups of tens or hundreds of nerve knows to be incorrect.ln this case, by cells. ln such cases, the information of permitting a penaltyfreeescape in the interest is the time interval between in- middle of a frame. a compromise is achi dividual electrical impulses (action po eved between easeof programming and no tentials) in a nerve call. Histograms of correction routine at all; relatively few interpulse intervals and post stimulus measurements are made in one freme, so firing rates are routinely computed in re that repeating a frame is not onerous, and search laboratories: we can now make such the final tables remain accurate. techniques available to undergraduates as ln summary. this application of micro well. Finally, some neurophysiological ex computersto data collection illustrates periments (such as recording evoked poten several aspects that 1 believe are charac tials from the scalp) require signal aver teristic of good interactive programs. The aging tecause the signals are much smaller program is permanently in an EPROM and can than the background electrical noise. Once thereby be called by single keystroke. again, microcomputers can make without requiring that the user know how signalaveraging tecnniquesavailableto torun a tape or diskloading routine: undergraduates. also it cannot be accidentally written Digitizing transient signals and sig over. The program providesabundant nalaveragingboth require analogto di prompts. Data input and output arealways gital conversion, but presentlyavailable indecimal format. Theprogram can be AtoD boards that interface to a parallel reentered and redirected withouterasing port are too slow for use in a neurobiolo existing data, making it more convenient gy lab. An action potentialcan be as and more forgiving of errors. Errors of brief as 1 msac, and if its shape is to b which theuser is eware can be corrected preserved in digitization, at least 10 at the time of entry. Each of these fea samples should be taken. This requires a tures is. 1 think, important in bringing sampling interval of 100 usec. Analog Dev microcomputers intothe laboratory. The ices (P.O. Sox 2110, Norwood, MA 02062) philosophy behind this approach is to make manufacturesan integratedcircuit, Sbit the computer friendly to user who is not AtoD device that can make one conversion expertin programming. As colleague ex every 25 usec and that costs less then pressed it. "You don't expect students to $25.I have used their device in the tna build an oscilloscope in order to use one, log converter board whose block diagram is so why expect them to programthe micro shown in Figure 5. The converter requires computers that they use?" two control lines, one to start conversion and one to signal that the data are ready, APPLICATION 2: plus eight date lines. These line: Ora ANALOGTODIOITAL CONVERSION connected to Port A and to CA1 and CA2 of Our second application will eventually the AIM. PortS is connected to a beused by more students than the first digitaltoanalog converter, for playback one, but the program is still in an early of digitized data. The PortS control stye*. so 1 will describe it only briefly. lines, which are not otherwise needed, are Students in neurobiology laboratories typ used for input and output of trigget ically rbcord highlyamplified, transievit pulses. electrical signals from the nerves of ani My interactive program for A/D conver mals such as frogs or crayfish. The sig sion, like the HiPad program, will reside Science and Engineering 85

riga an EPROM end will make themicrocom puter sees like smart instrument to the user. The major inputroutineprovides continuous diiititton with scrolling display, so that theoscilloscopescreen resembles moving chart. Receipt of a trigger pulse causes digitization to cease SAW IN after' a preset interval, followed by con pmA tinuous replay of the lest three pages of data tthree pages are about the limit fo. flickerfree displeya. The user eon else CAI choose to output the digitized data slow ly, for playback to chartrecorder. CA2 Chart recordershave frequency passbands from DC to 100 Hz, at most, while the ori ginal signal has 'portent components up to about 5 kitzs thus an espnsion of the timescale by a factor of i00 is necessary to writs the data /accurately on a chart recorder. The timescale can be almond/ad easily by inserting a timing loop in the playback program. Digitization at 10,000samples per second would appearto require huge am ounts of immoral, but a comparison with the oscilloscope sweep speeds that aro used by nevrobiologisfs to display data shows that the memory requirements are not excessive. Cal -A single sweep typically displays from 10 Mee to 0.5 sec of data so that 194 bytes of memory would hold two to eighty swims, In *any cases, when longer sweeps of data are of interest, high rates of figitiza tion are not needed, and the effective ca pacity of the memory can be extended. Thus, in summery, therelatively low cost of icroceputers opens menu new pos sibilifiee for using them in Cie laborato 'tent5. ry. Theprogramming effort that is re quired can be extensive, end some of the interface hardwaremay have to be con structed, but theresults give students enormously greeter cap.bilities for na lyzing data. It is !.portentthat under graduatesinscience receive experience with such techniques, for theywiil work in world in which computerized canon.- tion and analysis of data will be routine.

96 1111=11migr4amr. 86 NECC 1980

CLASSICAL MECHANICS WITH COMPUTER ASSISTANCE

A. Douglas Davis

Department of Physics Eastern Illinois University Charleston, IL 61g20

-This paper describes the use of physical intuition and a solid under= computer and elementary numerical standing of what's going on in a given analysis in the teaching of a course situation. in classical mechanics. This is a traditional, rigorous, calculus-based Computers are used in the following mechanics course. The use of-the areas: computer allows students to solve 1. Introduction to variable forces. problems somewhat before the analytical 2. Introduction to integral calculus. solution is developed. Such prior 3. Investigation of harmonic motion. solution allows them to anticipate 4. Alternate approach. the analytical solution and greatly S. Central force orbits. aids in their understanding. Computer- generated solutions also allow the INTRODUCTION TO VARIABLE FORCES investigation of interesting problems whose analytical solution would other- In any good, solid introductory wise be beyond the scope of this physics course students will have solved course. essentially all of the basic probltis involving motion caused by a constant INTRODUCTION force. Variable forces require the use This paper describes the use of of calculus for a solution and are a computer and elementary numerical usually not covered in detail in an analysis in the teaching of a course introductory physics course - even if it in classical mechanics. This is not is nominally calculus-based. a computer-based mechanics course. Rather it is a traditional, rigorous Harmonic motion is the motion of a course in classical mechanics. Computer- body under the influence of a linear generated solutions are used as simply restoring force and can be written as one more tool to teach the real physics F X whereF is the force, X is the of the situation. position, K is the spring constant which determines the strength of the force, and The first week (three SD-minute the negative sign indicates that the lecture.) is devoted to teaching force always acts to move the body back "conversational BASIC" just enough to the origin. A classic example is a BASIC that even students with no prior body of mass M attached to a spring with exposure to computers can go to a spring constant X. But harmonic terminal and write the simple programs oscillators have more wide spread use we shall use. Thus, students can than that. A thorough understanding_ of immediately write, run and change their harmonic oscillators is useful, even own BASIC programs. The TAB function necessary, for understanding such diverse in BASIC allows even neophytes to things as automobile suspension systems, obtain graphic results very quickly. radio receivers, and ultra-violet In addition, the results of some absorption by the atmosphere. instructor-written computer programs are used. Both student-written and A major foundation of classical instructor-written programs give students mechanics states that a force acting on another tool-- another point of view, a body will cause it to accelerate. This enother handle -- to use in developing acceleartion is directly proportional to r

Science and Engineering87

the force and inversely proportional to operation, the antidifferentlation of a the mass and can be written in the form function. To stress the idea of an of a = F/m or, as is more usually done. integral as the area under a curve or as in the form of F ma. Beginning with an infinite sum, a homework problem Is this in the second week of class students assigned that asks for the sum of the write a simple iterative program to solve areas of small rectangles bounded by the for the motion of a harmonic oscillator. quadrant of a circle. The quadrant of a The essentials of the program are: circle can be broken into five, ten, 100 LET F K * X perhaps 20 or even 100 small rectangles 110 LET A = F/M whose individual areas can be calculated 120 LETVV+A* D by hand. As more and more smaller and 130 LETX=X+V* D smaller rectangles are used, their total 140 LETT=T+ D area comes closer and closer to the 160 PRINT T, X, V actual area of a quadrant of a circle, 160 GO TO 100 wrZ /4 or 0.71164r2. An integral, since it where F is the force; K, the spring is a sum of an infinite number of constant: A, the acceleration; V, the Infinitesimally small rectangles gives velocity; X, the position; T, the exactl that result. To make this point times and 0, the time increment T. um* s a ably clear, the students are Pditional details of the program allow asked to continue by breaking this D to be small for greater accuracy yet quadrant of a circle into 100, then 1000, have only a manageable amount of data and finally 10,000 tiny rectangles. This to be printed out.A more sophisticated problem would be unrealistic and futile numerical analysis routine, like the Runga- to attempt by hand. but is an easy and Kutta method, could be employed (I). interesting problem for the computer. But since the object of all this is' to understand the physics of the motion rather INVESTIGATION OF HARMONIC MOTION than extreme numerical accuracy, the F = -KX describes a simple harmonic simpler method seems preferable. While oscillator. Addition of a frictional classroom discussions are usually damping force turns this into a much more limited to writing a program in BASIC, realistic damped harmonic oscillator. this procedure is readily adaptable to a Such a damping force might represent hand-hald programmable calculator (2). mass and spring wiggling under water, or in cold molasses, or the resistance In a The TAB function in BASIC allows a radio or the shock absorbers on a car. student to get graphic output readily by Its inclusion drastically changes the changing the PRINT statement to: technique and approach necessary for an 160 PRINT TAB (40 + 30 * X): "*" analytical solution. But long before the which centers the output on column 40 students are concerned with the details when X 0 and has a scaling factor of 30 of an analytical solution, they have to PRINT an asterisk in column 70 or amply investigated the behavior or motion column 10 if X has a value of +1.0 or of this damped harmonic oscillator. The -1.0. If considerably larger or smaller only change to the earlier computer amplitudes are expected the scaling factor program is to redefine the force, includ- is changed accordingly. ing the damping force. 100 LETF=-K *X-C* V Such graphic output allows the With this change, the students can now students to see the motion -- and investigate the motion for various initial investigate its dependence upon Serious conditions and various values of the parameters -- in some detail before we damping coefficient C. Underdamping, begin the rigorous analytical solution of critical damping, and overdamping become the same peobiem. Knowing more about terms of real significance describing the behavior of a system makes finding certain particular and characteristic a mathematical solution all the easier motions of the oscillator. and more meaningful once it is obtained by more traditional means. Resonance Phenomenon .or the behavior of a driven or forced harmonic oscillator INTRODUCTION TO INTEGRAL CALCULUS has application throughout technology. The fact that an integral represents Tuning of a radio is but one example. the area under a curve is of vital impor- Long before the students hear the ominous tance in physics. Yet students sometimes phrase "inhomogeneous second-order complete the average course in integral differential equation," they will have calculus knowing an integral as simply an learned much about the characteristics of

96 88 NECC 1980

solutions to Just that. Again, one varied; the air drag coefficient is also simple change to the initial computer varied. programis all that is required. The equation defining the force now becomes CENTRAL FORCE ORBITS 100 LET F = -X*X-C*V+E*SIN (10T) Gravitational forces, electrostatic where E is the strength of an external forces, and forces on an isotropic driving force which varies sinusoidally harmonic oscillator are all examples of with angular frequency w.Resonance central forces, forces which depend only can clearly be seen and investigated by upon an object's distance from a certain changing various parameters and observing origin. Central forces occur throughout the effect those changes have on the nature. Any course in classical mechanics output. will spend a considerable amount of time and effort investigating central forces ALTERNATE APPROACH in general and the inverse-square law of Some very interesting real-world Newton's Law of Universal Gravitation in problems are difficult to solve analyt- particular. The orbits of interplanetary ically. Others can be solved analytically space probes, comets and planets all are without too much difficulty, but under- both important andinteresting. standing the full meaning of the solution may not be entirely clear to students. Again, the computer can enhance this For both situations a computer-generated aspect of classical mechanics. For solution is a very useful alternate exampleafter the analytical solutions approach. are derived and discussed, e desktop mini-computer with attached x-y plotter It is an easy matter to discuss the can be brought into class. As the class trajectory of an object thrown and then watches, orbits are drawn for various acted upon by Earth's uniform gravita- initial conditions. The conditions for tional field -- as long as friction a circular orbit become obvious, and through the air is neglected. That's a the real meaning of escape velocity is reasonable approximation for many sit- made entirely clear. uations. But it is also a rather interesting problem to include air This computer and plotter also allow resistance and then inviii1WITIU the students a glimpse at how planets trajectory. Air resistance is not would move if the universe had been neglegible if a student throws a designed differently. Orbits for a force crumpled wad of paper at a waste can that has a radial dependence of or if a naval cruiser fires a shell at a target ten kilometers away. 1 or 1 or rz.9 It isan easy matter to return to the original computer program and modify or whatever one likes can be handled just it to handle a ballistic trajectory with as readily as the 1 of the real force. linear air resistance: 100 LET Fl = -C * V1 This always sparks'students' interest and 105 LET F2 = -C * V2 N * 0 leaves them with an experience somewhat 110 LET Al Fl /N more tangible than just the discussion 115 LET A2 = F2/I1 of an equation. 120 LET VI V1 + Al 4 0 125 LET V2 - V2 + A2 * 0 CONCLUSION 130 LET X = X + V1 *0 Classical mechanics forms a very 135 LET Y a + V2 * 0 important foundation in the education of 140 LET T = T + 0 physicists and engineers. Andalmost 150 PRINT T, X, Y Invariably, it is considered difficult by 160 GO TO 100 students. Both student-written and C is the air drag coefficient, G is the instructor-written computer programs offer acceleration due to gravity, the quantities an additional tool to aid students in with a 1suffix refer to horizontal or understanding mechanics. Students find x-components, and quantities with a 2 the computer enjoyable (even those with suffix refer to vertical or y-components. little or no background who also find This program yields a data table of hor- its use challenging). They report that izontal and vertical positions which its use greatly aids them in gaining a students are asked to plot in order to see thorough understanding of the analytical the trajectory. Initial conditions of solutions, for as they see how a system muzzle velocity and firing angle are

99 Science and Engineeringgg

actually responds, the mathematics eventually derived to describe that motion becomes much more meaningful.

REFERENCES (1) Aolosics:F. Vierling. "Harmonic An Notion". Pulcolomifommision on College Poysits, 1969, Page 35.

(2) R. Eisberg, Applied Mathematical Physics enc ngtfi a cu a ors, c raw- 9

i 0 90 NECC 1980

COMPUTER. AUGMENTED VIDEO EDUCATION IN ELECTRICAL ENGINEERING AT THE U. S. NAVAL ACADEMY Michael W. Hagee Tian S. Lim Richard A. Pollak United States Naval Academy Annapolis, Maryland 21402 (301) 267-3492

ABSTRACT an important part of the learning process This paper describes a computerin the Department of Electrical Engineer- augmented video education (CAVE) project :ng and is in part a philosophy of teach- being undertaken in the Electrical Engi- ing at the Naval Academy. The somewhat neering Department at the United States slower or weaker students find this type Naval Academy. The project is designed of aid essential in any course in which to produce a series of modules involving applying basic principles to solve prob- computer graphics display and integrated lems is a primary objective and also used computer-controlled television to help as a measure of achievement. These EI engineering students (non-electrical) as sessions were found, not surprisingly, to well as non-engineering students learn be remarkably similar. They begin with a the essentials of electrical engineering. quick recap of pertinent theory followed Each module contains a quick recap of the by the solution of a basic problem apply- theory and basic sample problems, ing the principles just reviewed. As time followed by an exercise to test students' _permits, the problem solving is repeated proficiency. with variations in the hope that the student builds both,experience and con- I. INTRODUCTION fidence. In fact, the confidence of the Although there are different areas students that they can succeed by apply- in which a midshipman at the United ing the different. principles td slightly States Naval Academy can major (from different situations is of paramount nuclear physics to English literature), importance in such a course. A typical all midshipmen regardless'of major are RI sessions is also characterized by required to take a two-semester survey usually covering only one or two of the course in electrical engineering during stated learning objectives of the program. their second class (junior) year. Two Sessions covering the same material are different two-semester courses are repeated with different students. In this offered. One is the core course program sense they are inefficient and tax the for non-engineering majors and the other instructor not necessarily just for the is the engineering core course program time involved but also because even a for engineering majors. Both courses skillful and patient instructor can find cover basically the same material at it difficult to maintain enthusiasm slightly different levels. Both courses through constant repetition. are a mixture of theory and practical work and cover just about every major II. DESCRIPTION OF COMPUTER-SUPPORTED electrical engineering area. INSTRUCTION SYSTEM This unique requirement for all The Academic Computing Center at the Students to successfully complete a U. S. Naval Academy has developed a two - semester college level electrical computer-supported instructional system engineering course has presented some (CSIS). It is intended for use as a time-consuming problems. In the past, drill/practice/tutorial tool to aid USNA one-onone extra instruction (El) instructors in computer presentation of sessions have been the main tool in their course material. The same learning helping those non-technically oriented material can also be used as a test. One students. feature allows the instructor to build an El sessions involving an instructor exercise in basic building blocks called and one or two students have always been frames. A frame may be in the form of a

101 Science and Engineering 91

question, a comment, or a help sequences All extra-instruction modules are or it may be a means of letting the designed to give initially a brief (one student control branching.The frame- or two frame) review of the objective in type designator determines how the com- question. However, the heart of the puter is to analyze the student's response. modules is the drill and practice frames. if such a response is necessary.Another These frames can select problems and important feature is that frames may c'ntain circuits at random and through the use numeric and string variables and function of predictable wrong answers (PWA) guide expressions -- all under the control of the student through the necessary steps the author/instructor. This feature to a solution. The instructor/programmer allows the instructor to present the same can incorporate video anywhere in the functions (question types) to all program, from using educational television students, but each studenk. gets a dif- for the initial introduction to having ferent set of values for the variables certain segments serve as help sequences and functions (questions) used in the for the PWAs. both the graphics frame. Furthermore, the instructor can terminal and the video cassette player control such things as the number of are completely controlled by the computer, tries a student is allowed for answering the student is relieved of all coordina- a question correctly, whether the frame tion tasks but is still an active part in contains graphics, the number of times a a multi-media presentation.A control frame is to be presented with different interface was designed at the Naval variables substituted, and allowable Academy to take signals from the author's correct answers. program and physically control the video A team of eight instructors from the cassette player. There are commands such Department of Electrical Engineering is as "GO TO" and "PLAY."These commands currently working on a project using CSIS. transport the video tape under author/ The current project will incorporate instructor or student control. The those phases of aid to students into computer really acts as a coordinator. packages or modules of material available The student, not the computer, can control at any time in the Academy's audio visual the instructional flow. This control centers. results both from answers to presented The project team is producing a problems and from input to periodic series of modules of video and computer decision frames. Figure 1 shows an materials which could be used by students arrangement of the color television as a self-help, extra-instruction tech- monitor, color video cassette player, and nique. Bach module is keyed to one of the computer graphics display terminal. the recurrent stumbling blocks encoun- In what follows, we describe some of tered each year by non-engineering the computer-aided NI programs. majors in their first course in electrical engineering. These modules are A. Voltage/Current Division being designed to replace some of the Figure j2 depicts a typical question regular extra-instruction tutored/taught frame block diagram. Figure 3 reflects a by the faculty. specific question frame in the voltage A study of past exams, student division module. As the student enters evaluationst and faculty questionnaires this particular frame, the program assigns indicated that all students, from the random values from previously stored files superior to the weaker, experienced to all circuit variables.The component common stumbling blocks. Interestingly across which the voltage drop is to be the problem areas were common to. both calculated is also randomly selected. the engineering and now.engineering Based upon the values assigned and students. These areas covered such component selected, the program calculates topics as network reduction, Thevenin's the correct answer, six predictable wrong Theorem, Norton's Theorem, voltage/ answers (PWA), and one unpredictable wrong current division, RC and RL transient answer. The student then enters his analysis. answer. It is our conviction that the com- In Figure 4 the user has entered a puter is most significantly exploited value of 65, and the computer has re- as an educational tool when it is used sponded with the unpredictable wrong in conjunction with other activities. answer message.In Figure 5 the student All module design work is approached entered the correct magnitude but had the with the ide.,41 of complementing class wrong polarity which caused one of the PWA work with practical drill and practice messages to be displayed. Finally in exercises which reinforce the major Figure 6 the student has answered correct- learning objectives. ly and received an appropriati congratula- tory response.

102 92 NECC 1980

FIGURE 1

f4; r,

7.1 -4 -tt

The PMAs are .sot limited to a one- circuit as shown in Figure 7.NO is or two-line message but can be expanded asked to answer the questions and the into several supplementary frames or a picture remains on the screen until he video presentation.Upon completion of has provided an answer. If the answer is these help sequences the user is usually correct, he moves on to the next topic. given the same or similar questions to If the answer is wrong the first time, teat his understanding. he is given a second chance. If his After a successful answer, the secondanswer is still wrong, he is program checks to see whether the given three choices: (1) mare review, student has received the required number (2) take the test again, or (3) stop. of presentations. If not, a new circuit This information is given in Frame 13 as with different components and values shown in the Appendix. All he has to do would be randomly selected and presented. to make the choice is to type in REVIEW, If the number of presentation CONTINUE, or STOP. If he chooses REVIEW, criteria is satisfied, the program or CONTINUE, an appropriate picture will determines whether the user has answered appear on the screen to provide the infor- a predetermined number of questions mation he has asked for. If he chooses successfully to continue to the next to STOP, he will be asked if he wishes to objective. If not the student is sent comment on the program. He can either to another series of help sequences: type in "NO" (no comment) or make a otherwise he is directed to the next comment and then sign off. objective. All variables, circuits, com- III. SUMMARY ponents, number of presentations, and Computer-augmented video education is number of attempts are under the author's a highly interactive process that is dif- control. ficult to describe verbally. The best understanding can really only be achisoved 0. RL Transient Module by sitting at a carousel/terminal and This program consists of 28 frames. running the module.A good computer-based As an example, suppose the student has instruction system must satisfy several reached Frame 8 and is presented with the different communities: the educator-

103 Science and Engineering 93

programmer, the instructor, and most a system with a demonstrated high reli- importantly the student. ability, we believe we have developed a Student response to the first set of pedagogically sound and interesting set mcdules has been heartening, as a large of modules that will complement the lumber indicated that they found the regular classroom instruction. The modules to be helpful and enjoyable.Many student can use the system to increase students tried the programs after hearing basic electrical engineering skills that they were of help: on the night through drill and practice. The in- before each major exam approximately 258 structor can anticipate to spend less of the class worked one or more modules. time on Er sessions and therefore can W. feel it is important to develop pro- devote more time to students with serious grams that will be challenging and com- difficulties. pelling and that will encourage students This type of customised instruction not only to try other topics, but also offers the added benefit of diagnosing to recommend the programs to other students' specific problems in given students. concept areas as well as measuring vari- The twelve modules that were used ous teaching methods, since one can covered material of about 1/4 of the monitor the responses/comments of course, and approximately 758 of the students to the given material. students in the course used one or more programs. It appears that the availabili- REFERENCES ty of additional topics would have re- 1. R. Pollak and J. Schwab, "uSNA - sulted in a greater percentage of the Computer Supported Instruction students using the modules, since other Systems Sophisticated, Generative topics were suggested by both usersand and Easy to Author," Proceedings of non-users. The advisability of provid- the 16th Annual Meeting of Associa- ing a complete set of supplementary tion for Educational Data Systems - programs is being studied at this time. Higher Education. 1978. A total of six sections out of 45 were randomly selected to authenticate 2. M. B. Sousa. "Computer Augmented each module. Three of these sections Video Education." Educational were designated control groups while the Technology. FebruaiiWW7W. 46-48. remaining three were the test groups. Although data are still being analyzed, initial indications point to a high FIGURE 2 correlation between the test groups and the higher test averages. (The complete Individual data analysis will be available by the Help time of the conference.) Sequence There were two major problem areas pointed out by the students. The first, not surprisingly. was delays caused by slow computer response. The chaining process used in the display of graphics caused some fairly large delays especially when the number of users on the system was in excess of 100. This problem must be dealt with (it will be studied this summer ), since these delays may have prevented some students from using the programs and could dampen enthusiasm for any interactive computer materials. The second most often recorded comment was a pedagogical one.The N students had a strong desire to see the school solution to a problem. This was Fail true even though they may have answered (Help) the problem correctly. There was also Frame a strong preference for seeing the specific problem missed worked in the help sequence rather than seeing a, Next general review of the theory involved. Success Through the use of computer Frame graphics display terminals, well engineered and flexible software, and 94 NECC 1980

FIGURE 3 FIGURE 5 QUESTION #1FRAME 51 QUESTION #1 FRAME 51 You will be given three chances to answer You will be given three chances to answer all questions correctly.If at any time all questions correctly. If at any time you feel you would like to review the you feel you would like to review the voltage divider technique concept, type voltage divider technique concept, type "00" when a question is asked. "30" when a question is asked. (The calculator mode may be entered by (The calculator mode may be entered 1 typing "$CS ".) --Wait for the circuit.- - typing "#QA".) --Wait for the circuit --

What is the value of the voltage drop - What is the value of the voltage drop across R2? across R2? R1=1500 OHMSR2=670 OHMS R3=560 OHMS R1=1500 OHMS R2=670 OHMSR3=560 OHMS R4=600 OHMS Vs=6.3 VOLTS R4=600 OHMS Vg=6.3 VOLTS ANSWER? ANSWER? 65 I'm sorry but our answers don't agree. I think your approach is incorrect. Please TRY AGAIN or type 00 to review. ANSWER? CIRCUIT 7 4.9 You have the correct magnitudewbut the wrong polarity. Try Again.

CIRCUIT 7

FIGURE 4 QUESTION #1FRAME 51 You will be given three chances to answer all questions correctly.If at any time you feel you would like to review the voltage divider technique concept, type "00" when a question is asked. (The calculator mode may be entered by .typing "$CA".) --Wait for the circuit.- - What is the value of the voltige drop across R2? R1=1500 OHMS R2=670 OHMSR3=560 OHMS R4=600 OHMSVs=6.3 VOLTS ANSWER? 65 I'm sorry butyouranswers don't agree. I think your approach is incorrect. Please TRY AGAIN or type 00 to review.

V1/4A. R3 CIRCUIT 7

1 05 Science and Engineering 95

APPENDIX FIGURE 6 QUESTION #1FRAME 51 You have missed the problem. You will be given three chances to answer You have 3 choices as follows: all questions correctly. If at any time 1. Type REVIEW for more drill. you feel you would like to review the voltage divider technique concept, type 2. Type CONTINUE if you would like to "00" when a question is asked. try the test problem again. (The calculator mode may be entered by 3. Type STOP to stop. typing "SCA".) --Wait for the circuit.- - ANSWER? What is the value of the voltage drop across R2? M1=1500 OHMS R2=670 OHMS R3=560 OHMS R40600 OHMSVs=6.3 VOLTS ANSWER? 65 I'm sorry but our answers don't agree. I think your approach is incorrect. Please TRY AGAIN or type 00 to review.ANSWER -1.91 You have the correct magnitude but the wrong polarity. Try Again. LAST CHANCE ON THIS QUESTION. 21.9GROOVY(

it R3 CIRCUIT 7 R471.

FIGURE 7 QUESTION #5FRAME 8 TEST 1 Consider the followingcircuit.

1 2

R2 E-=

Let E*10 VOLTSR1=3 OHMS L=30 mH R2=6 OHMS The switch has been in position 1 for a long time (at least 5 time constants). What is the current through the inductor? ANSWER?

1U Structured Programming

USE OF PROGRAMMING METMCDOLOGY IN INTRODUCTORY COMPUTER SCIENCE COURSES

Elizabeth Alpert Hartnell College 156 Homestead Ave. Salinas, CA 93901 408 758-8211 x 431

INTRODUCTION readable, and understandable programs "Twenty-five years ago our job was to should be the goal of modern programmers. instruct the machine; now it is the Enormous contributions have been made machine's job to execute our program." toward reaching this goal. It is the re- E.W. Dijkstra sponsibility of computer science educators not only to lecture about methodology but In August 1979 I participated in the to incorporate the new methodologies Ninth Annual Computer Science Institute thoughout the computer science curriculum. at the University of California, Santa The purpose of this paper is to discuss Cruz. The institute was dedicated to Pro- those aspects of programming methodology gramming Methodologya one-week intro- that have particular relevance for computer ductory course followed by a two-week science education and to make some sugges- lecture series.The speakers were an tions for their incorporation early in impressive group - -many of the members of the curriculum. IPIP Working Group 2.3 plus several other prominnt computer scientists. I expected THE SCOPE OF PROGRAMMING METHODOLOGY to receive information I could use in the The major objective of programming classroom, but I did not anticipate that methodology is to increase the program- the material would lead to a total re- mers* ability to design and implement evaluation of what and how I was teaching. programs. Programming methodology The nature of computing is changing addresses problem solving techniques, rapidly. The proliferation of microproces- program reliability and adaptability, pro- sors has spread the availability of com- gram correctness and structure, guidelines puting and created an increased need for for partitioning large program tasks, and what Wirth (1) calls "programming know- software tools. The computer science how." In addition, advancements in micro- student, whether in a four- or two-year processors and LSI technology have paved program, must develop an appreciation of the way for the development of new what methodology is and how to use it. programming techniques. Economizing on memory space or eliminating an instruction or two are antiquatedprogramming goals. The design and implementation of correct,

96 Structured Programming 97

PROGRAM DESIGN ming projects. Most introductory textbooks provide completely specific problems. The Modularization students are even given the input/output ....a separation of concerns." formats. It would be a worthwhile exer- E.W. Dijkstra cise for instructors to assign practical problems with which the student has Most programming problems deal with familiarity--e.g., registration process- more than one issue. Thinking about all ing, payroll- -and require the student to the issues at once is confusing and do a complete specification of a problem. difficult to do. By separating the issues and dealing with the complexities one by ALGORITHM DESIGN AND DESCRIPTION one, a program becomes naturally parti- "Nice descriptions of neat algorithms tioned into modules. Programs written in are not gifts from heaven; they are a modular structure are easier to think formally designed." about, modify, maintain, and understand. Dijkstra and Gries How to partition a problem is not necessarily obvious, and therefore it is Once a program has been formally difficult to teach. What should be taught specified, algorithms can be designed for is that partitioning a problem is the each of the parts. Before the algorithm first step in problem solving and program is designed, however, it is necessary to design, and is a worthy goal in itself. select a notation to describe the algo- Most introductory textbooks discuss rithm. The notation must be concise, modularization, but do not include enough understandable, and formal. exercises to provide students with an opportunity to apply it. It is up to the An Introduction to Formal Descripticns instructor to develop sophisticated One of the most powerful tools for problems simple to solve algorithmic design is the notation (or after theproblemhas been partitioned. language) developed by Dijkstra (2). It The whole cycle of changing, adding, is essentially a calculus for the deriva- deleting, and exchanging modules should be tion of programs using predicate trans- made a part of all programming projects. formers for each statement. A predicate transformer is a rule for deriving from a Specification result predicate (post - condition] a predi- Specification is a statement of speci- cate corresponding to the set of states fic requirements. It is a process that is that ensure that the statements will termi- independent of composing the program and nate and establish the truth of the post - is not part of the solution. Specifica- condition. The statementlin Dijkstra's tion provides a means of communication, language include: Ale, abort, assignment, and in fact should minimize what the and the guarded construcii-IT---fi, do--- programmer needs to know. To be useful, od. A guard is a Boolean ail-reel-Ion at the specifications must be precise and well- Vied of a statement.list; a list may be structured. Parnas (2) believes that executed only if the corresponding guard specifications stated in terms of exter- is true. nally observable phenomena are preferable The statements are defined by their to specifications given as another pro- predicate transformers. The mathematical gram because they are user-oriented and semantics of a statement S and any post- understandable even by the non-programmer. condition Rare given by the weakest pre- They do not leave room for guessing about condition such that S establishes R. This the requirements nor do they suggest is denoted by wp($,R). particular implementations. There are For the assignment statement the various techniques of formal specification predicate transformer is that appear in the programming methodology literature, but regardless of the method wp( "x := E", R) = R-x:=E the common guidelines, as suggested by where R._ denotes the predicate obtained Parnas, are: ag:=E state everything that is required by substituting all'frae occurrences of state nothing that is not required x in R by some expression E. A definition of the predicate transformers for - leave no room for doubt. all the statements can be found in (3). It is essential that students learn A description of the c-erational semantics to appreciate the necessity of formal for each statement V" Lows. specifications and learn a well-structured The "skip" stay sent is executed by The "abort" statement specification methodology. Practical doing nothing. experience in writing formal specifica- signals failure.An assignment statement tions should be required of some programis denoted as "x : E" where x is a

1 u 98 NECC 1980

variable and E an expression.The semi- Since one of the two conditions must be colon denotes sequencing; the execution of true the prbgram will never abort.Note "Si; S2" will result in Sl being executed that if x = y both conditions will be before S2. The "TF" statement is denoted true, and it is indeterminate which state- as ment will be chosen; nor does it matter (3). if B+ SL Programs using the do---od iterative 1 1 construct can be developed in much the 11 B2 + SL 2 same way. Let us consider developing a program to establish a result R that states that S is the sum of the elements B -to SL n n fi in an array A of n elements where n > 0. Obviously, the establishment of the truth where the Bi denote guards and the SLi are of R requires a loop structure.What must be found is a generalization of R, i.e., statement lists. In the execution of IF any one of the SLi will be executed if the a relation P, that can be easily made true initially and that can be held true during corresponding Si is true.*The Q "bar" the iterative process. acts as a separator between otherwise un- In generalizing, the variable i replaces the constant n. ordered alternatives.The "DO" statement, Thus, P can be defined to be the relation denoted by that S is the sum of all elements in the array up to A (i) and 0 < i < n.Using do B. SL 1 this generalization, the result, R, is B+ SL established if P is true and i = n. 2 2 Since P must be true before executing 0 the loop, conditions must be found that 0 B SL establish this state.Setting i to 0 and n n od S to 0 will establish P - -the sum of no e lements is zero. During the iterative continues to executeany one of the alterna- process, computational progress must be tives whose guard is true until all of the made to ensure termination. Stepping i guards are false. towards n by increments of 1 will do that. Keeping P invariant then requires adding Some Examples of Formal Descriptions A(i) to S after incrementing i. All that Describing algorithms using this nota- remains is the choice of the guard for tion can be accomplished through some the loop. Since i = n is the condition faifly simple formal methods.Let us con- for termination, i 0 n must be the condi- sider a situation where it is desirable to tion for remaining in the loop.The establish the result as a = 7 where a is a program has now been derived: program variable. The program a := 7 will establish the truth of the result since S to 0; i := 0; the initial state is true (i.e., 7 = 7 is T, and w( "a := 7", a = 7) = T). The pro- do i 0 n + gram a := 6 will not establish the truth of a = 7 since there is no initial state i pal i + 1; S := 5 + A(i) that will guarantee it (i.e., 7 = 6 is F) and wp( "a := 6, a = 7) = F). od Let us next consider the problem of establishing, for a given x and y, a result A more robust program results when where m, a program variable, contains the the guard is weaker.A guard i 0 n is larger of x and y and either x or y when weaker than i < n. If, for whatever = y. For the result to be true when reason, n were less than zero, the loop m x, x must be > y. For the result to guard i < n would cause an immediate skip, be true whennt= y, y must be > x.The and termination of the loop with no conditions x > y, y x become the guards notification of error. for the program statements. A common solution to this same problem is the following if x2 y + m := x S im A(1), i tm 2 0y,x+m:= y do i 0 n + 1 + fi S ;NB S + A(i); i i + 1 od *The IF aborts if no guard is true. 100 Structured Programming gg

The second solution is more efficient--one whose elegant solutions can only be under- pass through the loop has been eliminated. stood when derived formally, but for the However, for n <:1 the second program is types of problems introduced at beginning not correct. The program could be modi- levels the pre- and post-condition asser- fied to the following equivalent program: tions as well as the iterative structures can be defined in a less formal manner. S A(I); i 1 Dijkstra, himself in fact, often does just that. do i 0 n What is important is that students be taught to design algorithms and be i + 1; S S + A(i) encouraged to express those designs in a notation that is independent of the pro- od gramming language used for implementation. but both will result in an infinite loop When algorithms are designed with a for n < 1. particular language in mind they are usually overly complicated and often in- Reasons for Using Formal Descriptions correct. There is no great accomplishment By this time the reader might be intraining students to think in ?OMAN thinking that the examples have not shown (or any other programming language). The anything new. That is exactly the point. real goal should be to train students in The solutions may be the same but they how to think analytically and provide have been arrived at through a totally them with tools that can assist them in different process.It is the methodology that task. that is important as a program design tool. These solutions could probably have Other Formalisms been written Without the analysis, but for At this point the reader might be more sophisticated problems the program wondering why there has been no reference solution may not be as obvious or intui- to flow charts, structured programming, tive. And even if one were to arrive or Stepwise refinement. The design intuitively at the same program, that the methodology that has been presented is development of the algorithm can be actually an historical outgrowth of these formally verified is significant. techniques. By following the methods outlined by Twenty-five years ago, when program- Dijkstra, solutions to problems can be mers were concerned with the details of devised and described in a notation that is machine execution, flow charting might quickly understood. The algorithms can be have been a useful technique. Now, easily discussed because of the simplicity however, it is antiquated as a design tool. of the language. Initial formulations of ThiVoit appropriate use of the flow chart algorithms can often be *massaged" pro» is as documentation--a description of the ducing simpler or more efficient solutions execution of the program. It is unreason- without changing the assertion about their WarariXpect students to try to design validity. algorithms using flow charts. NOreover, Additional examples of this approach they hardly ever draw the flow charts oan be found in the appendix. It is before coding a program even when suggested that the reader solve the prob- instructed to do so.They usually draw lems first and then compare the solutions flow charts after the program has been with the algorithms. coded and is running. Instead of being The greatest challenge in computer chastised for this, they should be praised. science education is teaching students how Structured programming is more an to design algorithms.The intuitive approach implementation methodology than a design May=work for some students but it certainly one. The constraints of structured pro- cannot be depended on. Illustrating the gramming for design purposes are fairly design of a few basic algorithms is primitive. The Dijkstra methodology has -usually helpful, because the students equivalent basic constructs, and stronger can transfer the techniques to related guidelines for combining them. Students problems. Po- example, different algo- should certainly be taught structured rithms for manipulating the elements in a implementation, but will find that imple- list may share many of the same menting well-designed algorithms requires properties. very little additional work. An initial reaction to the Dijkstra Stepwise refinement still plays an notation and the methods that are described important role in design, especially for slush more formally in his book is that the larger problems. Dijkstra uses refinement methods are too mathematical and beyond the in his notation when describing complex comprehension of most beginning students. problems. Hehner C41 has suggested adding That charge may be true for the underlying to the Dijkstra notation a formal method theory and the Mere sophisticated problems of refinement that eliminates the need

110 100 NECC 1980

for the do--cd structure; and may further The programmer who truly understands the simplify the notation. basiez will be more valuable than one Students usually are successful when who thinks he/she knows all the using refinement techniques. That an complexities. English phrase or descriptive term can be included in the design process(and its Documentation and Style precise definition done separately), simpl Documentation and style are ath.r. fies thinking about a problem. The prac- areas that can improve reliability. Docu- tice of writing down simple concise step* mentation should include not only a is one that instructors should follow in description of what the program is the classroom whenever a problem is being supposed to do but also information re- worked out.Too often students attempt to lated to the design. Self - documenting solve problems by thinking in the syntax of techniques such as explicit declarations the programming language when just a simple and choice of meaningful variable names description suffices as a first step. are helpful. Developing a style of programming PROGRAM IMPLEMENTATION should go right along with developing "Software is undoubtedly the major formalisms. Indentation and formatting source of unreliability in most of the program text is a relatively computer systeks today." simple matter, yet it greatly enhances the J.J. Horning reader's comprehension. Most current textbooks provide Unfortunately, the design of a correct examples of documentation and style. With- algorithm does not ensure that its imple- out constant pressure from the instructor, mentation will be reliable. Whereas the however, there is little transfer of what costs of hardware are plummeting, the the student sees in the text to the costs of sortware keep rising. Seventy programming assignmerts. percent of programming effort is spent on maintenance and much of that main- Robustness tenance is repair to unreliable or incor- The concept of robustness was Men- rect software. In view of this situation, tioned briefly in dealing with program one of the important goals of implementa- design. It is relevant as well to program tion should be reliability. This section implementation. P program is more robust of the paper presents some suggestions if it functions properly when given a taken from Horning (5) for creating more wider range of input values. The inclu- reliable programs. sion of a few precautionary measures to handle the events that aren't supposed to Language occur can contribute much to the relia- The goals of simplicity, under- bility of a program. These measures can standability, and maintainability are as be language, machine, or data directed. important for program implementation as At present there are no exception handling they are for program design. A simple, schemes that are infallible, but neverthe- elegant algos.hm is easier to implement less, it is vital that beginning program- than one that .ias not been carefully mers realize the importanca of such planned, but an effort must be made to considerations and develop an appreciation maintain the simplicity.This effect can for the need to make programs robust. be achieved by choosing programming language constructs that are simple and CONCLUSION easy to understand. Well-designed "Well it Iprograsq works, it must algorithms can be implemented using only be right." a subset of a total languages Anonymous Another adVantage of choosing a subset is that the programmer can become This paper has presented certen more familiar with the constructs and concerns of methodology in program design truly understand how to use them effec- and implementation. These concerns should tively. The use of a language structure also be the concerns of computer science that is only vaguely understood can be educators. The need for computer literacy dangerous. courses is being stressed by college and Very often in beginning programming university administrations across the language classes instructors cover as many nation. Soon every college student, of the language structures as possible. whether in a two- or four-year curriculum This is actually a disservice to the will be enrolling in a computer science students. A more successful approach would class. The question is--what and how are be to throw out all the complex structures we going to teach these people? and consider only the very basic ones. Structured Programming 101

For years, the philosophy has been 3. Dijkstra, Edsger W. A Discipline of that students need to get on the machine Prramming. Prentice-Hall, Inc. quickly. It didn't matter much what they 1976. did there. If we are teaching people how *4. Hehner, Eric C.R. do Considered od to program, it does matter.Many students A Contribution to the Programming take only one course in programming and Calculus. University of Toronto. use that experience in their future *5. !WOW J. J. Effects of Program- endeavors. With the increase in small ming Languages on Reliability. business and home computers this pattern Xerox Palo Alto Research Center. will increase. The decision must be made then whether these people should merely learnt programming language or really learn how to program. APPENDIX The methodologies described in this The following are examples of algo- paper are not difficult to understand or rithms developed by formal methods pre- teach. They can be included in all elemen- sented by David Gries during a five-day tary courses now being taught. If they are Introduction to Programming Methodology taught as part of the programming process class at the University of California, so that students initially learn to do Santa Cruz, August 1979*. things correctly, the quality of programming is more likely to improve. Example 1. Coincidence problem Teaching all the details of any pro- Given a function f with M values gramming language is of transitory value. where Languages change, machines change, new versions of software are released, and no f(0) < < f(2) < < f(14-1) two manufacturers do everything the same. But teaching methodology has lasting value. and a function g when N values where Just because a program works doesn't mean it's right.What must be conveyed to g(0) < g(1) < g(2) < < g(N1) students is that there is more to programming than just coding. They need to coast the pairs that: are the same. develop an appreciation for. programs that For example, for are simple, elegant, readable and robust- - the kind of program of which Dijkstra f = 3,5,8,12,15,18 says, "Ain't it a beauty!" g = 1,3,4,7,8, 11

Acknowledgements the count c would be 2. I would like to thank Professor William M. McKeeman and the University of Solution 1. m := 0; n vs 0; c gm' 0; California Extension for giving me the opportunity to participate in the Institute do n 0 M and m 0 M of Computer Science at OC Santa Cruz. I am grateful to Kenneth Friedenbach for hin if f(s) < g(n) + mtr.m + 1 contributions to the development of this paper. I would also like to thank f > g(n1 * n vs n + 1 Professor Fred Schneider of Cornell University for his comments on an earlier f (21) = g(u)-0 n := n + 1; version of this paper. m g= M + 1; REPERENCES C t= c + 1

*1. Wirth, Niklaus. The Module: A fi System Structuring Facility in od Higher-Level Programming Languages. Insirtut fur Informatik, Zurich, The loop invariant is that c contains a *2. Parnas, David L. The Role of count of the number of CO = g(i) where Program Specifications. University 0 < i < m, 0 < j < n. Termination is of North Carolina at Chapel Hill. ttri condition-M = M, provided f(m-l) < g(n) or n = N, provided g(n-1) < f(n).

*These were distributed at the Program- ming Methodology Lecture Series, Ninth Annual Computer Science Institute, University of California, Santa cruz, August 1979.

112 102 NECC 1980

Example 2. Different Values Given a function f where

f(1) < f(2) < f(3)< 1,

count the number of different values in f(1 :M).

Solution. 2. m 1= 1; c := 1

do m # M

if f (m) = f(m 1)1- m 1m m + 1

f(m) # f(m + 1) +c := c + 1; m 1= m + 1 fi od

The loop invariant is that c contains 1 + the number. of f (i 1) # f (i) for 1< i< a and 1< m e m.

The program may be simplified to:

aso ctat1 do m # M

m 1m m 1 if f (m) = f (m + 1) 'skip" I f (m) # f (m # 1) y c := c +1 fi od

1'3 Structured Programming 103

FORTRAN 77: Impact on Introductory Courses in Programming Using FORTRAN

Frauk L. Friedman Department of Computer and Information Sciences Room 381 Spedumnslell Temple University Philadelphia, Pennsylvania 19122 (215) 787-1912

ABSTRACT A first course in teaching problem solving and language. The feature Will normally have been structured progtamming using FORTRAN is briefly introduced first, and a brief description of Its described. The features of the new FORTRAN 77 syntactic form provided. The problem provides standardwhich in the author's view impact most additional motivation for mastering this new fea- significantly upon the course are summarised, and ture since the problem solution would be much more the effect of those features upon the structure difficult without it. and content of the course is discussed. Some of the problems shown in Fig. 1 emnhonifn skills that are fundamental to programming, such as INTRODUCTION finding the largest value inadata collection, Oa the third of April, 1978, The American searching an array for a specified item, and sort- National Standards Institute (ANSI) approved a ing. Other problems relate to a variety of appli- new American National Standard for the FORTRAN cation areas: business-oriented problems (checking programming language (ANSI 78]. This standard, account transactions and inventory control), games designated as FORTRAN 77, is a revision of the (bowling score computation and Tic -Tac -Toe), 1966 American National Standard FORTRAN. It is statistical computations, computer graphics, and expected to have a significant impact upon the text editing. use of the FORTRAN language in a wide variety of By concentrating on problems that require the applications areas. The new language should also introduction of additional features of the FORTRAN have a considerable effect upon the use of language for a reasonable solution, the motivation FORTRAN as a convenient language for teaching for these features becomes readily apparent. Once introductory programming, since it provides a the essentials of the festures needed to solve a number of significant pedagogic advantages over particular problem are introduced, class dis- its predecessor. cussion focuses on data description and algorithm The features of FORTRAN 77 that support design. It is occasionally the case that other these advantages are the subject of this paper. FORTRAN features are introduced as the algorithm A summary description, with examples, of each i developed, refined, and finally implemented. feature is described, and the effect of the The essentials of these features are described, and feature upon the structure and content of an examples of their use are given, usually within the introductory course is discussed. Before pro- context of the problem at hand. ceeding to these topics, however, an outline description of such a course is presented SUMMARY OF NEW FEATURES (see also EFK 77 and FK 78]). The new FORTRAN standard describes two levels of the languages FORTRAN (sometimes referred to COURSE STRUCTURE as Full FORTRAN), and Subset FORTRAN.Whereas the Figure 1 contains an outline of the topics FORTRAN subset was previously described in a covered in the course, a time scale for these separate standard (American National Standard Basic topics, andalist of problems that are FORTRAN, ANSI 13.10-1966), the description of associated with each topic. The course is Subset FORTRAN is now included in the description oriented around a set of two -dozen problems of the full language, and the old standard has which illustrate a variety of problem-solving been withdrawn. techniques. Most of these problems are solved The guiding criteria used in the development in their entirety, from the analysis and initial of the standard were CBRAI 78]: algorithm outline, through to the final flow- 1. the inclusion of only those new features diagram refinements and FORTRAN program. proven through actual usage Each problem is used to illustrate the 2. the inclusion of new features that en- application ofanew feature of the FORTRAN hance the portability of programs

11 4 104 NECC 1900

Week

1-2 Introduction to computers, Computation of gross and net programs and prorcam- salary for one person (the ming languages program given to students to FORTRAN and-the basic run on the computer during the computer operations first week)

3 Problem Analysis Computation of Sum and average Algorithm.development of N items and refinement Flow diagrams

4 One and two alternative Inventory control, finding decisions largest number, simulation of WHILE loops a bicycle race Compiler role in trans- relating structures

5 Data types in FORTRAN Checking account program, prime List-directed formatting number identification, Centi- DO loops grade to Fahrenheit conversion Arithmetic expressions and functions

6-7 Lists and subscripted Computation of table of fac- variables torials, finding an item in a Searching a list list, frequency distribution of Index computation exam scores. Table lookup via direct computation and search arrays

8 Block-IF decisions struc- Scoring a bowling game, drawing ture a bar graph, sorting an array Generalised DO loop Next iteration and loop exit control Nested Structures

9-10 Functions and subroutines Simple statistical package Argument lists and global Sort/merge package data Program system charts

11 Use of Format Statements Mortgage interest tables

12-13 Logical expressions General search subroutine, Character string processing finding parameters of a Extracting substringe DO loop header, text editor Replacing substrings program system

14 Multi-dimensional arrays Matrix inversion, status of a Array input and output Tic -Tac -Toe game, printing Computer art block-letter patterns, scheduling class rooms

Pit. It Course Outline and Assigned Problems

3. minimal increase in language or processor 6. production of a more precise description complexity of the language. 4. avoidance of features that conflict with the previous (1966) standard t The new standard describes programs written 5. elimination of features in the 1966 in FORTRAN 77, and not the processors (such as a standard only under clearly demonstrated circum- complier or interpreter) of these programs.The stances implementation of a standarconforminn processor 1 s Structured Programming 105

Is to be inferred (rose the standard. The standard CHARACTER*120 BUFFER is to be interpreted as specifying only the minimum CHARACTER CARD (SO), ITEM reggirementm of the language. Thus a standard -con - CHARACTER*10 FRAME, IN1TLS *1, LNAME fotaing processor for the language vast be able to handle all standard - conforming programs according The length of each characi :r variable and array is to the rules of the standard. It may, in addition, fixed by the declaration statement. In this however, have extensions for features such as bit example, BUFFER is declared as a character string manipulation or errs- processing that are not of length 120, while CARD is taken to be an 80 specified in the language.It is then the decision element array of character strings of length 1. of the user whether or not to conform to the Character constants consist of strings of standard when writing a program.Of course, characters enclosed in apostrophes.Character standard-conforming programs usually will be variables and array elements may be given values in portable to all machines supporting a standard - the same manner as other typed elements in FORTRAN: conforming compiler; non-standard programs may through the use of READ, assignment, and DATA not be as portable. statements. For example, the statements shown This section contains a list and description below all have the effect of assigning the string of the new features of Tull FORTRAN which affect IVORY JOE to the variable FNAME, as previously most significantly the introductory course just declared. dew:tibed. Those features discussed that have not been included in the subset are so designated. READ (FMTID10, UNIT" S) FRAME Users of systems not supporting the full language 10 FORMAT (A) should sake adjustments as appropriate in the curriculum changes suggested in the last section Input Card IVORY JUE of the paper. 1234567890 The relevant features of Full FORTRAN are listed in Table 1 and summarized in the remainder FRAME a 'IVORY' // of this section. The material presented is in no FRAME 'IVORY JOE' way intended as a complete description of these DATA MAME / 'IVORY JOE' / features; in fact, it barely scratches the surface. Additional examples and detail may be found in the The READ statement illustrates the use of Brainerd paper and in FORTRAN 77 introductory cifiers for format and unit numbers.Other programming texts (see [DR 78], and D0 79]). input ou;put specifiers (for end-of-file, error conditions, etc.) are also allowed (see (NO 80]). TABLE I The old form NEW FORTRAN 77 FEATURES MOST RELEVANT READ (10, 5) FRAME TO INSTRUCTION IN INTRODUCTORY PROGRAMMING is still permitted with the expected caveat that 1. The character data type the first item listed specifies the unit number and 2. List-directed formatting (not in the the second, the format number.The use of PMT and subset UNIT have the added advantage of allowing order- 3. The block-IF (if-then-else) decision independent list specifiers in an input/output structure statement. 4. Generalized form of the DO Statement The use of the A descriptor by itself in a fotnat is also new to FORTRAN.When the length of Arrays 5. the element to be transmitted is not specified in 6.Expressions: the PARAMETER Statement the format, it is taken from the declared length. and mixed -mode arithmetic The statement 7. The SAVE statement.

FNAME 'JOE' 1-22&SheraeLar_Data_Tga& 'IVORY' // ' Perhaps the most significant change in the Illustrates the use of the character string concate- standard is the addition of the character data nation operation. Substring and string comparison type, which now replaces the Hollerith type. operations are also permitted in FORTRAN 77, as It was the use of the Hollerithtype that made well as intrinsic function opurations for determin- many FORTRAN programs difficult to understand, to ing string length and for character.to -integer and check out, and to transport'from one computer to integer.to.character conversion.User-defined another having a different sine storage cell. character functions are also permitted in FORTRAN (Although the Hollerith type is no longer in- 77. (The FORTRAN 77 Subset does not support cluded in the standard, it is expected that most concantenation, substrings, or character functions). major manufacturers will continue to support this feature in their FORTRAN 77 processors). 2- List - Directed Formatting (Not available in the Some examples of the declarations of character Subset) variables and arrays are: The new FORTRAN standard permits the uae of a

1sG 11

106 NECC 1960

statement label, an integer variable that has been The logical function PRIMA shown in Fig. 2 (see assigned a label, a cbatactet exptession, or an (BRAT 78] provides a good illustration of the use asterisk for the designation of a format. Pot and advantages of the block-IF. example, statement lists i), 11), ill) and iv) ate allowed in FORTRAN 77 and produce identical LOGICAL FUNCTION PRIME results. INTEGER N, DIVISR IF (N .LE. 1) THEN I) WRITE (6, 150) 'SUM N (N + / 2 PRIME = .FALSE. 150FORMAT (A, I6) ELSE IF ('1 .!Q. 2) THEN PRIME = .TRUE. II) ASSIGN 150 TO LABEL ELSE IF (110D(N, 2) .EQ. 0) THEN WRITE (6, LABEL) 'SUM 0 1, + 1) 2 PRIME = .FALSE. 150FORMAT IA, 16) ELSE DO 10 DIVISR 0 3, INT(SQRT(REAL(N))), 2 ill) CHARACTER FORM*6 IF (MOD (N, D1VISR) .EQ. 0) THEN FORM '(A, I6)' PIER .FALSE. WRITE (6, FORM) 'SUM RETURN END1F 'Iv) WRITE (6,1(16, 16)' 'SUM = ', N * (N + 1) / 2 10 CONTINUE PRIME 0 .TRUE. An astatisk is used to specify list-directed ENDIF formatting, as determined by the input output RETURN list, the ptocessor, and the fora of the data. For END input, the type of each data item is determined by the FORTRAN system (rather than a user format description).Additional information, such as the Fisz_k An Illustration of the Block-If Structure position of the decimal point in a real number is also determined in this manner.Data items may This function shows the use of the Block-IF be separated from one another using blanks or in implementing a decision structure with four comas. alternatives, and a decision structure having one Given the card alternative (inside loop 10). A two-alternative Block-If may be implemented in the form (219-40 -0677' 'LITTLE LOOT' 801.35 The statements REAL RATE IF (condition) THEN CHARACTER SSNO*11, NAME*24 INTEGER FOURS If condition is true statement READ*, SSNO, NAME, HOURS, RATE sequence executed would produce the following result: ELSE .1 WEE= 11121111RATE If condition is false statement riODZI.V41414-Wi55-, 1.35 sequence executed ENDIF As shown, character strings read via list-directed formatting must be enclosed in apostrophes; the 4-Generalized Form of the DO 100 Statement_ four data items shown lathe card are separated The previous sample also illustrates the more from one another by one or mote blanks. general DO loop feature provided in FORTRAN 77. In list-directed output, the widths of the The form of the FORTRAN 77 DO loop header state- fields used to print the listed elements are ment is: determined by the type of the element and the DO sn loopvar = expl, exp2, exp3 processor.The width of each character string to be printed is determined by the declared length where of the string. However, real and integer'data I) expl, exp2, and exp3 represent the are printed in fixed-width fields regardless of initial, terminal, and step value expressions the magnitude of the value to be printed; the respectively field widths ate pre-determined by the processor II) expl, exp2, and exp3 may be any integer, and are normally not alterable by the programmer. real, or double precision expression having positive negative, or (except for exp3) zero values (the use 3-The Block -IF Decision Structure of expressions is not permitted in the Subset) The block-IF structure will considerably ill) Loopvar may be any integer, real, or reduce the reliance upon the GOTO and the label double precision variable in FORTRAN ptogramaIng. This structure, with iv) an optimal comma is permitted after the appropriate code indentation, can greatly enhance terminal statement label, an. the readibility of programs written in FORTRAN 77.

12 7 Sbuctured Programmkp 107

The number of times * DO loop is executed REAL X(2*SIXS) is called the trio count. The trip count specified DirtA MI), I 1, SIZE) I MAO I by the previously described loop header is computed WRITE (6, *) X(6*X-J) as VAX(iNT(v2-v1 + v3)/v3),0) Example 11) illustrates the use of the PARANITIR statement, which is used to attach a symbolic same where vl, v2, and v3 are the values of the expres- to a program constant or parameter.(The UNARM sions expl, exp2, and exp3 respectively, and INT 'totem/et is net included in the Subset).If the truncates the result of the expression argument symbolic name is not of the defeat implied type, if it is not an integer value. The following its type scat be specified before its appearance is relationships must bold between v,l, v,2 and v3: e PARAMETER statement. Expressives are permitted to the right of the equal symbol, but they map vl < v2 and v3 > 0 contain only constants or previously defined symbolic constants. or 14 > v2 and v3 < 0 Symbolic constants may be used in expressions in the same way es variables.They Imp, in addition, be used in specificities and DATA stateliest* If the value of the trip count is not positive, the in places when only constants bed been previously loop will not execute at all. It is important to allowed. Such uses are illustrated is limo 3 and note that this is contrary to the convention 4 of Example ii), abets the symbolic constant SIZE adopted by many processors based upon the previous is used where only constants were previously standard: that loops were executed at least once allowed. regsrdisss of the values of the loop parameters. Example ii) also illustrates the use of the The previous standard did not specify what was to implied -DO in be done for loops written with an initial value DATA statemeut (not allowed in the Subset) and the moralisation of the fore of sub- exceedingthe terminal value at loop entry. The scripts allowed is /OSMAN 77.Any integer exchange will not effect the execution of programs passion soy be used to specify a subscript, that were written in accordance with the previous provided the value is within the bounds specified standard, ant it way affect those that were for the corresponding dimension in the array written in violation of the standard. declaration. (The ell standard limited subscript specification to expressions of the form 5-AII!1! The Full FORTRAN language defined in the Cl * V t C2 standard provides three major changes from the previous standard: where C Cwere integer comstants, mud Ewes an 1, 2 i)arrays smyhave up to seven dimensions Integer variable). ii) the specification of a lower subscript With regard to the formation of expressions, bound is allowed the major difference between the 1966 standard and iii) subscripts in an array reference may be the new NORMAN is the latitudes of slued-mode any integer expression (see also, section 6). arithmetic. Integer, real, double precision, and Items i) and ii) above, are not permitted in complex operands map appear in arithmetic expres- the Subset. sions except that double precision sad complex The specification of the lower subscript operands may not appear in tbe'esme expression. bound is indicated through the use of a colon, as The type of an expressive is determined by In exmainlegoperand-operstor-operand triples --the REAL X (-315) type of the result of each triple is determined by the typo of the two operands involved.For ex- which defines a nine-element array X with elements ample, if X, 1, and J are integers, and 14, J-2, x(-3), X(- 2)...X(0)...X(3). If the lower bound then the result of the evelostive of the StitIMOSt is omitted, it is assumed to bs one. X - 2.3 I/J 6-oprestiooss laglaggsgLitugging_sgjAmt is 4, es determined es follows: i) divide I/J; slims I'S and Jet are both Erode Arithmetic integers, the result of this divisive is an sere are places in the nunr1ORTBAN integer, 2. in which expressions ore permitted where only 11) add the rod value 2.3 and the integer constants, variables,' or restricted forms of ex- 2; since 2.3 is real, the reedit of the division Freestone Were previously allowed. For example, is first converted to real prior to addition which expressions are now allowed in output statements then yields a result of 4.3. (see i. below), as subscripts (ii), as array ill) dimension bounds (ii), and es indexed -DO pare the red result 4.3 is assigned for the lateen wettable E, resulting in the wrist nuiSn' meters (see section 4). Sons examples are seat of the truacatosivalue 4. i) WRITE (6, 130) 'SUM ', N * (N + 1) / 2 ii)INTEGER SIZE, I, J, lC PARMISTER (SIZE 10) 108 NECC 1980

7-The SAVE Statement parameter. Other examples are illustrated in the Contrary to what a number of FORTRAN users have sample program shown at the end of the paper. become accustomed to, there is no requirement in In the view of this author, mixed-mode either the 1966 standard or the new standard to arithmetic offers no advantage to the student in an retain the values of local variables in subprograms introductory course. On the contrary, the use of from one execution to the next.Programs written mixed-mode expressions requires additional care and under the assumption that local variables were a more sophisticated understanding of expression saved were non-standard and would not execute evaluation than is desirable or even necessary correctly on all.proccssors. at this level. The automatic type con- While saving the value of local variables in version required by the mixed-mode expression subprograms used to be rather cumbersome, the provides very little programming convenience and FORTRAN 77 SAVE statement can be used to simplify has the added disadvantage of being hidden from the this process.For example, a small subprogram to user.The use of the type conversion functions increment a counter and check for overflow night such as INT (real-to-integer with truncation), AIRE appear as follows: (real-to-nearest integer), and REAL (integer-to- real) in avoiding mixed-mode. arithmetic should be SUBROUTINE EMIT encouraged. It is perhaps not too surprising that the INTEGER MAXVAL character data type, Bloc{ -IF, and list-directed PARANETER (MAXVAL st 100) formatting have had the most influence upon the INTEGER COUNT course. For, as an analysis of the course struc- DATA COUNT /0/ ture outline indicates, the major emphasis in the NAVE COUNT. course is upon problem solving, rather than the COUNT w COUNT + 1 details of the FORTRAN language. These three fea- IF (COUNT .GE. MAXVAL) THEN tures make it even easier to concentrate on PRINT*, 'COUNTER EXCEEDS MAX VALUE OF ', MAXPal, problem- solving techniques and algorithm develop- PRINT*, 'COUNTER RESET TO 1.' ment, with less emphasis upon the language COUNT 1. 1 implementation considerations. Yet each contrib- RETURN utes to this effort in a different way. ENDIF The character data type makes it possible to RETURN introduce the concept of a character string, and END the reading, printing, and comparison of strings at a very early stage of the course. For example, The SAVE statement causes the value of COUNT to be character string constants are used in list - saved between calls to the subprogram BUMPIT. directed output statements as early as week 1, in order to provide descriptive labels or headers for IMPACT OF NEW FEATURES UPON INSTRUCTION the values printed by the first program run by Only a few of the new features found in a student. By week 4, the student is reading and FORTRAN 77 have been described in the previous printing character strings using list-directed section. As indicated, the list is restricted to formatting and simple character string comparison. those features which are felt to have the greatest During week 4, data types are discussed in detail, impact upon instruction in introductory programming and a more formal presentation of the chsracter courses using FORTRAN. Yet of the features dis- type is Oren.Finally, by weeks 12 and 13, cussed, only three, the character data type, the students are writing programs requiring the use of Block -rF structure, and list-directed formatting some simple but fundamental string operations such have substantially contributed to major changes in as aubstring extraction, comparison, insertion, the structure and content of the course. Of the deletion, and replacement. Most important, all remaining features, the generalized DO loop work is now done in a totally machine-independent (introduced in week 8), implied-DO in a DATA fashion, without the previous requirement of having statement (weeks 6-7), additional flexibility in to store data of one type (character) in a memory the use of expressions (spread throughout the cell of a different type (real or integer).This course), and the additional array features (week 8) latter feature has the added advantage of providing are creatures of increased convenience having only additional compile-time checking for data type and a minor impact upon the course. The SAVE state- operator consistency. ment (weeks 9-10) is a feature that Mould be The Block-IF structure eliminates most of the understood by all students working with subprograms, need for COTOs and labels in the implementation of but it has no other impact upon the course. algorithms in FORTRAN 77. The only remaining COTO The PARAMETER etatement (introduced in week 4) needs now can easily be restricted to implementa- provides a vehicle for discussing the notion of a tion of the WHILE loop structure and loop exit and program parameter (as opposed to an in -line con- next iteration steps. The Block -IF makes it stant). This statement provides a convenient means possible for instructors and students to concen- for a Programmer to attach a WOO to &constant trate on the specification of decision steps in value (programmer parameter) that has special pro- terms of the tasks to be accomplished and the gram significance. A value representing the maxi- condition(s) of selection, without concern for the mum size of an array is one example of a program details of using labels and COTOs. The resulting

.1r9 Structured Programming 109

implementation is not only easier to write, but also to prepare and submit for computer entry.The easier to understand and far less subject to error. capabilities of the list-directed feature are then This point is perhaps but illustrated via one slowly expanded informally throughout weeks 2 possible rewrite using COTOs and labels of the through 4 until finally, in week 5, it is executable portion of the function PRIME (see Pig. formally described,and a brief overview of how it 3). works is presented.Exclusive use of list- directed formatting in all problems studied and IP (N .GT. 1) GO TO 1 assigned is then continued until week 11. PRIME .FALSE. RETURN CONCLUDING COMMINTS IP (N .NE. 2)co TO 2 Of all the new features of FORTRAN 77, list - PRIME directed formatting provides the greatest added RETURN convenience to instructors of introductory 2 IP (MOD(N, 2) .NE. 0) GO TO 3 programming courses (be they in FORTRAN, BASIC, PRIME .FALSE. PASCAL, or any other language). The block -IP and RETURN character features are not far behind in this 3 MAX, IPIX(SQRT(PLOAT(N))) respect. But list-directed formatting allows DO 10 DIVISR 3, MAX, 2 students to begin to do input and output IP (NOM, DIVISR)am.0) 00 TO10 immediately, and to continue to do even slightly PRIM .FALSE. sophisticated input /output throughout the first RETURN tea weeks of the course (including the study of 10 CONTINUE subprograms) without the added complication PRINS .TRUE. introduced through the use of formats. RETURN The following program (Pig. 4) illustrates the use of list-directed formatting, the character Fig.31 TheFunction PRIME Without the Block..IF data type, the block -IP structure, and the PARAMETER statement. Arrays are used in the pro- Even with indentation, the logical structure gram only for illustration, not because they are of this code is obfuscated considerably because of required. The program was run on a Control Data the COTOS and labels. Corporation Cyber 174 using the University of The single and double alternativeforms of Minnesota FORTRAN compiler, 1177.For the example the block-IF are introduced during week 3 of the input entries show below course, and the general form is presented during 6 week B.By this time, however, students have been Bird' 5 doing considerable programetag, using the COTO 'Mickey Mouse' 2 only for the implementation of the MILE loop 'Kermit T. Frog' 6 structure. (Even this use could have been avoided 'Ace Bandage' 95 had the FORTRAN 77 standard contained a VSILE-like 'Carmine Burma' 7 loop structure in which the repetition condition '!hiss Piggy' 9 could be specified in the structure header). Another advantage of the block-IF is increased the generated program output is' similarity in the structure of student programs. METIER OF EXAM SCORES IS 6 Students (and instructors, too) are now better able NAME RATING to understand the programs of others Aid to assist SCORE ra BIRD 5 SATISFACTORY in checking out and correcting programs that fail. NiCKST MOUSE 2 UNSATISFACTORY Finally, there is the matter of list-directed MART T. FROG 6 SATISFACTORY formatting. List - directed formatting allows the ACE BANDAGE *** INVALID SCORE ** instructor to delay considerably any discussion of 95 CAREN& BVIAta SATISFACTORY one of the most detailed, unpleasant features of 7 OUTSTANDING the FORTRAN language -- formats.While it is true MSS mat 9 that formers are an important feature of the THE NUMBER OP OUTSTANDING SCORES IS 1 FORTRAN language, and should not he overlooked, the TEE NUMBER OF SATISFACTORY SCORES IS 3 study of formats contributes little to student /RENUMBER OP UNSATISFACTORY SCORES IS 1 mastery of the fundamentals of choosing data structures and designing and Implementing THE NUMBER OF INVALID SCORES IS 1 algorithms. It is for this reason that formate have been delayed until week 11 of the course, REFERENCES after the presentation of subroutines and functions (ANSI 783 American National Standard Programming is complete. It is indeed a shame that the list - Lansuage FORTRAN, American National directed formatting feature is not included in the Standards Institute, New York, 1978. Subset. It is hoped that most Subset implementors [BRAT 783 Brainard, Walter S. et. 81., "FORTRAN will view it as top priority for support in their 77", CACI (21, 10), October 1978, pp. processors. 806-20. In the course, list-directed formatting is introduced in the first program given to students 110 NECC 1980

C PROGRAM TO PROCESS A SET OF EXAM SCORES RANGING BETWEEN 0 AND 10 C AND CLASSIFY ACCORDING TO OUTSTANDING (8-10), SATISFACTORY (4-7), C OR POOR (0-3). COMPUTE AND !RENT FREQUENCY COUNTS. C C PRANK L. PRIEDMAN 12-9-79 C INTEGER MINOUT, MINSAT, MINSCR, MAXSCR YAMMERoamouT 8, MINSAT - 4, MAWR s 0, MAXSCR 10) INTEGER MAXCNT PARAMETER (MAXCNT 100) CRARACTER*24 NAME (MAXCNT) INTEGER SCORE (MAXCNT), N, Ix, I CRARACTER*20 RATING INTEGER OUTNR, SATNR, UNSNR, ERRCNT C C READ AND VALIDATE N READ*, N IF (N .GT. 0 .AND. N .L14MAXCAT) THEN mu:RINT*, 'NUMBER OP EXAM SCORES IS ', N

' NUMBER OF EXAM SCORES IS NOT VALID, MAX IS 1, MAXCNT PRIM*, 'PROGRAM TERMINATED.'

C C DO 20 II * N READ*, NAME(II), SCORE (II) 20 CONTINUE PRINT*, 1

. nine*, NAME SCORE RATING' C C PROCESS EACH RECORD.DETERMINE AND PRINT RATING ALONG WITH C NAME AND SCORE. INCREMENT APPROPRIATE COUNTER. OUTNR * 0 SATNR * 0 UNSNR * 0 BRUNT-0 DO 40 I 1, N .GT. MAXSCR) THEN

ELSE:1P (SCORE (I) .GE. mum, THEN

' ', RATING

C PRINT COUNTS mum ' PRINT*, THE NUMBER OP OUTSTANDING SCORES IS ", OUTNR PRINT*, ' THE NUMBER OF SATISFACTORY SCORES IS ', SATNR PRIME*,9 THE NUMBER OF UNSATISFACTORY SCORES IS ', UNSNR plaw, " PRINT*, THE NUMBER OF INVALID SCORES IS ', ERRCNT C STOP END LENIts Somas' FORTRAN 77 Program

1 21 Structured Programming 111

ERN 783 Davis, Cordon B., and Thomas R. Hoffman, TORTRANI A Structured, Disciplined Style, McGrew-Rill, 1978.

CFK 773 Priedaaa, frank L. and Elliot B. Roffman, Problem Solving and Structured Program: Lag in FORTRAN, Addism4esley, 1977.

CPR 783 Friedman, Frank L. and Elliot R. Koffman, "Teaching Problem Solving and Structured Programming in FORTRAN," Computers and Education (2. 3), Perganon Press, January 1978, Pp. 235-45.

(35 793 Rune, J.N.P., and R.C. Volt, ProarennimR FORTRAN 77: A Structured Approach, -711*W1779.

CND801 Meissner, Loren P. and Elliot T. Organick, FORTRAN 77 Featuring Structure Programming, Addison - Wesley, 1980. 112 NECC 1980

USING MODEL-BASED INSTRUCTION TO TEACH

. . PASCAL

Bogdan Czejdo

Warsaw Technical Univ. Brigham Young Univ.

I. INTRODUCTION 177:1iiiiWer we introduce the concept of model-based instruction, first defining 'model' and then presenting a simplified view of the modeling process. A way of classifying models is presented, and a process for preparing a set of complementary models for use in model-based Jastruc- tion along with the characteristics of this type of instruction.The paper then concludes with the application of model- l'Iten 1.711, °mattes ofNot .16 based instruction to computer-assisted instruction. The internal models students created in II. THE ROLE OF MODELS IN TEACHING their %mon minds Allow them to understand A model is defined in Webster's Bic- and remember concepts and perform pre- tionaryw to be a*: scribed actions. The students may create these models on their own (transformation 1. copy, image Il) based on descriptions provided in text-1' 2. pattern of something to be me& books and lectures.Alternatively, the 3. archetype internal models used by the student may be 4. description or analogy used to based on external models (transformation 13) help visualize something that already created by the textbook author or cannot be directly visualized lecturer (transformation 12). S. system of postulates, data, and The creation and use of internal models inferences presented as a mathe- is a complex psychological process which matical description of an entity will be mentioned only briefly here. It or state of affairs should be recognized that the creation of . For the purposes of this paper, a model models by the student.is a difficult and is defined to be any diagram, table, pic- time-consuming task. During lecture, there . ture, or figure which helps the student is usually insufficient time for the student to understand and remember a'concept or to formulate suitable models on his own. perform an action which is a part of a The fuzzy and incomplete models that may set of teaching objectives. As shown by come to mind are soon forgotten because they research in a variety of teaching areas04, are not strongly impressed on him through models play a very important role in the reinforcement by the teacher or by his own learning process. performance. A simplified view of the modeling process Indeed, transformation Ii is difficult in learning is represented in Figure 1. for the student for a varie y of reasons. The student may be untrained in the creation of models or have a weak imagination. Because of an incomplete understanding of the full subject, the models he creates may turn out to be unsuitable. Unlearning We have chosen a subset of meaning In initial model and relearning a more which corresponds with the meaning a4Aquate model may be very difficult: of model in our paper.

123 Structured Programming 113

For all of these reasons, it is important for the teacher, and in particular the computer science teacher, to pay careful attention to the subject of models and to provide the student with useful external models. (In the rest of this paper, the term *model" will refer. to external model.) - -

III. TYPES OF MODELS A wide variety of models have been examined and evaluated for potential use in teaching PASCAL programming.These Figure 3. Simple Tree Structure models have been drawn from several textbooks(3,5) and have also been created by the authors of this paper.What constitutes Each circle represents being in a parti- the model and the object to be modeled cular state (at a particular level in this depends on the point of view of teacher and case). student. A computer program, for example, It should now be obvious how one can com- is often a model of some real-world activity bibe elementary submodels into a more com- or process. However, in the context of this plex structure. In Figures 4 and 5, a paper, the computer program will be consideredmore complete model of an interactive lan- to be the object to be modeled. guage is drawn using the level and tree One of the best ways to classify models structures of Figures 2 and 3. is by their structure. Six basic structures have been identified: - graphic Command level - array - mathematical Edit level - text - compound - parallel Insert level Most of the models we use are gra phic structures, four of which are leviilic-tfees Figure 4. Level Structure (hierarchies), networks, and domains. Levels and tree structures can be described using the example of an interactive language. On the microcomputer systems used by the authors in their introductory PASCAL course, there is an interactive command language which is interpreted by a system monitor or operating system. Each user starts at the command level. By typing an "E", the user can go to the edit level. To return to the command level from the edit level, the user can type a *QV sequence. Figure 2 shows simple level structure.

Command level 1QU Edit level

Figure 2. Simple Level Structure

Levels are represented by horizontal lines. The transitions are represented by labeled vertical lines. Figure 3 shows a simple Figure 5. Tree Structure tree structure. The root or first level is drawn at the top of the figure but, of course, the reverse option is also possible. In the

I 0 4 4.41 114 NECC 1900

tree model, each state has at most one There are, of course, a wide variety of "parent" state making it a strictly hier- other graphic structures that might be archical structure. If this restriction is useful in other fields of learning. relaxed, a network structure results.An A second type of structure is the array example of a network structure is the syn- structure. A simple one-dimensional array tax diagram in Figure 6 for a BEGIN-END IIMEcalled a list, a two-dimensional block; array is a matrix.Array structures find their greatest utility in modeling computer memories (core, disk, etc.) and in representing data.See Figure 9 for an example of an array structure which describes the name, type, and value of PASCAL variables.

PASCALVariables Figure 6. Name MS .Value Syntax Diagram for BEGINEND Block count integer 43 surname string Jones Another *Wimple of network structure is real the model of the MAK computer shown in Pi 3.14 Figure 7. alpha character G condition boolean true CPU Disk Memory Figure 9. Name/Type/Value Model

Mathematical structures constitute a thilrEYfroriodel structure consisting of a string of symbols (letters, numbers, Figure 7. mathematical operators, etc.). Figure 10 Model of Flow of Information in Illustrates one type of mathematical struc- Minicomputer flan ture, the Backus Normal Form model. It is an alternative model for the same BEGIN-END block as described by Figura 6, but with a The tree and network structures consist more mathematical flavor. of nodes [states) connected with directed lines. The final graphic structure to be discussed in this paper, the domain, con- ablock)staBEGIN END sists of a set of'areas which either over- lap or are disjoint. Figure 8 illustrates ts=l, a simple example of a domain structure that is a model for teaching the Boolean operations OR, NOT, and AND. Figure 10. Backus Normal Form for BEGIN-END Block

For completeness, we add a fourth type, the textual structure, which is nothing but text711173ifairiEgiage. An article, letter, or book all contain such structures. We introduce this type of model here, not because we want to study books as textual structures, but because we want to alio', Figure S. textual structures as small components of larger models. Domain Structure model Having introduced various types of for Teadhing Boolean Operations structures, we are now in a position to describe compound structures. A compound 12 Structured Programming 115

structure is one that is composed of several PASCAL program. Thus, a parallel structure types of substructures For example, it is consisting of one model to represent the possible to combine two types of model struc- information flow and another to represent tures (array and graphic) into one compound the human performance (operator actions) model as shown in Figure 11. seems especially appropriate and of general applicability.

l- 000e0Ao1 Neon) b) isrtrage IV. EVOLVING PARALLEL STRUCTURES no SI MUM) 8 SIRING Effective teaching of a comgix Model or 01 IOTWOO) of a set of parallel models involves a tvoCibelei Mt 101 intabol) ZOMBA Ps mum' Menet= Mit CU) ImOseldb) series of steps on lessons. The models N - mune upon which these lessons are based must be carefully selected and designed so that the student's understanding grows in a smooth and orderly way. This process suggests a fooest404 NM) MIMI) family of closely inter-related models culminating in the final parallel structure which has been selected to represent the system under study.The model chosen for a particular micro-lesson may be related el d) to the models used in previous micro- lessons by; (1) incorporating some new feature or component into a previous model. (2) synthesis of several previous models into a new model. (3) providing a more exact description of some part of a previous model. As an example of this process, consider noun U. the model of a TERAK computer in Figure 7. 01144444446, User Streuture 414 owl GeephIesi Steuctego One can introduce the concept of a work Iete Cempound MK 40. file by expanding this model as shown in Figure 12. The last, and probably the most significant, model structure to be described here CPU Disk Memory is the rallel structure. A parallel structure s similar to the compound struc- rMOnitorIc file ABC ture in that it consists of a collection of several more elementary models. However, a parallel structure differs from the compound structure, as the sub-models within IKeyboarci I workfile the parallel structure remain distinct and physically separate, related only in that they represent complementary aspects or Figure 12. TERAK Information Flow features of the same real-world object. The advantage of several parallel models Another example of model evolution can over one compound model lies in the sim- be seen from a comparison of Figures 5 and plicity that can be retained in each com- 13. Here again a new feature is added to plementary sub -model while at the same time the model to create a more extensive model. representing all the necessary features of the system. For example, the models in both Figures 5 and 7 represent certain features of the TERM microcomputer. Figure 5 illustrates the steps necessary to have the computer perform certain actions, while Figure 7 shows Ale flow of information within the system. Thus, these two models form a parallel structure. Another example of parallel structure is shown in Figures 6 and 11. Figure 6 depicts the syntax of the PASCAL language (in a sense, the steps or actions in programming) while Figure 11 represents the computer memory (information flow) associated with a

126 116 NECC 1980

Model 2

Model 1

Reality

Figure 15. Problem Solving Using Models

The problem-solving process illustrated in Figure 15 consists of 6 basic points: 1. PR-Practical problem to be solved 2. Model M2 3. MS2- specified model M4 4. Model M1 5. MS1- specified model M1 6. PP-Practical performance By specification, we mean the process of applying the model to a specific situation Figure 13. Tree Structure by selecting one state or path of the model. The choice of teaching objective for It should be noted that the models of each micro-teaching unit can now be speci- Figures 5, 13, 7, and 12 are inter-related and constitute an evolving parallel struc- fied by the triplet of transformations: ture. Figure 14 depicts such a structure T T T 1, 2' 3 where M1 and Mi' represent the models of Figures 5 and 13 and M2 and M2' correspond where each T is defined as follows: to Figures 7 and 12. Each connecting line (Ri) represents the relationship MS2 T(PRMM) between two models or the relationship of 2 1 2 a model to reality. MSM = T2 (MS2,(MSM, MM)1

Microteaching Microteaching PP= T(MS ) 3 1 Unit 1 Unit 2 1 1 These transformations represent:

T - The transformation of the Model 1 M, for a given practical problem, PR. T - The transformation of the model 2 for a given specified model, M12. T- Practical perfoimance for the 3 specified model MS1. Figure 14.

Evolving Parallel Structure Models enable us to describe teaching objectives more precisely, which is, of V. TEACHING OBJECTIVES IN MODEL -BASED course, a very important condition for the INSTRUCTION success of teaching. Another advantage of Once an evolving parallel structure of this way of designing teaching objectives models has been designed, it becomes fairly is connected with the analysis of problem straightforward to define the teaching solving, as one can concentrate on under- objectives. The objectives can be des- standing and performance more than remem- cribed in terms of the models, the rela- bering. tionships between the models, and relationships of the models to reality. V/. MODEL -BASED INSTRUCTION The purpose of each microteaching unit Having prepared specific teaching ob- would be to solve a practical problem in jectives, we are in a position to design some particular area. The process of pro- the micro-teaching lessons in detail. Two blem solving using models is shown in phases .re involved: _first we teach the Figure 15.

127 Structured Programming 117

models and the relationship (the generali- In various practical situations, it may zation processes), and second we teach how be appropriate to relax one or more of the to apply them (the specification processes). above requirements for model-based instruc- An example of this process is shown in tion. In addition, the diagram of teaching Figure 16. (Figures 15 and 16) can be modified depot:dins on the student's role in the creation and use of the model. When_the model is an -algorithmic description of a process of performance, then the process of specification of the model is impossible, and we have algorithm-based instruction rather than model-based instruction. Some models require that certain parameters be supplied by the user to complete the model. If this is the case, then the transformations 2-3 and 4-5 of Figure 15 will consist of a series of steps instead Figure 16. of just one step. Diagram of First Phase Other models require some work to con- of Microteaching Lesson struct using simpler models. If the process of construction is given by some algorithmic rules, then the diagram of teaching A description of the steps in Figure 16 will not be substantially changed.However, is as follows: if the creation of one model is described in terms of other models, then in the pro- 1. General description of a device (TERM computer). cess of problem solving we will have two phases in place of one. The first will be 2. Introducing the basic elements of the connected with model construction for the Model, M2, and relations (Keyboard, practical situation, and the second will Monitor, CPU, and disk drive). be connected with applying this model to 3. Basic elements of the model, MI (modes). some practical action. 4. Description of the first task to be performed (transfer from keyboard to CPU). VII. THE APPLICATION OF MODEL-BASED INSTRUCTION TO COMPUTER-ASSISTED 5-6. Generalization of the model, M2, so INSTRUCTION that it represents task 1 (adding in- On the basis of the principles of model- lormation flow from keyboard to CPU). based instruction, a CAI facility has been 7-8. Generalization of the model, M1, so implemented which teaches the PASCAL lan- that it represents performance con- guage on the TERAX computer. Plans are nected with task 1 (adding actions E underway to also implement this facility on and I). the TICCIT system. 9-15. Repetition of steps 4-9 but for task It appears that MEI is a very useful 2. method in CAI because: 1. Computer graphics provide an effec- Thus, the first phase is the process of tive way of displaying models and It includes the generalization of a model. their transformations.Model-based introduction of new aspects to the model languages are much easier to imple- The second related to each elementary task. ment than natural languages. phase is the process of specification and is 2. The strict definition of teaching based on model specification as shown in objectives makes it very convenient Figure 15. Again, we repeat this phase to evaluate user responses. This is several times with increasing student activ- very important in tracing each step ity until students can solve any problem in in the process of problem solving. this area. Three types of conversation blocks have Now we are ready to describe the basic been designed as shown in Figures 17-19. characteristics of model-based instruction: 1. Parallel structures 2. An evolving sequence of models 3. Teaching objectives which are based on the evolving-parallel structure 4. A method of teaching based on moving from a practical situation to a model and then back to practical performance. Figure 17. Simple Conversation Block 118 NECC 1980

Figure 17 shows a simple conversation block. In Figure 20, IB represents an informa- It consists oft tion block to be displayed on the monitor. B1 is a simple conversation block, 82 is P 4. the problem to be solved a teaching conversation block, and B3 is C 4 correction block + 4 right answer an evaluation conversation block. It is, of course, possible to repeat conversation - 4. wrong answer blocks B2 and B3 until the lesson is mastered by the student.

VIII. SUMMARY In tE717751r, we described the use of model-based instruction to teach the PASCAL language. In the classified set of models, we found and defined a parallel ri,nrs IS. Teaching CobvOtaatiest 'lock structure. We next showed how to transform this structure into a sequence of evolving models that constitute an evolving parallel structure, the basis for defining teaching In Figure 18, several simple conversation objectives. The characteristics of model- blocks are combined into a teaching sequence. based instruction were then described. In each simple conversation block, one of Finally, the application of this teaching the transformations, tl, t2, t3, performed methodology to computer-aided instruction by the student, is evaluated. was shown, using as an example the teaching of the PASCAL language.

REFERENCES

1. Webster's Seventh New Collegiate Dic- ti.Taly, 1961 ed. G 4 C Merriam Co. Z--Yakowski, Ludwik. Nauczanie Problemowe w Szkole Zawodowej. Warszawa: WSIP, 1974. 3. BowIes, Kenneth L.Microcomputer Pro- blem Solving Using PASCAL. New Yorks Figure it. svaleatioft COOVO,SSUOR Sleek Springer-Verlag, 1977. 4. Jensen, Kathleen and With, Niklaus. Pascal User Manual and Report.New Yorks Springer-Verlag, 1970. S. Schneider, G.M., Weiengart, S.W., and Figure 19, in comparison with Fiaure 18, Petit:me, D.M. An Introduction to Pro- shows the process of evaluating two students' gram:deg and Problem Solving with Pascal. solution to a problem. The students' possible New Yorks John Wiley, 1970. responses are represented by Al, A2, A3e and Niemierko, Boleslaw. Testy Osiagniec A2 is the 6. A4. Al is the correct answer. .Ilholnxch. Warszawa: WSIP, 19/37. wrong answer, but the answer is covered by (MINI -MICRO COMPUTER) PASCAL the rules of model M1. A3 is the wrong VERSION 11.0. Institute for Information answer not covered by the rules of M1. A4 Systems, La Jolla CA, 1979. is the exit path followed when a wrong answer 8. Czejdo, B., Kozinski, W., Kwiatkowski, is repeated. There are as usual correction S., Kipinski, E. "Zastosowanie Maszyn blocks, C2, C3, and C4, which differ depend- Cyfrowych do Automatyzacji Przetwarzania i ing on the response of the user. Przekazywania Informacji."Prace Naukowe The conversation blocks described above Politechniki Warszawskiej Elektryka, are, together with the information blocks, 47 (December 3977). the basic elements of a CAI micro- lesson. 9. Chanon, R.M. "A Course in Programming The diagram in Figure 20 is an example of and Practice: Toward Small Systems." this type of CAI lesson. SIGCSE BULLETIN, 1 (February 1978): 224- 228.

Mara 20. Disven Of CAISIteteleMpos

1 9 Structured Programming 119

STRUCTURED MACHINE-LANGUAGE: AN INTRODUCTION TO BOTH LOW- AND HIGH-LEVEL PROGRAMMING

David G. Hannay Department of Electrical Engineering and Computer Science Union College, Schenectady, NY 12308 (518) 370-6273

INTRODUCTION Students in introductory computer interesting perspective. courses are often frustrated by the unfamiliar processes involved in solving CAPABILITIES bewildering algebraic problems, particular- Rl -Dl is a robot that has not outgrown ly if their background in mathematics is his training wheels. It can move back and weak. But these same students can learn forth in a straight line and position it- the fundamentals of structured programming self over one of ten squares (memory cells) quickly and painlessly by directing the numbered 0 to 9: activities of a simulated robot, R1-D1. forward RI-DI is an indirect descendent of the turtle used by the LOGO group at M.I.T. to 0 1 1^11MISIMMINI119 introduce elementary school children to the computer. The immediate reward of seeing (- backward a turtle move across the floor or a robot It can be directed to move to a specific across the screen makes even non-mathe- square or to take a position relative to matical students excited about programming. the current square, moving forward or This pedagogical principle of immediate re- backward one square at a time. enforcement is as valid for college students R1-D1 can also do a limited amount of as it is for young children. arithmetic. Each of the ten cells can But this visible movement is one of store an integer number. It can also add the least significant of Rl -Dl's abilities, or subtract from an on-board accumulator. for it has the arithmetic capabilities to Decision making is accomplished by perform simple computations and the logic comparing the contents of the accumulator capabilities to illustrate the fundamentals with the contents of the current cell then of structured programming. executing a single command (or group of RI-D1 was also designed for ease of commands enclosed in parentheses) based on programming; each command consists of a the result of that comparison. For example, single letter or digit. A program con- you could have the robot execute a command sists of a free-form string of commands only if the accumulator was less than the with virtually none of the syntax rules contents of cell 15. involved in the typical programming lang- Finally, R1-D1 can be programmed to re- uages; hence a student is free to concen- peat one command (or a series of commands) trate on such programming techniques as a specified number of times. This struc- decision making and loop control. ture can be enriched by the use of the This particular robot has been used decision making commands to exit from the successfully in introductory programming loop prematurely based on certain conditions. classes by means of a simulator written For example, it could add a number to itself in PASCAL. This simulator traces Rl -Dl's seven times, or until the sum exceeded 1000, -movements and calculations on a comand- whichever comes first. by-command basis. Once students have Input and output commands do not exist been introduced to programming via the as such. When the simulator is run, you robot, the concepts of programming in a specify the initial contents of each of the higher-level language become easier to ten cells; their contents are displayed explain, and the concepts of computer automatically after each command is executed. organization can be taught from a more However, if you wish to teach about I/O

13 120 NECC 1980 instructions as such, reading from and enjoyed even by children in elementary writing to the cells can be considered as school. (While in this mode tl-e last two input and output respectively. Or, while groups of instructions, skip and program discussing computer organization, you may control commands, are not allowed, as they, treat these instructions as memory fetch by definition, involve multi-character and store. command sequences.) In direct mode, a complete command SUMMARY OF COMMANDS string is given directly to the simulator Miscellaneous Commands: which then begins execution. This allows "H" Balt. the students to use multi-character command " " Temporary Pause.(Allows entry of sequences with minimal knowledge of the immediate mode commands.) host operating system. Students communi- "N" Make a Noise. cate directly with this one program without "@" Execute commands from an indirect learning to use a text editor, compiler, or command file. other system programs. Movement Commands: Finally, as the system text editor is introduced, students can edit and save pro- "F" Move forward one square. grams for RI-DI and modify them as they see "13" Move Backward one square. the results generated. This approach not "0" Move to square O. Only helps in teaching such concepts as de- "1" Move to square T. "2" Move to square B. bugging by tracing, but also gives the "3" Move to square T. students a chance to use a subset of,the Move to square 1% system text editor on a manageable piece "4" of text since programs are typically only "5" Move to square 14 "6" Move to square B. one line long. "7" Move to square T. APPLICATION TO VARIED AGE LEVELS "8" Move to square T. "9" Move to square I First graders enjoy watching the robot move back and forth across the screen as Data Transfer Commands: they enter commands in immediate mode. "W" Write contents of accumulator High schoolers can be shown graphically the into current cell. relation of mult":)lication to addition and "R" Read contents of current cell division to sub'*.ction as illustrated by Into accumulator. the sample program in the next section. Arithmetic Commands: College freshmen gain a glimpse of both machine language and PASCAL programming "8" Zero out accumulator. concepts. "I" Increment accumulator by 1. "D" Decrement accumulator by 1. SAMPLE PROGRAM "A" Add contents of current cell A program which would take the initial TO accumulator. contents of cells 1 and 2, and put their "S" Subtract contents of current cell from accumulator. sum in cell 3, their product in cell 4, their quotient in cell 5, and their remain- Skip Commands: der in cell 6 is given below: "E" Execute next command only if accumulator is Equal to contents 1R2A3W 1RCZ2PA4W of current cell. Z5WIRC6WP(6R2LXS6W5RIW)11 "G" Execute next command only if accumulator is Greater than This program not only illustrates several contents of current cell. important programming concepts such as "L" Execute 'nett command only if loading and storing, looping, decision accumulator is Less than making, etc,, but also serves to remind contents of current cell. students that multiplication can be per- Program Control Commands: formed by successive additions, and that division can be performed by successive "P* Perform a command as many times as the current value of the subtractions. repeat count register. The first section of the program reads the value from cell 1, adds the value in "C" Load repeat count register from the accumulator. cell 2, then stores the result in cell 3. The second section of the program picks "X" Premature 'Alit from a perform up the first number and loads it into the loop. repeat count register. The accumulator is MODES OF OPERATION cleared, then the second number is repeated- The R1-Dl can be run in any one of ly added to it. 'The product is then stored three different modes: immediate, direct, in cell 4. or programmed. In immediate mode, com- The third section of the program first mands are executed as soon as the key is clears the quotient to zero and sets up struck on the keyboard. This immediate the dividend as the remainder. Then, each feedback has proved very beneficial in time through the "P" loop the divisor is eliminating some of the students' fear subtracted from the dividend, and the re- and has made the robot understood and mainder and quotient are updated.As soon

131 Structured Programming 121

as the dividend is less than the divisor, the loop exit is taken and the program halts. One of the major disadvantages of this type of programming is the obvious lack of on-line documentation. But since the programs are so short and simple, this lack of documentation has not proven to be a handicap to students.

CONCLUSION Decision making and loop control on R1-D1 follow the general outlines of the PASCAL IF..THEN and REPEAT statements respectively.The equivalent of a GO TO statement has been purposely omitted, thereby teaching students the concepts of structured programming from the very beginning. Parentheses can be used, when necessary, to treat a group of commands as a single command, just as BEGIN..END are used in PASCAL. Students in introductory computer courses learn the fundamentals of structured programming quickly and painlessly by directing the activities of the robot. A few sessions with R1-D1 at the beginning of the computer science curriculum leave students both more knowledgeable and more enthusiastic about programming than they were in the pre-robot period.

SIMULATOR PROGRAM LISTING

PROGRAM R1D1(INPUT,OUTPUT); ( DIRECT EXECUTION VERSION ) ( THE WELL STRUCTURED ROBOT ) ( BY; DAVID G. HANNAY ) VAR COMMAND { CURRENT COMMAND ) CHAR; ACC, ( ACCUMULATOR ) COUNT, f REPEAT COUNT REGISTER ) LEVEL, { DEPTH OF *PERFORM" LOOPS ) LOC, { CURRENT LOCATION OF R1D1 ) PC, ( PROGRAM COUNTER ) PTR ( TEMPORARY POINTER ) t INTEGER; CELL { MEMORY CELLS ) ARRAY(0..9) OFINTEGER; PROG ( PROGRAM STORAGE ) PACKED ARRAY[1..00) OP CHAR; DOUBLE, f SET OF PREFIX COMMANDS VALID ) f SET OF VALID COMMANDS ) t SET OF CHAR;

PROCEDURE PERFORM (FIRST, LAST, TIMES: INTEGER); FORWARD/

13 122 NECC 1980

PROCEDURE VALID ASSIGN; BEGIN DOUBLE tm [IDA,AGA,ALA,APAI, VALID := , I($ I), ipr,rci,sx,1 END; ( OF VALID ASSIGN ) PROCEDURE BLANK SCREEN (ROW: INTEGER); BEGIN GOTOXY(0,ROW), WRITE(CHR(27):',V) END; ( OF BLANK SCREEN )

PROCEDURE ERROR (WHICH: INTEGER); BEGIN GOTOXY(0,22), CASE WHICH OF 1: WRITELN( 'ILLEGAL COMMAND'); 2: WRITELN('OUT OF BOUNDS'); 3: WRITELN( 'PARENTHESES CHECK' )= END: ( OF CASE ) EXIT(PERFORM) END; ( OF ERROR ) PROCEDURE SCAN (VAR Pl, P2: INTEGER; LIMIT: INTEGER); VAR P3, P4: INTEGER; BEGIN IF PROG(PC) <> I(' THEN BEGIN P1 := 0; P2 := 0 END ELSE BEGIN PC v PC + 1; P1 := PC; WHILE PROG[PC] <> ')' DO BEGIN IF PROG[PC) = I(' THEN SCAN (P3, P4, LIMIT); PC :0 PC + 1; IF PC > LIMIT THEN ERROR(3) END; P2 := PC 1 END END; ( OF SCAN ) PROCEDURE SKIP (LAST: INTEGER); VAR Sl, S2: INTEGER; BEGIN PC : PC + 1; IF PROG[PC) IN DOUBLE THEN PC PC + SCAN (S1, S2, LAST); IF S1 <> 0 THEN PC := S2 + 1 END; ( OF SKIP ) 133 Stnotured Programming123

PROCEDURE PERFORM; VAR P1, P2, RC: INTEGER; BEGIN RC := TIMES; WHILE RC >= 1DO BEGIN GOTOXY (50 , 2) ; WRITE (' LEVEL :',LEVEL,' REPEAT COUNT: ',RC); PC := FIRST; WHILE PC <= LASTDO BEGIN COMMAND :0 PROG[PC]; GOTOXY(0,12); WRITE(PC,': ',COMMAND,"); IF (COMMAND >= '0') AND (COMMAND <= '9') THEN LOC := ORD(COMMAND) ORD('0') ELSE IF COMMAND IN VALID THEN CASE COMMAND OF

READLN; ow: BEGIN GOTOXY(0,22); WRITELN('* * * HALT * * "); EXIT (PERFORM) END; IF': BEGIN LOC := LOC + 1; END:IF LOC > 9 THEN ERROR(2)

BEGIN LOC := LOC 1; IF LOC < 0 THEN ERROR(2) END; 44'4 CELL(LOCJ := ACC;

ACC := CELL(LOCJ;

ACC := 0; 'I': ACC := ACC + 1;

ACC 1= ACC 1; ' A' ACC := ACC + CELLILOCli '5': ACC :0 ACC CELL(LOC);

IF ACC <> CELL(LOCJ THEN SKIP(LAST); sw: L': IF ACC = CELL(LOCJ THEN SKIP(LAST); BEGIN SCAN(P1,P2,LAST); PERFORM (P1 ,P 2, 1)

1 3414 124 NECC 1980

PC := P2 4 1; END;

'C' COUNT := ACC;

BEGIN PC := PC 4 1; LEVEL := LEVEL + 1; SCAN(P1,P2,LAST); IF P2 = 0 THEN

P1 := PC; P2 t= PC; PERFORM(P1,P2,COUNT); PC t= P2 END ELSE BEGIN PERFORM(P1,P2,COONT); PC t= P2 4 1 END; LEVEL := LEVEL - 1 END; 'X't BEGIN RC := 1; PC := LAST; END; END( OP CASC ) ELSE ERROR(1); WRITE(CHR(27),'J'); GOTOXY(6*L0C+10,12); WRITELN(10,ACC,'>'); FOR PTR := 0 TO 9 DO BEGIN GOTOXY(6 *PTR410,13); WRITZ('1',CELLPTR),11.) END: GOTOXY(50,2); WRITE('LEVEL: ',LEVEL,' REPEAT COUNT! ',RC); PC t= PC 4 1 END; { OF PC LOOP I RC := RC - 1 END; ( OF RC LOOP ) END; ( OF PERFORM ) BEGIN( R1 -D1 MAIN PROGRAM ) VALID ASSIGN; WRITETIPROGRAM? '); PC := 1; WHILE NOT EOLN DO BEGIN READ(COMMAND); PROG(PC] := COMMAND; PC := PC + 1 END; COUNT := 1; LEVEL :0 0; LOC t= 0; ACC t= 0; WRITELN('C E L L S'); FOR PTR := 0 TO 9DO BEGIN GOTOXY(6 *PTR +11,10); WRITE(PTR) END; PERFORM(1,80,1): END. ( OF PROGRAM 1R1D1' ) ACM Elementary and Secondary Schools Sillacommittee

ACM ELEMENTARY AND SECONDARY SCHOOLS SUBCOMMITTEE PROGRESS REPORT,

David Moursund* Dept. of Computer & Information Science University of Oregon Eugene, Oregon 97403 Phone: (603) 686-4408

ABSTRACT. which several of the taskgroup leaders Although the instructional use'of will discuss their progress. computers at the Precollese level is In June 1978 the ACM Elementary and growing by leaps and bounds, it is still Secondary Schools Subcommittee was formed in its infancy. Over the short run many by merger of a Secondary Schools Subcom- of the major problems are correctly per- mittee and a Teacher Certification Sub- ceived to be due to a lack of adequate conmittee. Since that time ES J has work- or appropriate hardware, software, and ed diligently to identify some of the courseware.But there are two other major Problems related to instructional major problems that will be the domi- use of computers at the precollege level nant long-term considerations. These and to help solve them. Dug to very lim- are the problems of securing widespread ited financial resources ES in most agreement upon the ultimate goal(s) for cases can only hone to provide some lead- instructional use gf computers and the ership and to lay a foundation which may problem of teacher training. help lead to long term solutions. The Association for Computing Ma- This same two-year time span has. chinery's Elementary and Secondary witnessed a massive influx of computer Schools Subcommittee has been working facility into the schools. Reliable data for the past two years to identify some on how much hardware has become available of the major problems of instructional has not been collected. !le know. however, use of computers and to help lay a foun- that total sales of microcomputers by dation for progress towards their solu- Apple, Commodore, Radio Shack, and a num- tion.Some two dozen taskproups have ber of other companies total in the hun- been established. This paper is the dreds of thousands. The state of Minne- introduction to a NECC/2 session in sota has long been recognized r its 11.,:oursund is Chairman of the ACM leading role in the instructiot. use of computers. Through the Minnesota Educa- Elementary and Secondary Schools Sub- tional Computing Consortium it has made He is also editor of The committee. time-shared computing available to almost Computing Teacher, a professional jour- every school in ehe states Precollege nal aimed mainly at elementary and public education system. During the past secondary school teachers interested in instructional use of computers. two years these same schools have added

125136 128 NECC 1980

approximately 1,000 microcomouters, while we are headed. The overall long-term continuing to increase their use of the goals are neither clearly understood nor time-shared system. Minnesota has a pon- widely accepted by the people who will be ulation of about 4 million, or a little involved in implementing them. less than 2% of the total US population. Over the past decade computer liter- One can get additional insight into acy for all students has emerged as the this massive growth of microcomputer use major goal in instru "tional use of com- in education by talking to a variety of puters at the nr -" ege level. Ini- teachers and by attending local and na- tially comput4 acy tended to mean tional professional meetings, ss it has computer aware It was thought by been my privilege to do. For example, many to be adeqa students could each year the northern California affil- comt to understano ..ome of the capabili- iate of the National Council of Teachers ties and limitations of computers and of Mathematics holds a meeting at Asil- thus gain some insight into how computers omar, near Monterey, California. Usual were affecting the world and their lives. attendance is about 1,500 educators. Bob There was relatively little mention of Albrecht, well-known author and leader in giving simdents substantial training in the computer education field, has attended use of computers or inter iting their this meeting for many years.According use into the curriculum. Lnitially, at to Bob the 1978 meeting included 4 modest least, it was clear that resources to do amount of computer activity, and there so were not available within the.foresee- were 10 microcomputers available for dem- able future. onstrations and hands-on use. But the future proved difficult to At the most recent 1979 meeting foresee accurately, and large-scale in- there were computer-related talks sched- tegrated circuitry became commonplace. uled in parallel with every math session. As the price of computers dropped and There were some 66 microcomputers avail- availability increased, the meaning of able for hands-on experience, and there computer literacy changed.Now the ex- was a well-organized software swap.The pression tends to mean a functional, software swap was set up by the Computer- working knowledge of computers. Uaing Educators (CUE), a northern Cali- The analogy with reading and writing fornia group that is lass than two years literacy expressed by Art Luehrmann (2) old. By the end of the Asilomar meeting is highly instructive. We can imagine CUE had well over 200 members. (1) being involved in education at the uriwn At the Asilomar meeting I talked of the invention of reading and writing with dozen of teachers who are just get- and entering into discussion as to their ting started in the instructional use of role in education.We can imagine edu- computers. The most typical question cators getting bogged down on issues such went approximately like this: "Our as the brand of pencils and paper that school has $X to spend for computers. are best or lamenting the poor ouality What should we do ?'The typical figure and small, quantity of books that are mentioned ranged from a thousand dollars available, But these turn out to be or so up to many times that amount. short-term issues, and are certainly not Two conclusions seem evident. the major long-term problem. Over a time First, many people with very little for- span of a few thousand years, and aided mal training or experience in the in- by the invention of movable iype and high- structional use of computers are now be- speed printing presses, reading and writ- coming involved; indeed, they are being ing come to be integrated into every as- asked to take major decisions that will 'pect of human intellectual activity. affect how computers are used in their What have proved to be continuing schools. Second, while the primary con- Problems in reading and writing are il- cern is still hardware, there is growing lustrated by looking at examples, such as awareness on the part cf novice users in mathematics and music. In each field that software is els° a major issue. there he been a need to develop appro- On the average, the novice computer priate symbols, notations, and vocabulary educators know little about work done by in order to represent the key ideas.Key others nor of the higher level problems ideas had to be developed and preserved. faced by the field of instructional use As this knowledge accumulated, educatora of computers. are faced with the problem of hc.4 to teach it and what students should learn. WHERE ARE WE HEADED? Now shift your attention back to the In my opinion one of the major pro- problem of inatructional tse of computers. blema of instructional use of computers If our technologically oriented aociety is there is little agreement as to where continues to prosper then we can easily

137 ACM Elemental), and Secondary Schools Subcommittee 127

imagine that eventually computers will be machines,and we are just beginning to as readily and conveniently available as see how massively difficult the software books, pencils, and paper are today. problem really is. Within that framework a single clear-cut We can better understand the soft- goal for instructional use of computers ware/courseware problem by examining one seems eiident to me.We want to integrate of the major directions of expanded com- use of computers into every aspect of hu- puter activity, computer-assisted inman intellectual activity in the same struction. The idea is for the computer manner that we have done for reading and to take over some of a teacher's func- writing.The computer is a universal tions, interacting with students to en- tool, a new aid to problem solving, and hance student learning. Material is we want all students to develop a high needed at every grade level and in every level of functional knowledge and skill discipline, and the past 20 years of ex- in its use. perience of the PLATO project, as well as This ultimate goer. will take many many other computer-assisted instruction decades, centuries, or even milieu's to pro3ects, points out the difficulty of accomplish.Progress will be made by developing good quality instructional identifying problems that face current software and cvircware. Developing good educators and working to solVe these pro- software and computer-oriented courseware blems.The ACM ES has identified about is more difficult than developing the two dozen problems and established a like more traditional materials currently in number of task groups to work on them. use. nreover, the current market is The next several sections of this paper small, the number of people involved in discuss some of these problems. developing materials is modest, and the machinery for coordinated national ef- SHOULD COMPUTERS BE USED IN EDUCATION? forts in development or distribution is There is growing support prom tea- not yet well established. chers, school administrators, school It is clear, then, that regardless board members, parents, and others for of hardware progress, software and course- the instructional use of computers in ware will continue to be major problems education. But much of this support is for many years to come. Some federal still tentative, and much of it is based funding is being made available, however. upon incorrect insights as to how com- For example, Judy Edwards (4) heads a puters will affect education. Robert three year, $200,000 per year project to Taylor of Columbia Teachers College work on educational software for micro- heads the ES J Arguments taskgroup.His computers. review of the literature led him to as- Another approach is being strongly semble a book of readings (3) on the encouraged by ES J and has also received arguments in favor of instructional use the backing of the International Council of computers that have been so well ex- for Computers in Education (5). Local or pressed by various people over the past statewide groups of computer-using educa- 15 years. Perhaps the main conclusion tors who can readily participate in soft- to draw from Taylor's work is that many ware and courseware exchange are being formed. people are busy reinventing the wheel. Such local groups bring people together Most people currently entering the com- for personal interaction and are proving puters in education field are not taking to be a highly effective mode of infor- the time to learn what ib already known mation dissemination. Those interested and thus are expending considerable ef- in starting such a groupplease contact fort redoing what has already been done. the author of this paper.

HARDWARE, SOFTWARE, AND COURSEWARE CURRICULUM CONTENT QUESTIONS PROBLEMS Computers are a general aid to pro- 3 Several of the ES taskgroups are blem solving in every academic discipline, concerned with the problems of hardware, although they currently are much more software, and courseware; as mentioned used in some areas than in other;. ES earlier, many people view them as the has taskgroups concerned with computer major issues that need to be resolved. uses in mathematics, the sciences, the The hardware and software problems are social sciences, the humanities (includ- closely related.We have a growing num- ing art and music), business, and voca- ber of options in educational b vdware, tional education. The task in each case with the quality and capabilit these is fairly similar.Eventually use of a machines continuing to improve -ut computer as an aid to knowing the subject there is a lack of software compatibil- matter and to solving problems from these ity between different manufacturer's disciplines will be routine. So far,

1 3 128 NECC 1980

however, relatively little progress has But calculators have hid essentially occurred. no impact upon the elementary school math- We find the most progress in mathe- ematics curriculumtSpeculations as to matics, Calculator use in secondary school why would fill at least an entire paper; mathematics is now common, especially in however, it seems that a major factor is the more advanced courses; strangely (in that teachers lack appropriate knowledge the author's opinion) use of calculators and skills. in the remedial math courses has been more What, then, can we expect to happen slow to gain acceptance. The use of com- with, computers in the elementary schOol? puters is taught in some math courses, One answer is computer-assisted instruc- but we seldom find a math course where tion, with the main emphasis being on the the math content has changed to reflect computer taking over some of the instruc- what computers can do or how one uses tional processes currently handled in them in problem solving. Almost always other ways.Proponents talk about tea- the computer is an add-on topic and is cher-proof materials and point out inad- used in conjunction with learning the eouacies in the current instructional pro- traditional topics in the traditional man- gram. Opponents point out the inadeoua- ner. cies of commuterized instructional mat- In very few secondary schools one can erials and discuss the wisdom of this find a significant change in the science type of change in our instructional de- or business curriculum due to commuters. livery system, As in math, computer use tends to be add- Very few elementary schools current- on in nature, to help students learn the ly attemot to teach studenta about com- traditional content. Thus it is a'very puters 'We have only modest insight as rare student who emerges from high school to what is appropriate. Again, the major with a functional knowledge of the com- drawback is teacher knowledge. puter as an aid to problem solving in a variety of disciplines. TEACHER KNOWLEDGE The problem here is immense.The Every teacher knows how to read and content and coursework for every disci- write and makes use of this knowledge in pline needs to be rethought and redone teaching. Every teacher has substantial in the light of computers and their ca- insight into the use of reading and pabilities. writing as a tool to learning the disciplines s/he teaches, and in solving the ELEMENTARY SCHOOL problems of these disciplines. The great A single taskgroup is attempting to bulk of instructional materials builds bring order out of chaos in elementary umon students' abilities to read and school education, which represents about write half of all of precollege education. Al- Contrast this with teacher knowledge though this group is not expending much of computers! It is a truly rare achool energy on the issue of calculators, that that has even one teacher who uses com- particular issue gives good insight in- puters as an everyday tool in cooing with to some of the overall difficulties the yrobleme of his/her disciplihe. Thia, faced by the field of instructional use then, is the major problem.We are.ask- of computers. ing computer illiterate teachers to help Calculators are now ouite inexpen- students to become computer literate at sive and reliable. Even at the retail a functional level. level one can buy a 4-function machine This problem is being attacked in with 4-key memory and liouid crystal many ways.Among these are teacher cer. display for under $10. Since such a tification requirements, pre-service machine will last for years, an elem- courses, in-service courses, self-directed entary school could provide its stu- study, and so on. Many teacher organiza- dents with essentially unlimited access tions recognize the problem and are in- to calculators at a cost of perhaps S2 cluding computer talks and computer ta- per student per year.The use of cal- torials in their professional meetings. culators at this level has been studied All of these things are necessary, and in many research projects and has re- all are helping. But progress seems slow ceived the backing o such organiza- relative to the magnitude of the task. tions as the National Council of Teachers The educational world bas yet to accept of Mathematics and the National Council the use of computers in precollege educa- of Supervisors of Mathematics. Many tion as a major goal, and thus to begin books of calculator materials are now to devote the resources necessary for available for use in the elementary rapid progress. school.

13 ACM Elementary and Secondary Schools Subcommittee 129

REFERENCES Lion in 1980. I. Computer-Using Educators.Founder 4. Edwards, Judy. She is Director of and president of this organization is Computer Technology Northwest Re- Dr. William (Sandy) Wagner, Mountain gional Educational Lab, 710 S.W. 2nd, Vtew High School, Mountain View, CA Portland, OR 97204, 94041. 5. International Council for Computers 2. Lenhrmann, Arthur. "Should the Com- in EducationsThis non-profit pro- puter Teach the Student, or Vice- fessional organization publishes The versa?" Proceedings of the Spring Computing Teacher, a journal for i37 Joint Conference, 1972. Reprinted ucators. Seven issues are planned in Creative Computing, vol. 2, no. for academic year 1980-81, and the 6, Nov-Dec 1976. price for US subscribers is $10. 3. Taylor, Robert.Tutor, ToolTutee: Write to ICCE, %Computing Center, The Computer in the School. leachers Eastern Opegon College, La Grande, College Press, scheduled Tor publica- OR97850,

140 130 NECC 1960

COMpUTING CONPITINCIIS POR BCSOOL MOMS

Robert P. Taylor James L. Poirot James P. Powell Towbars College North Texas University North Carolina State University Columbia University (817) 788-2521 (919) 737-2858 (212) 6784484

IN/ROPUCTI014 by teachers at the school level as belonging to On December 4-5, 1978, the Elementary and one of three sets. The Met set encompasses Secondary Schools Subcommittee (313) of the ACK those basic universal computing oompatencies Curriculum Committee met in Washington, DC to required for any school teaching, regardless of begin formally laying out curricular and teacher level or subject. The second encompasss those training guidelines for the integration ofcom- additional computing competencies needed only by puting into the elementary and secondary schools the teacher **must teach computing es a subject of the country. Building upon a lengthy initial in its own right. The third encompasses' addi- discussion and a review of earlier papers and tional computing related, subject - specific com- documents dealing with the ease problem, the sub- petencies needed by teachers of reinjects other committee outlined what it would attempt to dc. than computing. This process identified several tasks that This paper outlines the competencies in all together would constitute the overall work of the three sets. It Incorporates critical suggestions subcommittee. finally, on the basis of interest received as a result of wide circulation of two and expertise, each participant in the l$3 meeting earlier papers on the topic (1,2). We also trust was assigned to a working task group to further it will stimulate useful discussion and criti- define and carry out one of the identified tasks. cism. We hope it provides some guidance to those Among the tanks identified was one dealing wondering abet teachers should know about com- with teacher training: to define the scope and puting. unbutance of teacher training needed to integrate coeputinginto the schools. Accordingly, a COMPUTING COMMITUNCIsS Halpin BY TENCaiRs teacher !reining task group was formed. Three sets of computing competencies This paper is a product of that task group. The Wet (1.0) includes those which all It deals only with a subset of the issues and teachers must haft, regardless of their level or areas related to designing overall,computer - discipline, evenifthatdiscipline isthe literate teacher training. To appreciate its teaching of computing itself. The second set focus and accept some of its omissions, one should (2.0) includes those needed only by the teacher be aware of the following conctrsinta that task of computing as a subject. It should be noted group, placed on itself.first, it was unanimously this second set presupposes the Met. The third agreed that definitions should be in terse of ore. set (3.0) includes additional competencies for potencies to be achieved rather than in terms of teachers who use computing to support or enhance programs or courses to transmit those competan* instruction in subjects other than computing. cies. Socund,beesuse the computing coepetencies leery teacher should acquire the competencies needed by the teacher who must teach computing as listed in the first set (1.0) and the competen- a subject are more extensive than those needed by cies listed in either the second set (2.0) or the other teachers, the competencies needed only by third set (3.0). the computing teacher should be treated es a separate module. Third, though integrally related to 1.0 s Universal computing competencies needed 1w each other, the competencies needed by the teacher ell teachers are quite different from those needed by the These are computing competencies which ell teacher's teachers, the staffs of institutions school teachers should have to teach effectively actually doing the teacher training. It was in a society permeated by computes. They casts agreed, there:ors, that specifying the compete*. to either cc both Of two goals: (1) to under cies needed by the teacher's teacher would be gated computing and (C) to use computing. They separate module of work. It was also agreed that can be stated partially in terms of competencies it should only be undertaken after the competen- listed in ACM's "Curriculum '78r (3)and per cies needed by the teacher had been specified. tially in termed different competencies derived The task group saw the coerstencies needed from other sources. >t number of such other

14 ACM Elementary and Secondary Schools Subcommittee 131

sources are listed in the references at the enact systems includingcomplete student this paper. They refloat the abundance of diverse progress record keeping; and educa- work that has taken place in the past decade tional administrative systems. relating computing to education. 11.4 Wuman/Nachine Relationships: Includes artificial intelligence, robotics, 1.1 Comfttencies computer assistance in decision making In terms of these universal competencies, (e.g., medical, legal, business), _Ames teacher should: simulations, and computers in fiction. C1.1 be able to read and write simple pro- 11.5 Information on computers, in educations grams that work correctly and to under- Includes periodicals, important books; stand bow progress and subprograms fit on-line inquiry sources such as ERIC; together into systems; professional societies such asacts, C1.2 have experience using educational ARCS, NCTRI time-sharing networks, application software and documentation; networks of compuuter users; and hard- C1.3 have a working knowledge of computer ware vendor groups. terminology, particularly as it relates In 11.1 procedures or algorithm are at the to hardware; heart of computing so teachers should learn what C1.4 know by example, particularly in using they are by implementing simple examples such as: computers in education, some types of a procedure to average a class's grades, a game problems that ate and some general types to guess what number the computer is thinking of, of problemsthat are not currently or a procedure to display a large box on the amenable to computer solution; screen and then mate it shrink to disappearance. C1.5 be able to identify and use current Teachers need to be able to implement such information on computing as it relates procedures in only one language but should be to education; able to read at the same, or a greater level of C1.6 be able to discuss at the level of an difficulty in a second so that the idea of a intelligent layperson some of the-his-- language, its strength and limitations, springs tory of computing, particularly as it from personally experiencing functionaldif- relates to education; and ferences between two languages. Within the Cl.? be able to discuss mural or human-impact limitations of simple programs,teachers should be issues of computing as they relate to taughtto write well-structured code, easily societal use of computers generally and readable by others, and to document their code in educational use particularly. acceptable fashion. The ahoy* competencies should be transmitted 11.2 should be integrated throughout the within the general preparation programs for all course of study. In order to successfully use teachers by having those progress include the computing the vocabulary of the field must be topics listed below (11.1 through 11.5). For understood. What is the difference between those being trained to teach computer science, tape, a floppy disk, and a disk? Why use one over those topics will represent only a small subset of the other? These are general ideas but some whet mat be learned about computing and its uses minimal understanding of thesis necessary to use (see Section 2.0). For all other teachers, how- computers. ever (apart from the subject-specific competen- 11.3 should certainly familiarise the cies covered in Section 3.0), these topics cover teacher with several ofthe existing well- much of what must be learned about computing by developed CAI systems cited in Section, 3.0 below. the teacher who is to be minimally literate. ?etchers will not develop new ideas about what could be done in their areas unless they see the 1.2 s Topics of Study bestof what has been done; neither will they get T1.1 Programming "Topics: Includes develop - a full understanding of what can notbe sent ofsimple algorithms and their reasonably well dons by computer without such implementation in a programing lan- exposure. Tor example, acquaintance with the guage, programming style, debugging and physics system designed by Bork (4), with CCC verification, endtask-specific pro- drill and practice systems based on Suppes's work gramming for educational applications. (5), or with PLATO work(6), should serve to 11.2 COmoubtr, lerminologv; Includes soft- acquaint the teacher with the CAI issues. ware le.g., operating Systems, time Since many teachers end up in administration sharing system) hardware (e.g., CRT, and since administrative uses of computers affect tape, disk, microcomputers) and miscel- the teacher, WOO introduction to one or more laneous items (e.g., documentation, representative administrative systems should be testing, vendors) . included. A student record system would probably 11.3 Classic Applications of Cosentino in be a reasonable choice for illustrative study; it moons Includes representative deals with information familiar to the teacher experience withproblem solvingand without using financial details some might find text manipulation;simulation, drill difficult to understand. and practice,and complextutorial Teachers should also be familiar with super

11 2 132 NECC 1980

calculator modes of computer use as a classic systems of two or more programs, application. As homecomputers become more C2.2 be able to determine whether they have common, perhaps little formal work in this area written a reasonablyefficient and will be necessary. Clearly,word processing must well-organized programs be covered= every teacher does so such word pro- C2.3 understandbasic computerarchitec cessing manually that none should be left ignorant tonal of how much word processing help the computer can C2.4 understand the range ofcomputing give. topics that are suitable to be taught, With respect to T1.4, the long range and the as well as the justification for immediate implications of computing as a form of teaching these topics, artifical intelligence should be taught. The C2.5 know what educational tools can be excitement of learning to think about thinking and uniquely employed in computer science of contemplating the powers and limits of human education. intelligence are so significantly linked with com- The first three competencies are of the sort puting experience that thisaspect mustbe commonly needed by all computing professionals, studied; the opportunity is too great to pass up. and are listed in "Curriculum '711" as among those Artificial intelligence pay best convey both the to be covered by the undergraduate computer power and limitation of computing in education. science degree program. Competencies C2.4 and Acquaintance with any of several perspectives on C2.5 are not orAmsonly needed by all computing this experience is essential and can be taught professionals. They are essential only in the using such projects as the WOO work (7) or the preparation of computer science teachers. SOLO work (I). A growing body of fiction about lormal training in mathematics is computing can also contribute effectively to the considered crucial for the teachers of computer teacher's insight into the emerging world (9,10). science.While the specification of this meths - With respect to T1.5, teachers must know matical content is considered beyond the scope of where to look to keip abreast in this rapidly this report, it should be noted that this content changing field. Course work andinstruction may include topics which are frequently not part should therefore routinely call attention to and of formal mathematics programs (e.g. finite require that the trudge, use a range of sources mathematics, statistics). about computing and eddcation. For individuals who are to serve as a com- The ideas presented above represent a minimal puter resource person for their school or school set of competencies which every teacher should system,two additional competencies have been obtain. The topics Oreeented provide a framework identified. to achieve this minimal level of competency. In C2.6 Develop the ability to assist in the addition to these competencies,every teacher selection, acquisition, anduse of should also acquire the competencies listed in computers, interactive terminals, and either Section 2.0 or Section 3.0. computer services suitable to enhance instruction. 2.0 s Competencies needed forthe teacher of C2.7 Ds able to assist teachers in evalua- computing tion, selection, and/Or development of While every classroom teacher should have the appropriate instructional materials general set of computing competencies suggested that use computing facilities. above, the teacher of computing needs more. The likelihood that he or she will, in addition to 2.2 : Topics of Study teaching, beforced to function as a general These competencies should be gained through resource to faculty, administration, and students a series of courses and other vehicles developed only increases the need for more extensive compe- through joint efforts of teacher education pro- tency in computing. grams and the computer science program. We pre- Since much of the knowledge required for such sent below alist of topics that should be a teacher is similar to that required of anyone included in the program. desiring to be a computer professional, many of T2.1 Programming Topics: Includes ;advanced the computer competencies defined in the recent algorithms, programming languages, ACT4 Curriculum Committee report, "Curriculum '70" blocks and procedures, programming (3), apply to the teacher as well. This sector style, documentation debugging and therefore relies extensively upon that report. verification, elementary algorithm analysis, applications. 2.1 :Competencies T2.2 Software Croanizationt Includes cos- The core material recommended for teachers of pUter structure and machine language, computer science represents essential elementary data representation, eymbolic coding material, as well as material especially designed and assembly systems, addressing tech- for educators. Computer science teachers ehoulds niques,macros, program segmentation C2.1 be able to write and document readable, and linkage, linkers and loaders, well-structuredprograms and linked systems and utility programs.

1 el 3 ACM Elementary and Secondary Schools Subcommittee 133

T2.3 aardware Organizations includes and implementation of cosputing throughout the computer systems organisation, logic school, even when the other teachers know nothing design, data representation and of computing. transfer,digital arithmetic,digitel storage and accessing control, 3.0 Subject specific Computing competencies reliability. needed by teachers T2.4 Data Structures and illing_Processings In addition to the setof universal includes data structures, sorting and competencies needed by all school teachers, there searching, trees, file terminology, are additional level- and subject-specific sequential access, random access, file competencies which teachers should have. Any I/0. teacher will require at least one of these, but

T2.$ Computers in Society: . Computersand no one of them will be universally appropriate their effects on governments, careers, forall teachers. The definition! of those thought, law, personal behavior) competencies spring entirely from the vast and privacy and its protections information highly diverse body of experience with using security and its preservation. computing in education over the last decade. 12.6 TeachilMi_CbmPuter Sciences includes Seeress representing SOW of this work are listed (1) knowledge of learning theories as in the bibliography. The competencies can be they apply to learning about computers. stated generally, irrespective of the teacher's (2) knowledge of several appropriate eventual level of subjects the topics, though, curricular scope and sequences for a will vary considerably, depending on both. variety of program goals (e.g., literacy, careers, college preparation. per- 3.1 ComPotenotee sonal problem solving). In terms of these subject-specific 'Curriculum '70," along with avast amount of amp:tingles, the teacher should: researchin computer education,supportsthe C3.1 be able to use and evaulate the general inclusion of topics 22.1 through T2.5.Knowledge capabilities of the computer as a tool ofprogramming topics, software organisation, for pursuing various discipline.. or etc., are essential for the computer professional level-specific educational tasks; of today. C3.2 be able to use and evaluate alternative The teaching of computing is a unique cox» hardware and software systems designed paterprofession. Knowinghow to program, to function as tutors or teacher aids; however, does not, in itself, qualify a teacher C3.3 be familiar with information and quan- for teaching computer science. Materials on why titative techniques of study in the and how to teach computer topics included in T2.6 (teacher's) subject. are invaluable to the teacher of computing and These competencies should be developed by should be included within a program of study the teacher preparation program, tailored to suit training such teachers. the trainee's intended teaching leiel and Competencies C2.6 and C2.7 of the previous subject. Vs will not present an exhaustive list section are required for those serving as computer of topics corresponding to these subject- resource personnel for a school system. These specific and level - specific uosPeunotas. competencies should be gained through the computer Instead, we will present model topics for a few science program and the teacher preparation pro- selected subjects end levels. gross. The following topics will assist in developing the required competencies. 3.1.3 -Libidos of Stu& for Teachers of Early 22.7 Winced Tunics, in computer, Sciences Childhood arisary_grasea_imel-- TIC includes advanced topicsin computer TEC3.1 Computerises wimplh.gt opens - organisation, operating system, archi- gap Experience with a wide rends of tecture) .datebase systems, computer computerises but computing-related communications. activities that children van partici- 12.0 Computers, in Education: includes pate in to enhance their readiness to detailed knowledge of learning/teaching understand and work with cosputers. research as it has implications for 21C3.2 games and Simulations: Experience affective design of institutional com- with a wide range aflame end simula- puting systems and administrative uses tions that stimulate children to of computing in our educational explore and better understand funda- setting. mental concepts andstrategies of including study of computers in education learning. will increase the teacher's ability to as:Me a T1C3.3 Tutorial patens: Soperience with role ofleadershipin providing direction toe simple and complex tutorial systems school system in integrating computing into its focusing upon mathematics, spelling, curriculum. This additional computer background reading, and other elementory topics, should allow the computing teacher to act as a including bi- lingual variations of resource person to assist in fostering development such systems.

1 4 134 NECC 1980

=3.4 exploratory, programming systems: used to create and manipulate texts in Experience with well-developed explor- a foreign language. atory systems where child-appropriate Under TFL3.1, simulations based on relevant EA) subsystems such as robots are pro- activities andsituations in the language- grasmatioally manipulated by the child culture can provide insight into the languags in a discovery or problem- solving difficult to obtain in other ways. these, and approach. many popular computer games, should also be used Less work has taken place in the area covered provide a more informal language practice for by 14E3.1 than one might expect. Despite the learners.This practice can take two forms: (1) likely wide-spread availability of microproces- translatinq and (2) informally using the sors in the immediate future, computer less com- language. Teachers should practice translating puting activities can still be very useful, Sy all user text, of appropriate games and simula- contrast with heavily machine-dependent activi- tion, into the target language, thus preparing to ties, they provide a more contemplative, less have their students do such translation. involved opportunity to examine some of the funds - Teachers should also have wide experience with mental ideas connected with computing.They thus playing games whoa, text is entirely in the tar- allow those using them to deepen their under- get foreign language and which expect all player standing even if titet have access to computers. responses in that language. Such playing in a Typical examples may be found in some of Papert's foreignlanguage can be a valuable informal work (ll) and in Taylor (12). enhancement. Experience with a wide range of Vast quantities of games and simulations are such games and slwulaticns may else suggest new, available for TOC3.2, but careful choices should more appropriate ones which the teacher can be made in selecting them. Many are not well - Create or haveothers,including students, written, either from a programming point of view create. Some attempt to create such new material or from a child-user point of view, and no game or (or new variations of old material)should be simulation should be selected unless it succeeds part of every foreign language teacher's in both areas. Some of the best work in this area training. Naturally, where audio is available, at this age level came out of the People's Com- it should be appropriately used. puter Company under the initial stimulus of There are many examples of language drill Albrecht (13). and practice suitable for use under TFL3.2. Work With respect to TEC3.3, though,many tutorial such as that done by Suppes (17) at Stanford systems have been developed, not all of them are should certainly be familiar -to language good. The experience of the teacher should teachers, though alternatives certainly exist certainly include at least one good system and (18). Work in this area should rely es heavily some discussion of what lies behind it.The work upon audio and graphics as possible, thus cutting of the CCC group under Suppes is certainly a down on the automatic tendency to always trans- worthy example in this category (14). late from the native language and to develop Finally, new exploratory systems relevant competence only inreading' and writing the for 11C3.4 are smearing, but the pioneering work foreign language. is still only for illustration. In particular, TFL3.3 should ensure that teachers use a the =MCC work which produced SHALLTALe (15) and suitable computer text editor to manipulate text the LOGO work, particularly as Seymour Papert has in the target foreign language. With appropriate advertised and sustsined it (16), is outstanding. accent mark capabilities, sucheditors can Microprocessors can be the primary machine encourage language learners to practice much more used, but not so much that the trainee is left prose writing and thus enhance their overall com- ignorant of the advantages of larger systems. mend of the language.

3.2.2 : Tbeics of Study for Teachers of Foreign 3.2.2 :'lbpics of Study for Teachers of Physical Language --SW. Science -- TPS TP53.1 Genes andSimulations: Experience TPS3.1 Exploratory programming Systems: with a wide range of games and simula- Experience with well-defined explorations designed to provide cultural tory systems through structured, background and informal language discipline-appropriate languages. learning, using the culture as context Systems meetinclude graphic capa- and the language as the medium of com- bilities, be programmahly control- munication in the gems or simulotion. lable by the student, and be oriented 4114.2 Tutorial Systems: Experiencewith to discovery through problem-solving tutorial systems designed to enhance activities relevant to the physical the learning of a foreign language sciences. through a carefully arranged body of TPS3.2 Tutorial Systems: Experience with interactive experiences driven by com- tutorial systems designed to enhance petency-based, computer-administered learningof thephysical sciences testing. through a carefully arranged body of TP53.3 perste:: language, Text Editing: Expe- interactive experiences driven by rience with a powerful text editor

1'45 ACM Elementary and Secondary Schools Subcommittee135

competency- based, computer-adminis- List of Contributors tered testing. Vivian Coon, University of Missouri -Rolla 7253.3 Wass and Insulations Brperience with Richard Dinnis, University of Illinois etand-om gams and simulations Dan Isaacson, University of Oregon designed to enhance understanding of David Musson., University of Oregon specific physical phenomena or signi- John W. Hamblen, University of Missouri-Rolla ficant,past experinents. Atone R. Kamorewskii Bremen HighSchool, 7293.4 Clammog/Laborabery Midlothian, Illinois experience with automated sanagesent James Lockard, Buena Vista College of people, learning, time, and Dick Ricketts, Multnomah County Education resources including automated inven- Service District, Portland, Origin torying, laboratory Stan Troitman, Wheelock College infornatim/referance system; in general, uses of thecosputer to Cladd References provide the science teacher with more 1) Poriot, James, James Powell, John Romblon, time to west with individuals. *Mart Taylor. "Computing Competencies for 1123.5 Date Collection and Analysis: expe- School Teachers -- A Preliminary Projection rience with :systems which collect and for the Teacher of Computing." National analyse date including on-line Educational computing Conference gathering and analysis of experimental Proceedings, Iowa City, 1979. data and process control systems which 2) Taylor, R.P., John Hamblin, James L. collect, analyse and provide feedback Poriot, and James D. Powell. * Computing to the system. Competencies for Teachers -- A Preliminary Tapepossibilities under all three topics for Projection for All but the Teacher of Physical science teechtcs have been extensively Computing." National educational Cosoutift explored already by Bork (19)and others (20). Conference Proceedings, Iowa City, 1979. their careful work and well-documented analysis 3) Curriculum Committee on Computer Science should be extendedand incorporated in the (C3$).°Curriculmi701 Recommendations for trailing of physicalscience teachers. Such the Undergraduate Program in Computer training should include experience with relevant Science." (Preliginary Draft, dated Autumn examples which illustrate the three topics; it 1976). should also require each trainee to construct 4) Mork, Alfred. "Learning to Teach Via selected, similar, small modules ofcomputer- Teaching theCOmputer to leech.' Journal of supported instruction as a normal part of teacher Comoutereased Instruction, November 1275, training. vol 2. 5) Sums, Patrick. "Computer- Assisted =MAW Instruction at Stanford" in Man and This paper has addressed the computing compe- Computer, larger, 1970. tencies needed by pre-college teachers. These 6) Smith,' Stanley and Bruce Arne Sherood. competencise are listed in three groupings based "Educational Uses of the PLATO CmpUter on the involvement of the teacher in computer- System." SCINC4* April 1976 , vol 192. related activities. Because of the variation of 7) Papert, Seymour. "Teaching Children subject matter, implementation magpies have been Thinking," Logo Memo 2, NUT AI LAO, October given for teachers of- grades 1-4 (early 1971. childhood), for teachers of foreign languages and S) Dwyer,Thames. °The Act of Ideation* for teachers of physical science. These examples Blueprint for a Renaissance' Creative, are not meant to be In inclusive but to indicate Computing, Sept/Oct 1976. the level of necessary background knowledge. 9) MOsehmits, Abbe (editor). Inside Before graduating from a teacher training Information: Computers is fiction. program, all teachersshould berequired to Reading, Mass.*Addison4Iesley, 1976. acquire the first sat of competencies and either, 10) Taylor, lobes* P. (editor). Tam of -the. the second or third set. This requirement will Marvelous pocking: Thirty -Five Stories of prepare each teacher to use computing inthe Commuting. lbrrimtowa, 1447.: Creative classroom. Computing Press, 1980. No aotempt has been made to package the com- 11) Papert, Seymour. "Teaching Children to Be petencies into specific courses.It is felt that athematic:Ian* VS. Teaching About each environment will possibly require a different leatbesetica.0 Logout,* 4, MIS AX LAS, July technique for the introduction of the material. 1971. Another group of individuals that need to be 12) Taylor, H. P. 0Computerlase Competing for considered are the currant in-service teachers. Young Children or What to Do till the The competencies described above are as important Coyotes Cones" in Proceeding of the lid for them as for our future teachers. In-eervice World Conference on Omuta' in Education. courses must be developed to provide the indicated Morth Melland/Amsterdam, 1975. background for bi..8011,100 teachers.

116 136 NECC 1980

13) IOC (collective authorship). What to Do 28) IF/P Technical Committee for Education, MK You It t Return. Menlo Park, Calif.: Working Group on Secondary School People's Computes Company, 1975. Education:Computer Education for Teachers 14) Stipp's, Patrick and Elisabeth decker."Seel- in Secondary Schools: Aims and Objectives Mitten Studies of CCC Elementety School Cur- in Teacher Training AFIPS, Montvale, NJ, ti:Mums 1971-1975." Computes Curriculum 1972. Corporation, Palo Alto, 1976. 29) Zeuhrmann, Arthur. "Reading, Writing, 1$) fey, Alan. "A Personal *mates far Children Arithmetic,and Computing" in ;spraying of All Ages" in itOriesdines of KR National Instructional Productivity in Higher Educa- Conference, Soften, August 1972. tion. Educational Technology Publications, 16) Papert, Seymour. VersonalCompeting and the 1975. Impact on Education." Edited transcription 30) minas, Stephen. "Computer Literacy: The of a talk delivered at the Gerold P. *leg Next Great Crisis in American Education.". Nemeelel Confatencs, as printed in the pro- Oregon Computing,' Teacher, vol 6, no 1, Sept ceeding', Camputinii in College, and Oftivet- 1971. Iljr 197$ andmugaUniversity of Bova, 31) Noussund, David. "Report of the ACK Teacher 1978. Certification Subcommittee." SIGCSE Bulle- 17) E4F'ss, Petrick, Sober! Salt h, Marian Beard. tin, vol 9, no 1, Dec 1977. PP-8-18. "University-Level Competes - Assisted Instruc- 32) Poirot, James L. "A Course Description for tion at Stanford: 1975.° Instructional TeacherEducation in Computer Science." Science, 6 (1977), Elsevier Scientific sIGCsn Bulletin, vol. 8, February 1976, pp. Publishing Company, Amsterdam. 39-48. 1E) 1Tnabea. at al. °CAI in Language Education" 33) Poirot, J.L. and Groves, D.P. Beginning in Proceedings of 2nd World Conference In Computer, Science. Menchaca, Texas: Computers in Education. North Sterling Swift Publishing, 1978. flollandalsevier, 1975. 31) Poirot, J.L. and Groves, D.N. compute 19) Bork, Alfred. "Preparing Student-Computes Science for the Teacher. ,amcjaca. Texas" Dialogs Advice to Teachers."PCDP, U. C. Sterling Swift Publishing, 1976. Irvine, July 1976. 3$) Recommendations Regarding Computerk ugh 20) Dwyer, thongs A. "Some Principles for the School Education. Conference Board of the Eisen (Meg Computers in Education."Inter- Mathematical Sciences, Washington, O. C., national Journal of Nen-Machine Studies 3, April 1972. Jely 1971. 36) Taylor, 'Robert P. "Graduate Remedial Training in Computing for Educators." afts,References SIMMS Training Symposium Proceedings, 21) AiiiniN7,111111am F. 'Computes Science Pre- Dayton, February 1979. pimation forSecondary School Teachers." 37) Taylor, Robert P. (editor).The Computes in 11pulletin, S (1973). 1'45-47. the School: Tutor, 221, Tutee. New Fork: 22) "Computere and the roaming Society." Report Teachers College Press, 1980. Of hearings before USEROctober 1977, US 38) Technology in Science Education: The Next Govt Printing Office, 1.94B. Ten Yeats. National science Foundation, 23) Conference Board of the Mathematical Science Education Directorate, Washington, Sciences Committee on Computes Education. D.C., July 1979. MEMMIINIMmalLn.Mmbakilualt 39) Topics,k Instructional Computing, A Johan Education, ...CD1111, Washington, D.C. special publication of SIGCUE, 1975. 21) Conference on Isaiiithematicel Skills and learning. U. S.Department of Beath. Edu- cation, and Welfare, Euclid, Ohio, 1975. 25) Lennie, J. Richard. "Undergraduate Programa to Increase Instructional Competing in Schwas. ProoesdingeofIdathConferenceon Cote sting in the Undergraduate Curricula, last ignBilbge Web* 1977. 26) Dennis, Z. Richard, Dillhung, C. and Mitisnieks, 3. "Computes AotiVities in Secon- dary Schools in Illinois." Illinois Series onSlucaticeigApplications of 24, Univ. of Illinois, 1977. 27) 110.00014 D.M. "Problems of Implementation: Courses for Pre- and In- Service education ". for on and fps InSecondary, Schools. NortholollandPublishing Co., 1978. Invited Session

CAUSE PROGRAM AND PROJECTS

Chaired By Lawrence Oliver Program Manager CAUSE National Science Foundation Washington, D.C. 20550 (202) 282-7736

ABSTRACT PARTICIPANTS (Listed in order of Presen- Dirnation\about the Comprehensive EWE:GU-- Assistance to Undergraduate Science Educa- tion (CAUSE) program and a brief histor- A Brief Historical Overview of CAUSE ical overview will be presented with Computer-Related Projects emphasis on projects involving the use of Larry Oliver computers in education. Exemplary CAUSE National Science Foundation projects will be presented focusing on: 1) individualized testing using microprocessors; 2) minicomputers in under- Individualized Testing Using Micro- graduate laboratory science education; and processors 3) computer generation of instructional materials. Douglas MoCohm University of California, Davis

Minicomputers in Undergraduate Laboratory Science Education

Elisha R. Huggins Dartmouth College

Computer Generation of Instructional Materials

Michael'P. Barnett Brooklyn College of CUNT

1I c` 137 Tutorial

DATA BASES - WHAT ARE THEY? Arlan DeXock University of Missouri-Rolla 325 Math-Computer Science Bldg. Rolla, Missouri 65401 (314) 341-4491

ABSTRACT --WiTs7vorkshop will cover the three major approaches to ourrent data base hystems= hierarachica4vSODASYL, and relational. The relative advantages and disadvantages of each approach will be compared. In particular IBM's Database System IRS, _- Cullinane System IDMS, and Cincom's System TOTAL will be discussed.

149

138 Computer. Science Education

REQUIRED FRESHMAN COMPUTER EDUCATION IN A LIBERAL ARTS COLLEGE

David E. Wetmore St. Andrews Presbyterian College Laurinburg, N.C. 28352

(919)-27636!2x367

Computer education requires a definitive capabilities and all dour work was tionof the group to be educated* the fa- done in batch mode. Software constraints cilities (hardware and software) with limited our choice of languages to PL/1 which the education is to be accomplished* PLC* or FORTRAN. Most of our students are the goals of the educationalprocess: and not mathematically inclined* and so we the level and rigor of that process. In chosePLC because of its string handling practice* the first two factors oftende- capabilities. termine the last two. Over the next seven years* until the On the undergraduate level, there are acquisition of interactive hardware* our three fundamentally different groups to be computer education program evolved into a educated: potential computer scientists* fairly stable system.As our present sys- otherscience students* and everybody tem is a direct response to our earlier else. There issome agreement on the experiences* a relatively detailed des- needs of the first two groups. The needs cription of our old system follows. of the last group the educated members of society who are not mathematically or sci- Theprogram consisted of eight entifically oriented* are notso well three-hour sessions held weekly. The defined. This paper describesthe at average section size was about 25* and tempts* spanning a decade of one institu- there were usually six to eight class section to educate *everybody else.* tions. During the first week there was a shortintroductory lecture* followed by St.Andrews Presbyterian College iv the ideas of algorithms and flow diagram- a small (600 student) liberal arts col- ming. Students were given a simple bil- lege. In 3.969we instituteda science ling problem its flow diagram and its course required of all freshmen* *Selected program and wereaskedto punch up the Topics in Modern Science."This course is program* run it* and bring their output to both the terminal course for non-science the next session. maims and the introductory course for all Science PRJOVS.We felt that the ever in- At the second session the printouts creasing role of the computerin society were discussed*with emphasis on the de- required thatwe give all of our students bugging process. Then non-formatted 1/0 a familiarity with computers. We decided assignment* andunconditional branching that this could best be done by teaching statements were introduced. The assign- them the rudiments of a programming ment foroutside work was to modify the language. At that time we had no interac- billing program to handle multiple ous

139 150 140 NECC 1980

tamers by a loop and to write ane run a Second, the system must obviously be program to calculate averages. computer dependent. Inour old system, none of the assigned prograps would repay At the thirdsession we introduced the effort involved in writing and debug- arrays andconditional branching state- ging the programs. ments. The students were to modifytheir average programs to use arrays and also to Third, the material must becapable calculate the average deviation of their of ,gradual entry. New knowledge should be set of values. In the fourth session we required in small increments. Although held a review and then discussed comment our ,original sequence was modified quite lines andsimple output formatting.The frequently, we were never able to avoid students' assignment was to rewritetheir one sessionin which the stuaents were averaging program using comment lines and overloaded with information. formatted output to increase the under- standability of the program. Fourth, the system should be word-orientedinstead of number-oriented. In the fifth week characterstrings Most students aremore comfortable with wereintroduced. The assignment was to words than numbers. For example, we found produce a program calculating the average that interest inthecomputerincreased word length in a text. The sixth week was dramatically when we started to work with a continuation of character strings. The character strings. homework assignment was to write and run a program converting clear textintopig Fifth, the system must have under- Latin. standable errormessages. For beginning students a cryptic error message is prob- During the seventh and eighth weeks, ably worse than a simple statement that an each studentwasexpectedto design, error has been committed. write, and debug a program whichhe felt mightbe usefulto him in his academic Sixth, the system must possessthe career.Most students wrote suchthings attributes of any good tool: versatility, as checkbook balancing programs, grade usefulness, resistance to breakdown, obvi- .point average calculators, or simplesta- ous efficiency, and feedback to the user. tictical packages. We decidedthat the system which This educational venture was success- wouldbest combine the attributes listed fulin teaching students the rudiments of above was word processing. Word process- programming, the strengths and wealraesses ing includes programs to write text, alter of the computer and something of the po- text, move text within or between files, tential role of the computerinsociety. andtoformat text. The systems are not We were not successful in convticingthe only obviously computerdependent, they students that the computer vas a useful are impossible without a computer. It is teal, however, as most of etiom never used apparent that they will work withtextual itduring the rest of theiracademic material of any length or complexity. career. The major advantage of word process- In1977we ev:.:hased a Digital ing is its applicability. A freshman at a PDP-11/60 computer with eight video termi- liberal arts college knowsthat L4 will nals for stuetl-. use. During ourfirst write many papers over the course of the year wit% tails syst:....1 we did nothing more next four years. Any tool that will make than our former system by switching that writint easier is welcomed. toBASIC andusing the "MR' series to teach the language. The homework assign- Word processing systems areusually mentsremained unchanged. Our workcon- designed foruse by non-technical people vinced us that therearecertain attri- and are structured sothat much can be butes of an ideal curriculum "orteaching done with a few commands. Additional ma- everybody else about 1--vputera. I believe terial can be learned when its need is ob- that there are six sue" attributes: vious to the student.There areno quan- tumleaps in knowledge requirements. First, what is being done must be of Furthermore, because of VI,: usersfor obvious applicabilityto thestudent. which most word processing systems are de- Many college freshmen are not mathemati- signed, they arequitefail-safe, and callyinclined and do not foresee any use their error messages are usually clear and to them of the computer as numbercrunch- self-explanatory. er.

1 51 Computer Science Education 141

We chose UOT, a line-oriented editor The computer block lasts four weeks asour editing system. EDT is not as with fourassignments. During the first powerful as other availableeditors,but session the introduction ofthe manual, it is characterised by simplicity and ease editing program through appendingto of use. jinn was chosen as ourformatting files, and getting hard oopy are covered. program. It is capable of straightforward The students* first assignment is to acti- formatting with a very fewcommandsbut vate their accounts and write a file of at can be used for extremely complex format- least 100 lines in two different sessions. ting jobs. The student must hand in ahard copyof the file the next week. The textual material used inthe course is a 25-page manual consisting of For the second week the reading ma- the following information: terial is the remainderof the editing program section andthe text formatting Introduction material. This material is discussed and Logging on questions are answered. Thehomework as- Account security signment for the third week is to edit the Overview of the programs file produced during the firstweekso Logging off thatit will include formatting commands Files and file naming for case and margin control and forpara- graphing. A copy of the edited file is to Editing.Program be handed in. Entering the program Creating files During the third meeting the rema- Writing files inderof the manual is covered, and each Appending tc files student is asked to bringto the fourth Error messages sessionan output file (formatted text) Editing, an overview from the file produced during thesecond Text display week. Finding text Adding lines within text The fourth and final session is de- Deleting lines from text voted to questions from the class and work Changes within lines with individuals. The homework assignment (which is weighted twice as heavily as any Getting Hard Copy of the others) is to hand in, within three weeks, a copy of a computer - produced paper Text Formatting submitted for grading in another class. Introduction Case control An advanced word processing document Margin control is availableinthe college bookstore. Centering text The advanced document deals with additional features of the editing and formatting Displaying Files programs. This lastyear 174 students bought the required introductory text, and Account Directories about 50 copies of the advanced text were sold. (*) Spelling Checker Program Student evaluation of thecomputer Summary Table of Programs block hasbeen favorable. In our usual term-end evaluation it received an average rating of very good. Xi past years, prior Each new student atSt. Andrews is to word processing, the computer block was assigneda computer account. To use this usually rated as good to fair. account !*:. is necessary only tcego to the Furthermore, conversations withfaculty computercenterand obtain the account members outside the sciences indicate that nueber and password. Each May,accounts an increasing number of papers in upper aredeleted unless the student requests level courses are being computer produced. that it be kept. At present 230students Thus, although we have noharddata, we (37% of the student body) have active ac- feel that weare having some success at counts.About 22% of thestudents not our primary goal, that of showing all stu- taking a coaputer science course or the dents the utility ofthe computer as a freshman course have active accounts. tool.

152 142 NECC IWO

In addition, we feelthat word pro- cessingis an excellent computer application with which to show students the usefulness, strengths, and weaknesses of computer systems.

(0 Copies ofthe introductory handout may be obtained by sending a self - addressed, self - stamped ($0.70) large man- illa envelope to the author.

153 Computer Silence Education143

DEVELOPMENT OF COMMUNICATIONS SKILLS IN SOFTWARE ENGINEERING

John A. Seidler John G. Reinke

Department of Mathematics/Computer Science unieers'4 of Scranton Scranton, PA 18510

(717) 9S1 -7428

reasons for INTRODUCTION theway he reports are organized. Now such will an individual's programming skills beenhanced by an COURSE ORGANIZATION ability to present ideasorally and in The course isorganized writing? There is such more to around the programming tnan justwriting programs. presentation of student projects. For the first two Yet the emphasis in computing and weeks the students are given a general information science curricula orientation regerdingwhat is is on expected from them. programmingwitn little emphasis on many Daring the next three of the other essential tasks that surround weeks the students make their first oral presentations which moat be between 15 and areincluded in the programming and 25 minutes long process. Since 1975, we have been teaching and provide appropriate a course- that provides a broader view of general. descriptions 6( the Protect areas. The presentation software engineering. Tnis course, Should be directed toward convincing the audience of entitled Computer Projects, is an the importance of the projectAnd the undergraduate degree requirement that is impact it has on its environment. taken in the senior year.AS part of this The middle three weeks of thecourse course, studentsdevelop and document an are approved project of their choice andgive spentdescribing the various methods severalors presentations as wellas that can be used in the remaining reports several written reports. and the second oral presentation. Also, Normally, these projects begin during the students ate shown the the junior year. The projects can be done interconnections that should exist between the oral in a variety ofareasbut musthave a presentations and their various reculty\sponsor. Students are encouraged reports. The second half of thesemester to seek out faculty from other departments is devoted to the second oral presentations. who areinterested in using computing These are 45 to SSminute resources in someway to expand on presentations on the technical details of each project. existingprojects. In any case, all This talkemphasizes the projects musthavefaculty sponsors who appropriatenessof the technique selected are interested in the projects. to solve the problem. A key element in this course is its While the above sequence is occurring in class, organization. During the semester, each the studentsSubmit reports every student must give two lectures abouthis about two weeks.As a collection, these reports projectandwrite sevenreports. The are supposed to form a combination of the oral presentations with complete description of the project. thewrittenreports produces a total THE ORAL PRESENTATIONS of the project and makes the description Both oral presentations are evaluated students aware of a more global picture of by bothstudents the programming process as well as giving andfaculty member' present. Students are greded on a variety him a better understanding of the types of factors such as content, the useof of things that programmersandsystems visual analysts do besides grinding out code. aides,and the appropriateness of the level Subsequentsections of this paper of presentation for the describe the course and the rationale for audience. The first presentation is a 21 various items_ thatare included and the minute survey of the area of the project.

_I 'F 144 NECC 1980

rue purpose nere is to acquaint tne must be given in a way which would be audience wits tne problemarea and understood by a semi-technical manager or convince tnem oftne importance of tne executive. As a suggestion, students are project. told that they should produce figures that Tee second presentation is a 40 to 52 describe their project versus the status minutepresentation on technical aspects, quo, then emphasize howthese data are hardware, data organization, and arrived at legitimately. algorithms. Part of this presentation One purposeof these first three lb must InclUde, when appropriate, t reports is to give students a chance to structured walkthroughof a part of tire' collect their thoughts .during the first project. five weeksof the semester. The rest of the reports demand aconsiderable effort REPORTS on their part and require a thorough me seven reports, whentaken as a analysis of their projects. unit, should form a complete description of the project. Content of tnese reports Report Four. This report is a is a compromise between an ideal and what collection or-Teur analysisand design is realistic witnin the time frameof a documents. The first is $systems chart semester. In all cases, the reports are in the form of a tree structure that judged not just on their technical merits describesthetop-down organization of but also on the organization, grammar, and their system. Included with this chart is spelling. - brief description of the purpose of each module in the System. Report Ono. lids report is an The secondpart of this report is a exten ed abstract of the project and data chart. The data chart is in these should beaboutthroe to five typed parts--input, process,and output. The double-spaced pages long. the students input section describes the files used as are told to relate this report to their input to the system along with the record firstpresentation. Specifically, the formats of inpbt files. Each item ina first talk should elaborate on the general record is defined in terms of its type and problem area while tea extended abstract what role it plays in the system. The should only briefly mention it. output section describes both theoutput file and update files.By an update file Report Two. This tnree page report is we mean s file used both as an input and a requirements analysis. Here thestudent output. As in the input section, all must describe thehardware needs of his record formats and items in therecord project and describe now he came to his must be completely described. conclusions. To wqp the students, a list The process sectiondescribes the of general questions is distributed. This type, bound, and purpose of every major list includes: 'variable used. This section is grouped i. Wnat resources does the project first by describing globalvariables, require? (Primary memory, disk, then commonvariables, then a break down tapes, etc.) of other variables bymodule. By major 2. Time requirements - What effect variables, we mean that, for the sake of do various variables have on the brevity, the studentscan selectively time requirements? decide to limit the size of the data chart 3. Terminals - CRT, hardcopy, by not including all variables, for graphics, hign quality hardcopy example, temporary storage and loop -Which is ideal? indices. -Which are satisfactory? The third component in report four is -Baud rates? a 4Ata flowgraph. Again, due to time -Synchronous, asynchronous? limitations, the students are not required -Intelligent terminals, etc. to create a data flow graph of theentire 4. Program space requirements? system. Rather, they use several 5. Backup important items andshow how theyare transformed and combined with other data Report Three. This lip a justification to produce the results. Thedata flow document. The' students assume that they graph completes the systems chart in that are writing a document in response to a the systems chart shows the tree structure request by corpo ate managementfor of a system. justification of continued funding of the The fourth item in report four is a project. Here the students are required bibliography. This bibliography does not to defend the value of their project over just state the references. Instead, a the statusquo. Explanations of details brief statement mast accompanytech ,

.5,5- Computer Science Education 145

reference describing the relevant Report Six. This is usually the information obtained from that reference. easiest report for the students to write. Also,. the bibliography does not list only They are so wrapped up in thetechnical texts and periudicals. It also includes details of their project that the volume references to individuals who have of material that can be included provided information or assistance. It is contributes significantly to the grouped as follows: generation of this report. However, there 1. Text References are still problemsthey encounter in 2. Popular Periodical References construction of this report. (Time to Creative Computing) Their biggest problem is one of 3. PiEBNicef1717raical Re erences organization. The students are often so (all other periodicals) wrapped up in the project that they don't 4. individual Citations (persons know where to begin writing. who assisted) Surprisingly, theyfind it difficult to With report four, wehave a broad use the sectionsof report four asan based view of'the project. Thisreport outline for this report. Once they acts as a framework upon which the rest of realize the usability of reportfour in thereports can be built and integrated. generating report six, this report becomes well organized and easy to read. Report Five. Report five is a user's manual,esarr$7 the most difficult report Re ort Seven.This report contains for the students to write. The four sect ons7-61- first section consists fundamental problem is that_they know too of annotated and documented program much about their project, and they lack an listings. The documentation should appreciation of thedifficulties that explicitly indicate the connections others encounter (especially those who are between the code generated for the project not experts in computing) in simply using and the various reports. Section two is a the project. set of corrections to reports four, five, The key to improve this report is an and six, which allows students to modify emphasis on well-constructed ,examples. these reports ..o stay in line with the way Even this is a problem, however, because their system has evolved. the students usually create unusual Section three of this report is esoteric examples that are not typicalof antherdifficult section. Here students normal useoftheirproject. That is, are required to make a critical evaluation their examples are often like the horn :le of their system, pointing out flaws and examples we find in manuals that are deficienciesboth in their overall design produced by the various computer and implementation.This is followed by manufactures. section four, which describes what was Another problem tnat is evident in learned through the project. Included in this report is poor grammar. in order to this section are recommendations for correct this with aminimum effort, the modifications, enhancements, and ideas for students are told that the report is future projects. graded using the basics in Strunkand White's Elements of Style. Also, some time OSSERVATIOHS is spent on some theasics. There is no doubt that this course A final emphasis in this report is givesthe students a broader appreciation the importance of stating everything. of the entire hardware/software That is, many times students assume that development process. If there is a flaw, the readers of their reports have the same kt is not in this course: rather, it is background and knowledge that the authors that preceding courses do not require more have. As such, 'their report is not of thedevelopment process so that this completebecause there is still a course will be less of a shock to substantial amount ofKnowledge that is students. locket; up in their heads and not stated in Typicalof thevariety of projects black and white. that students have done in this course As a complement to report four,this are: report explains the details,critical 1. Developcant of relative and elements, andvarious interrelationships hierarchical data base in the system. Here, students are simulations encouraged to find the weaknesses inherent 2. General purpose software science in the various graphs and charts of report measurements system four. 3. Programmable calculator use in the physics laboratory

1ti6 140 NECC 1980

4. On-line accounting laboratory S. An enhanced editor S. Grapnics: Uktronics- Mablo interface 7. Graphics: 3-D perspectives with hidden line elimination S. Software for forcasting the university's financial status Byallowinga broad collection of projects, wa have obtained aninteresting by-product. Faculty from other departments, as well as various administrative offices, are more aware of the potential of our computing resources. Therefore, despitea stable, or slightly shrinking studentpopulation, we have created an increased demand for computing resources. This has given us an opportunity to obtain the variety of computing resources that we need to support our computer science program. Computer Science Education147

.1'

SYSTEMATIC ASSESSMENT OF PROGRAMING ASSIGNMENTS

Judy H. Bishop Computet Science Division Univetsity of the Ilitmeetstand Johannesburg 2001 SOUTH AFRICA (tel. (11)-394011)

SUNHARY style of program layout? The obvious way toassess a courseon The methods used to matk the assignments foe programmingis toassign merles to programs lets* firstyearcloses* orsmall third year written bythe students. However, it is not class,* do not seem approriate fora coitus evident st the outset that systematicassessment sized second year class (40-80 students). is feasible. Nor is it necessarily ap:4rent Program ossignmeets in first coursecan, for when features the meeker should look for and the themoat pert, be assessed by checking thee they weight to be attached to them.Thispaper work for certain samples of data (Noonan 19791 explains a method,taat hasbeen usedforfour althoughsometeachers ovoid this approach and yearsat two universities to mark course work at amass only by s special fora ofexamination second year level. The experience gainedshows (Rsdue1976, Troebetta 19791.At the other end that systematicassessment is possible and can of the scale, projects undertaken by third year be used to provide guidance to students as to and greduste students can be looked st what constitutes a good program. individually and marked over a widerange of criteria (Hell 1979]. Secondyear assignments BACKGROUND fell *members in the middle. One needs to give Thesecond year course in advanced program- marks forgood ptogtan and data design as well ming (Barron19751 is the backbone of the as for correct results, but the 'umbers involved computer science degree at both the Univetsity preclude spending hours agonizing over each of Southampton, England, and the University of program. In fact,the projects es set up the Witwatersrand, South Africa. Since 1974, it requirethe first three parts to be handed in at has been taught with the aim of turning outnot two weak intervals. Eachpertbuildson the just good, butexcellent progtammers. Like next so that it is desirable thee work handed in Noonan (19791, we felt that theprimarystudent on Friday be necked and returned by Monday. So ptoduct of sucha courseshould be a single, as not torely on theimpossible, end also largeprogramming project, split into four because typically some 40Z of solutions era not smaller projects. Two such projects mete handed in in final fora, specimen solutions ate designed to be usedin alternateyears, each distributed and students era free to usethese requiring a total of about 1000 lines Of Pascal instead of their ownin thenext phase. Mullis* 1974, 19751. Although e written Nevertheless, speed is a factot. examinationis required to tett a full knowledge As final complication, the within may be of data structuting end sorting techniques, the shined betweentwo or threepeople and this bulkof the final mark for the course is derived means that a standard must be well definedend from the project. At the outset, therefore, it understood. These thrisrequirements of quick, was realisedthat the marking of these projects accurate, end equal narking led to the design of bed tobe ofa higher standard than usual. the systematic method presented here. Very little guidance in grading can befound in the literature. Abehire (19781 in his THE METHOD otherwise excellent advice forteachers, skirts The methodquiteopenlyborrows ftom a the gables, stagily stating "Establish s grading technique used in faecal to mike programsmore policy ... Erode eachprogram ... Tell the readable. Instead ofworking innumbers, studentsyour grading policy." In petticular, marking schema is sat up vein words orphrases how does one neutralise the effect of subjective to deactibe the assessment of the solution under factors suches the student's previous perforem various criteria.Once all the solutionshave eoce, his presentation (nester essay),endthe been marked,weightsare assigned to the words preference of different markers for one or another end 4 final sack calculeted.This mark isthen

1 5S 148 NECC 1980

compared to the earlierestimatesmadeon impression. and the marks might be voderated one answer. (right, some wrong, none, way or the other.The six steps are still reading) 1. Divide thesolutionsintofive groups output design 0 (as required, based on evident output. improvement possible, shortcuts, 2. Set up s symbolic marking scheme. unformatted) 3. hark all the solutions sccording to the algorithm (elegant, competent, average, scheme. messy) 4. Assignweights to thewords used in achievement (morethanenough, fullfilled marking. requirements, gaps left, 5. Calculate a numeric mark. far too little) 6. Compare the estimatedandcalculated testing (complete, partial, insdeqeate) marks and moderate if necessary. self-contained (yes, no)

1. Rough Groupie. Divide the programs into groups according to Some comments on this scheme are: the evident output. 1. If there are no answers, the potential output A. Fully correct design will still-be assessed. 1. Answers partially wrong 2. In a text formatting problem, "shortcuts" C. Incomplete output, perhaps with an execu- would be given if, for example, formatting tion error suchascentering or right - justifying was D. No output - execution errorinreading done by spacing to an absolute column number phase instead of being based on the length of the C. Compilation errors line. "Unformatted " - would begranted when there is no centering orjustifying. These groups can be roughly translated into Possible improvements would be small things the traditional classes as follows: suchas leaving a line at the bottom of a page before printing the pagenumber. The A : First (802 +) (unless layout is rating of theoutput design should not be horrible or algorithms confused with that for achievements. messyor inefficleit) 3.The values given hereforalgorithm have B,C : Second beenfound to be applicable only when the or third(502 -792) (the largest group - marker is genuinely astonished at the ebould not get a First standard. A competent slgorithm is one in or Fall) which- D t Third - there are no redundant or out of place or fail (592-) (tricky group - they variables mustbe marked on - control structures are used correctly (e.g. criteria otherthan case, not cascading if's) results) loope cover short ranges (i.e. if a loop is Fail (492-) (depending on the level more than 20 lines long, part of it should of thecourse, these be in a procedure) might get zero) - the intention ofeachline is obvious, with commentsonly beingnecessaryto indicate yet-to-be-includedseztiens or 2. The Marking Scheme (Grading Policy) modificationsthat mightbe considered. A marking scheme is devised byselecting - the use of thelanguage is quietly apt, fiveor morecriteria on which the programs avoiding over- simplification and clever might differ, for exempla, the algorithm, design tricks alike. of output, schievement, and testing. In initial Most students' algorithms fell down somewhere e asignmente, the criteria mould emphasize points and are classed Si average. A messy program- S uch ISnest program layout or the correct use can be recognized a mile off! of procedures and parameters. In laterassign- 4. At somestsge, one wishes to emphssize that ments, a quick Ounce st the solutions hooded in proceduresshould bewritten, se far as willshow whether thee* lessons have been possible, es self-contained entities.This learned and ere nolonger worth assessing does not just meanthattheyshould have becsuse the standard is uniformly high. local variables,but that theycould be Each criterion is thenassigneda list of celled from different contexts and still between two andfivepossible values.These producethesameresults. This discipline are meaningful adjectives, spplicsblebothto is particularly important in a coursewhere thecriterion and tothe expected results. A one assignment leadstothe next. If typical scheme would bet procedures are not self-contained, they will have to be altered before being lifted out of

1 59 Computer Science Education 149

one program into another. programsillustrate. The other lists the 5. The testingctiterion is there to get importantpoints that should be commented on in students out of the habit of handing in only tutorials or ina handout, including moon Onetun. Veryofteninstructors Prefer a misconceptions or interesting approaches. singletunbecauseit is leas to mark, but in fact the opposite is true. It is easier 4. Weights to put the onus on the student to illustrate Only at this stage arenumerical Values aliat his program can do rather than have to assigned to the atlasusedin cIessifyinsthe deduce this from the listing. students' solutions. Each vales in each criteria is assigned a weight, so that the left Careful considetation mustbe given to the hand value*add up to thedesiredmaximum interaction ofindividualcriteria. A prima lark. Pot nest assignments, a maximum of 20is exampl isthat of testing versus achievement. adequate. Suitable weights for the sample If a feature is included, such as a test for an scheme above are: unexpected end-of-file,then a tun showing that it works should be handed in. This can lead to a progtan being given 3 1 0 achievement .6 fulfilled requirements answers (right, sommons, none, still reeding) testing Partial 5 if the end-of-file was checked for, but onlya output design (as tequited, tun with correct data wan handed in.On the 3 other hand, a ptogram that did not make the improvement possible, check would get 2 1 achievement = too little shortcuts, unformatted)

testing = complete 6 5 . 3 1 (2) because it had actually tested all that it could algorithm (elesent, competent,average, messy) do. Theweightings for these criteria (Step 4) 6 would be assigned suchthat the firstprogram achievement (mote than - enough, would get slightly more marks. 5 3 fulfilled requirements, gaps left, 3. Harking (grading) 1 The processof marking uses two tables. On fat too little) the first. values forthecriteria in the 4 3 1 marking scheme arefilled in for each student testing (complete, partial, inadequate) with the last two columns, "matk" and "calculated 1 0 class ", being left blank for the time being. eelfeontained (yes, no) On the secondtable, the group is recorded, alongside an "estimated class" and a comment. The estimated class represents one'sgut The maximum markin this scheme isactually teaction to the stmdents' efforts, and JO asfar mote than20because elegant algorithmsend as nany informal lurking systems go.The achievement note than required should be comment lists any relevant factors that woad rewarded by additional marks. These scores not show up in the marks and could be used at a couldwell becancelledout by the student later date to explain border-line cases.Typical fallingdown on sonsotheraspect, such as table entries would bet testing. Note that even the worst program vest getsome marks (for effort) so that sato dots TABLE I not appear for every criterion.

NAME ANSUERSI OUTPUT 'ALGORITHM ACHIEVE 5. Classifying J. Mullins rightas required' average ful.req Simple summations give the actual !arks for the first table. Prom these, classesate Coo t ITEST Iter MARICALC.CLAS1 assigned as follows: partial yes CLASS I MARX I 80* 16+ TABLE II 11-1 70-79 14-15 11-2 60-69 12.43 'NAM CROUP EST COMMENT . 'Good III 50-.59 10-11 J.MVIlinsl A I use of thelanguageI P 20-49 5-10 PP 0-19 0-4

Two further lists are useful. One records the names of students whoseprogramsat worth 6. Moderating distributing and notes thepoints that the Inthis finalstep the effort so far is

1 6 150 NECC

rewarded by comparingthe ,threeassessments: One of thecriticisms of the method may be - the rough grouping based on visible results that it does not encouragea fine enoughnet. the gut reaction class estimate Innearly all cases, the weights in the above - the mark calulated on detailed criteria. examplejump from 5 to 3 to 1. This is If thereareviolent disagreements between intentional. After fouryearsexperience, we the estimated and calculated classes, the have foundthat one cannot make the fine weightineeshould be adjusted first.Adjustment distinction between en algorithm that is was made in the above scheme becausea program excellent, very good, good, fair or poor. These with a very messyalgorithm got 16 marks and termsare too vague and are avoided in favour of really should not have got a first. Therefore competent, average and messy, with which one can messy was given a weighting of 1 instead of 2. identify a program more readily.Thus the marks do tend to bediscrete andcategorizedrather than continuous. Although the methoddoes notprovide any I II-1 11-2 III F built -In checks againstplagiarism, the fast pace with whichprograms canbe marked means Rough 12 ( 24 combined ) 9 that one'smemory is capable of detecting striking similarities. In four years, one case Estimated 11 13 5 of genuine plagiarism wasdetected andproven. 1 Numeroussimilar programswerefound, but by Calculated 12 3.%**.s.qiw10 6. 10 referring to the scheme, it was established that they differed in One or more important details. Usingthe methodhasdefinitely increased The above table shows acomparison from an our understandingofwhat constitutes a good actual assignmentgiven to 41 students.Only program. Ass result, our teaching has improved one program improved its rating on the and certainly thestandard of the student calculatedmark, Ails eight were demoted. The programs hasrisen over the years to such an value of the calculation method is evident in extent that me woulddisagreewitli Dolma and that close examinationshowed the estimates to Pozefsky (19791 who state that "Student- written be faulty. Three programs in Croup C hadtiny Programs accepted by computer science instructors errors andwouldhaveworked soon.They were areusually inferior to programs which exemplify estimated at II-2e. Then the scheme showedthat currently-accepted 'good' professional practice." they had shortcutor unformattedoutput in Our experience is that students produce programs additionand correctly placedthem in class which are just as good, if not better, than those III. On the other hand, another C program had in an "up-to-date programming shop". top marks for output, algorithm, and achievement Because the symbolic schemedoes nothave which compensated for the slightly wrong answers actualmarks, it can be shown freely to students and moved the class from a II-1 (estimated) to a so that they know what to aim for. The tables I (calculated). of detailed evaluationsare also invaluable in counteringqueries about results because the EVALUATION wordsareallthere to explain why the student Although the method takes a while to lost marks (e.g. shortcuts, inadequatetesting). describe, in practice it isa stream-lined procaas ehich can enable 60 assignments of about COHCLUSICN 300 lines each to beaccuratelyassessed in This method of grading isideal for medium abouttwelve hours. /he settingup of the sizedclasses where programming excellence is scheme takes about twenty minutes and assigning examined by practicalwork. It minimizes the the weights about tanminutes.The grouping, effects of prejudiceandprovides a balanced classifying, and moderating takeabout an hour. assessment of the factors that are being Men thisinvestment, the individual solutions particularlystressedat anyone phase of the can bemarked in tenminutes each. This project. It is adaptab'eto most types of compares favourablywith thenormof thirty projectsandhasbeenexpanded and used with minutes for an examination script. success in moreseniorclasses. It increases An important advantage of the method is that ones understanding ofgood programming, and it almost eliminatesprejudice of any sort. enables one to communicate this to students ina The example abtve was taken from the experience handyway, Finally,andmoat importantly, it of a marker whois usually quite generous and establishes confidence in systematicassessment showed that some 20Z of the estimates weretoo for both staff and students. high. On the other hand, there are some markers who equate "very good" With 7 out of 10 and ACKNOWLEDGEMENTS have a mental block against giving full marks. The AdvancedProgramming course vas devised With this scheme, they findthat they can by David Marron of the University of Southampton justify "complete" fortesting in their minds and it was at his suggestion that this method of and let the calculations take care of themselves. gradingwas designed,Thanksaredue to the

1 62 Computer Science Education 151

classes of 1975 to 1979 who responded withsuch enthusiasm to these ideas.

REFERENCES Abshire, Gary, "Techniques for Computer Science Teachers", SIGCSE Bulletin 10 (4), 42-46, December 1978. Barron,D.W., "Design and Construction of Computer Programs: A Course in Advanced Programming", Computer Studies Group, University of Southampton, 1975.Revised 1978. Deimel, Lionel and Pozefsky, Hark, "Requirements for Student Programs in Undergraduate Curriculum: Hon Huth is Enough?",SIGCSE Bulletin 11 (1), 14-17, February 1979. Hall, Colleen, "Third Year Project Harking Scheme", Computer Science Report, University of the Witwatersrand, July 1979. Hullins, Judy, "The Family TreeProject for Advanced Programming", Computer Studies Group University of Southampton, 1975. Revised at Computer Science Division, University of the Witwatersrand, 1978. Hullins, Judy, "The PlayStructure Project", Computer Studies Group, University of Southampton, 1976. Revised at Computer Science Division, University of the Witwatersrand, 1979. Noonan, Robert, "The Second Coursein Computer Programming: Some Principles and Consequences", SIGCSE Bulletin 11 (1), 187-191, February 1979. Radue, J.E. "On the Teaching and Evaluation of FORTRANService Course" SIGCSEBulletin (2), 32-35, June 1976. Trombetts, Michael, "On Testing Programming Ability", SIGCSEBulletin 11 (4), 57-60, December 1979. 152 NECC 1980

DATA STRUCTURES AT THE ASSOCIATE DEGREE LEVEL Richard P. Dempsey Computer Science The Pennsylvania State University The Worthington Scranton Campus Dunmore, Pa. 18512 717-961-4757

Recent design methodologies based on courses from the first year are: the structure of the data and the develop- 1) lst term. "Introduction to ment of data base management systems which Algorithmic Processes" This course use a wide variety of data structures add emphasizes solving problems using the weight to the importance of data storage computer as a tool.The language is and processing. The fundamental role either WATPIV or PL/C. played by data structures courses in 2) 2nd term. "Computer Organization applied bachelors degree programs also in- and Programming! This course emphasizes dicates its significance (1, 2, 3). the binary and decimal instruction set of Associate degree programs that train the IBM 360/370 Assembler Language. The students to be quality entrylevel objective is to get the student to see how programers with sufficient background to the computer operates at machine level. adjust to new trends in data processing This provides a solid background to make must provide these students with back- the student a better COBOL programmer and ground in both internal and external data debugger. All programs are run ou ASSIST structures (4). This paper describes the (described in (5)1. philosophy and content of such a course 3) 3rd term. "Introduction to Data taught as part of the associate degree Processing? This is not the usual course computer science curriculum at the using this title, but a high-powered be- Scranton Campus of Pennsylvania State ginning COBOL course. As the students University. have programmed in two languages already, this course moves fast. It covers such The objectives of the course are as topics as table handling, SORT verb, follows: sequential files on disk and tape, and 1) To present the structure and REPORT-WRITER. Emphasis is placed on functional characteristics of external programming style, documentation, and storage devices and their impact on file quality of code. Structured programming organization techniques. is emphasized throughout. The programs 2) To present concepts of internal are typical data processing problems. and external sorting and searching techniques. The students now have a solid, basic 3) To present the advanced COBOL programming background. Through the language elements associated with the "Techniques in Organization" course, this above. background will be reinforced while they 4) To present the concepts of in- learn to handle data in new ways. The ternal data structures. material in this course is presented with 5) To expose the student to the use an applications orientation and a minimum of libraries and utility routines. of theory. The objective is to give the 6) To make the student aware of student a working knowledge of data considerations in the design of data organizations and experience at using them. files. Some topics will simply be introduced and reinforced in following courses. This course, titled "Techniques of Organization," is offered in the first The order of presentation of the term of the students' second year. It is topics in this course is not always the a three-hour course. Prerequisite same. The nature of the programming

163 Computer Science Education 153

assignments, shifts in emphasis due to new good basis for what is happening in most of trends in data processing, and the makeup the data base management systems they will of the class itself all affect the arrange- surely face. Such knowledge is essential sent of topics. These topics, with to making effective use of these system. approximate times in parenthesis, are: Topic 6. Searching (1 week) A short Topic 1. Indexed Sequential Files time is spent on various searching tech- (1.5 weeks) A thorough working knowledge niques as they relate to the various data of ISAM is expected. Some time is spent on structures. This discussion further en- the physical aspects of ISAM files so that hances the students' understanding of the the students are aware of what is physical- significance of storing data in different 4. happening. The pros and cons of ISAM forms. files and the options available within it are discussed.The elements of COBOL Slotted with these six main topics are needed to process ISAM files are presented. some others, basically for programming Topic 2. Direct access files (1 week) purposes. Partitioned data sets are intro- Both relative and direct files are covered. duced to allow the use of libraries and the The concept of hashing is introduced. The COPY verb. These libraries are created for pros and cons of direct access files are the students, but the JCL and control discussed in relationship to the other file statements for IEBUPDTE are explained.. The organizations. The COBOL elements needed JCL and COBOL statements to handle sub- to process these files are also presented. programs are also presented. Topic 3. VSAM files (1 week) VSAM files are discussed, with major emphasis on The programming assignments for the the KSDS format. The physical character- course are all done in COBOL. Students are istics are presented and compared to those given all needed JCL, but all elements of of ISAM. Comparison is also made at the it are explained so they see the reason for COBOL language level between ISAM and UMW. each element of each JCL statement. All Topic 4. Sorting (1 week) External programs must be structured and completely sorts are covered briefly.One or two documented. They are evaluated on quality techniques are looked at so that the stu- and correctness.' dent can see the nature of what takes place Typical programing assignments are: in an external :oft. No attempts are mede Prog. 1. An input edit routine that to teach the student how to write an creates a sequential file on disk of the external sort. Internal sorts are also valid records. This file is then sorted discussed. In the first year the student in a second step by calling the system has programmed at least one internal sort, SORT/MERGE package via JCL. such as the bubble sort. Discussion here Prog. 2. An ISAM file created from is held to the relative efficiency of such the output of Prog. 1. Libraries and the sorts in relationship to data set size.A COPY verb are introduced. sorting technique, such as Quicksort, is Prog. 3. A report produced by sequen- presented as an example of a sort for tially processing the ISAM file from Prog. larger, internal data sets. 2. This assignment introduces the use of Topic 5. Internal Data Structures (3 the IEBISAM utility to allow keeping an weeks) Problems are discussed that empha- ISAM file as a sequential file. (Students size some of the limitations and problems are not allowed to keep permanent files on resulting from using only the basic inter- our system). nal storage techniques covered in the Prog. 4. It random update, including language courses. This discussion leads adds, deletes, and updates, processed into a working level presentation on linear against the ISAM file from Prog. 2. lists, stacks, queues, singly and doubly Prog. 5. A three-step program that linked lists, trees, and inverted lists. creates, updates, and produces a report Algorithms for handling some of these from a relative file. factors, as well as ways of manipulating Prog. 6. A program to implement the and allocating storage for them in COBOL Quicksort algorithm discussed in class. are examined. This assignment requires the use of COBOL Internal and external data structures subroutines and stacks. are compared. For example, a good tie-in Prog. 7. A program to build and of the similarities and differences would manipulate a singly linked list. be an investigation of how a set of records Some of the topics, such as direct that need a tree structure relationship can files and trees, are not implemented in bestored on external files using various this course. The next course, covering ad- file organizations. By this time the vanced assembler, advanced debugging, students are gettihg a good grasp of data utilities, and JCL, provides an excellent organization and its impact onprograming place for students to write programs using efforts, which should provide them with a these concepts. 154 NECC 19130

This course covers a lot of material 3. Fosdick, Howard and Karen Mackey, "A in a ten-weA term. Care must be taken not Cot:se in the Pragmatic Tools of the to get too seep into topics that don't Programming Environment: Description merit the attention in relationship to and Rationale," SIGCSE Bulletin, other topics. Creating more than one Vol. 11, No. 3, NUE7-1779, pp. 11 -13. course for the material is difficult as there are only six basic computer science 2. Mackey, Karen and Howard Fosdick, "An courses plus a projects course in our A- lied Computer Science/Systers curriculum. This limited number of e __aiming Approach to Teaching Data courses results from the University's Structures," S/GCES Bulletin, Vol. strong belief in the total edUcation of 11, No. 1, Feb. 1979, pp. 76-78. the student and concnrs with the guidelines set forth by %he ACM SIGCSE paper on 3. Beidler, John and John Meinke, "A. associate degree programs (6). Software Tool For Teaching Data Structures," SIGCSE Bulletin, Vol. A major problem with this course is 10, No. 3, Aug. 1978, pp. 120-122. the lack of a suitable textbook to cover such a broad range of topics.Most books 4. Little, Joyce Currie, "Computer Educa- only cover a subset of these topics end tion and Community Colleges," even then tend to be too elementary or too interface, Vol. 1, issue 1, Winter theoretical. The topic of internal data 1979, pp. 12-16. structures has proven to be a significant problem in this area. Barrodale et al (7) 5. Overbeek, R. A. and W. E. Singletany, has a nice approach for our applications Assembler Language with ASSIST, level but is not detailed enough; most Chicago: Science Research Associ- other data structures books are too de- ates, 1976. tailed antheoretical. 6. Little, Joyce Currie et al, "Curricu- This course has kept many computer lum Recommendation... in Computer science majors offthe streets a night or Programming," SIGCSE Bulletin, Vol. two, but the material and programs have 9, No. 2, June-WM-pp. 17-36. provided a soli.: background for our graduates. They become productive almost 7. Barrodale, et al. Elementary Computer immediately upon employment in such areaa Applications, Wiley, 1971. as application programming and technical support. Several have been able to use the course as a springboard into data base systems. This practical applications- level study of the area of data storage and organization has given them the experience and confidence to be quality members oi! the data processing field and to adjust quickly to the changing demands of their field. Integrating Computing into K-12 Curriculum

THE SCARSDALE PROJECT

INTEGRATING COMP'ITING INTO TH8 K-12 CURRICULUM

Thomas Sobol, Superintendent of Schools, Scarsdale, N. Y.

Robert Taylor, Teachers College, Columbia University, N. Y., N. Y.

Introduction communicate through tne computer

immediately and dynamically, in word, Within a few years every child in picture, and sound, with each other and America is likely to have at least one with others throughout the community whom personal computer. The potential impact they will never meet face-to-face. Thus upon schools staggers the imagination. At the nature of the communication will be the least, it is likely to move the focus transformed and the range of participants of education troll end product to process will be startlingly enlarged. and raise visual and auditory forms of information to a status rivalling that of This stage of mass computer use waa written language. Because ideas can be forseen years agot it arrives at last presented, explored, and expanded by human because computers have been improved, interaction with the computer, computing made smaller, and produced considerably is certain to transform the schools from cheaper over the last five years.Yet kindergarten upwards, its impact will be many of the questions raised in the minds as broad and deep as any intellectual of those who foresaw what was coming innovation in recorded history, including remain unanswered, How will use of printing. In addition to traditional computers affs.ct the child's development? communication, teachers and pupils will What are appropriate languages for young 156 NEM 1980

children to use in talking what changes computing might promote in to computers? What devices other than the education. traditional keyboard and screen or Given all these unknowns, if a public terminal/printer can, when attached to a school system takes the transforming computer, maximize its educational utility impact of computing seriously, where does for the young child?What should teachers it begin? Who does what with whom, at know about computers?Row expert should a what cost and with what type of success? teacher be in computing to make reasonably Though little research has been performed, effective use of it in teaching and though many more questions have been learning? What is the teacher's role in a raised than have been answered, and though classroom where every student has free nothing like a comprehensive intergration access to seductively engrossing computer of computing into the curriculum has been power? What is the impact of computer formulated, the computers are here, games on the child and what row should exciting possiblities beckon. To do such games have in the school?When nothing is unthinkable. This paper should the computer be used as a tool, reports one school system's response. when as a tutor, and when as a tutee?

The Scarsdale% Preipet Little significant research has been done on most of these questions, even by Two years ago students and teachers pioneers it computing and education. The in Scarsdale, NY, were doing little with research that has been done focuses on a computers. Three or four terminals in an narrow subsca of the issues such as the old office in the High School offered effectiveness of computerized drill and computer games to a handfotl of computer practice in arithmetic or spelling. There freaks, the terminals were down as often are few clear answers so far; often what as they were up, and nobody seemed to are advanced as answers are little more understand what the freaks were talking than strongly held hypotheses. There is about. Furthermore, inflatton and no ready-made curriculum, even in outline; declining enrollment had taken their toll there are po coherent texts; there is in the school system: summer school and little available teacher training; and driver education had been abandoned, and there is little public understanding of other programs were threatened. The time

. Integrating Computing into I12Curriculum 157

hardly seemed ripe for a brave new it. However, district history suggested curriculum venture in computing or that simply requiring teachers to take a anything else. new in-service course would not accomplish

everything. Grass-roots support, from However the community does enjoy an both teaching staff and community, would enlightened citizenry and teaching also be essential if the project were to faculty.Many were aware, however dimly, succeed. that computers were transforming the society and that they ought to be 19713/19744 transforming the schools as well. In the fall the District convened a Accordingly, in early 1978 the Scarsdale steering committee consisting of about a Board of Education appointed a citizen's dozen teachers and principals. Committee advisory committee to make recommendations members were charged with keeping their concerning the use of computers and colleagues informed of developments, computing in the schools.The committee's reviewing4progress, drawing up statements report, which gave the Board a platform of assumptions and goals, and making for action, recommended that computing be recommendations far the purchase of more introduced into the curriculum throughout equipment when appropriate. With the the school system and that appropriate consultant, this committee reviewed the steps be taken to purchase necessary Advisory Committee report, made plans for equipment and to train teachers in its the first in-service education efforts and use. The committee also recommended that decided on immediate hardware purchase. the district engage a qualified consultant to help in the effort.Work began The committee swiftly acted to immediately. acquire hardware, both to have it

available for the teacher training and to We believed the project should begin satisfy some long-standing student demands with the education of teachers --es many in the high school. However, the of them, system-vide, as possible. Too committee decided to limit first-year many school ventures in the past had acquisition to a minimum in the belief foundered because irre money had been that they could better make such decisions spent on expensive equipment than in after they themselves had had a year of helping people prepare themselves to use

1 6.cs 158 NECC 1980

training and experiencb with computers. in the classroom. First, FPL. an language

Because it was not clear that any one developed at Teachers College to teach

microcomputer would be best for programming, and BASIC were used to teach

everything, a mixture was acquired: four the crucial elements of programming.

Apples, file Pets, and one TRS 80. As Second, scheduled use of the ten

anticipated, the variety provided good microprocessors was built into the course

experience in making later acquisition so that every teacher participating had to

decisions. use them to finish the course. Third, selected articles by pioneers of computing As winter approached, Dr. Taylor met and education were read and briefly with all teachers in the school system to discussed. lorovide an overview of contemporary

computer technology and its implications Formally, the course was offered

for teaching and learning. (The teachers jointly by the district and by Teachers

actually met in several groups rather than College, Columbia University. The

in a single blocks English and Foreign participants could take the course for

Language, Social Studies, Art, and Musics Teachers College credit if they paid

Math and Sciences sand Elementary:The tuition, for local school credit if they

purpose of these overview meetings was to paid the nominal teachers institute fee,

inform teachers about the long range or for no credit, without charge.Of a

prospects and to build their enthusiasm faculty of 330, 122 enrolled, including

for beginning training. All teachers were several principals and the superintendent.

then invited to participate in the By the end of this course in late spring,

in-service course scheduled for 8 two-hour one third of the Scarsdale faculty had at

sessions throughout the winter months. least some familiarity with computer

programming and some hands on experience There were three main objectives to with a :ommon microprocessor. this first courses (1) to introduce the

teachers to the rudimentary concepts of Meanwhile, meetings were held with programming, (2) to get them over their parent-teacher groups and with members of

fear and accustomed to using the local press to explain the purposes of

sicrocomputers, and (3) to :mtter inform the project and to develop community

them of some potential uses of computers support.When time came to prepare a new Integrating Computing into K-12 Curriculum 159

school budget, the Board of Education had behind them were essential foundations for no trouble in i*,.:reasing its appropriation further thinking and for much of the

for computer equipment and additional action that made them obsolete.

training. 19/9/1980

Determining the next steps was less During the summer of 1979 a small easy. It is one thing to provide some group of teachers was paid to develop teachers with an exciting introduction to programming exercises for fifth and sixth computing, quite another to modify an grade children. (At this writing, at entire curriculum by integrating computing mid-year, many pupils have already raced into it. The vision shared by the authors through this material, despite the fact and the steering committee was only that till quite recently it might have general: computing should be incorporated been considered strictly high-school level into the curriculum.The problem was to material.) Based on the experience gained refine this general goal into specific, with using the first, mixed batch of detailed sub-goals.This task was all the microcomputers, the district selected and more difficult because most members of the acquired 19 more: 15 purchased from steering committee admittedly were not school funds, two donated by an outside experienced in current computing agency, and two purchased by parent technology let alone expert enough to groups. Each school then had a minimum of predict what computers would be like two at least two machines.More were placed or mere years into the plan when the in the junior and senior high school and curriculum changes might be realized. full-time aldes were hired there to staff Nevertheless, planning proceeded. a computing center in each.

To clarify what was underway, three Th:oughOuc the school year the documents (Appendices A, 8, and C) were Scarsdale Teachers Institute (the produced: a statement of assumptions, a in-service education teacher-run, set of curriculum goals, and a rough plan teacher-serving program which jointly for managing project activity over a sponsored the first formal introductory four-year period. Though almost computing course) has offered a series of immediately outdated by experience and mini-courses in programming in BASIC; events, these documents and the thinking

1 7 . .1Ms.mmb, 180 NECC 1980

certain junior and senior high school schools, with more to follow. And there teachers (primarily in math and science) is a broadening base of understanding and have begun exploratory use of computers in support of the project throughout the their classes; a concerted effort has community. been made to introduce computing and problems eneauntoreA programming to all fifth and sixth grade pupils in their classes;a selected group Any innovative project encounters of K-2 teachers have begun to explore inertia after the first enthusiastic push. dynamic graphics applications in beginning Scarsdale now faces the problem of what to reading and mathematics in their classes; do about the two thirds of the faculty who and formal in-service work of several have not learned about computers -- and kinds has been continuing.At the same what to do about the hundreds of time, planning for subsequent years enthusiastic children who have learned continued on various fronts. about them but must, perforce, be assigned

to the classes of teachers who have not. Units There is the problem of providing support

What has been accomplished thus far? to the willing teacher who is learning but-

To begin with, nearly one third of a who needs help. There is the ubiquitous district faculty which two years ago knew problem of money and of acquiring hardware little or nothing about computers now as rapidly as it can be used. knows the rudiments of programming, has And there is one other, more subtle lost most of its fear, and is beginning to problem that must perhaps beset all such try new things.A small cadre of teachers ventures in computer education until the has become enthusiastic and increasingly shape of the future is more clearly knowledgeable and gives promise of revealed. Computers have the power to leadership in the years ahead. Many change the ways in which people acquire pupils from the elementary grades through and extend knowledge, just as writing the high school have acquired a beginning changed such ways millenia ago.But until knowledge of the computer's capabilities we know better how those ways will change and of their own capacity to get the and how a school should be reorganized to computer to help them think. Computer capitalize upon those new ways, computer hardware exists in all the district's

17.1 Integrating Computing into K-12 Curriculum 161

technology is in danger of simply being computing that our understanding of it

harnessed for the pursuit of present goals will naturally reshape the way we think

within present modes of operation. To about everything. 4 employ computers thus is to miss much of Cencladjer school the revolutionary potential they represent districts and to heavily damp the enthusiasm and

insight they might otherwise foster in What might others learn from this

children.To finish by using computers project's successes and problems?We

for little more than enhancement within suggest the following:

the traditional curriculum and the 1) Begin with people, not equipment. presuppositions implicit in it would be The eager person who wants the tragic. The challenge Scarsdale and every machine and is ready to employ it is other school or school district faces is more valuable and catalytic than the to be sure this doesn't happen. machine that no one understands.

,The Goal 2) Develop broad support throughout

Using computing to practice your community and teaching staff.

mathematics and language skills in 3) Engage in on going, system-wide traditional curricula is helpful but planning. Do not leave matters to primitive, almost like using televisor an individual school or department primarily to display pages from textbooks. or to an outside consultant. Serious questions must be raised about

mere productivity of educational 4) Use a consultant or consultants who

approaches which limit visions of computer can bring not only technical

application to such narrow horizons. Some knowledge of computers, but an

of the skills for which computer understanding of schools and the way

assistance is so easily designed may people use their minds and

actually be of significantly less imaginations.

importance as computer access continues to 5) Concentrate resources of time, increase. Hampered by ignorance, enticed money, and attention on the project. by expectation, Scarsdale seeks a grander In this way you can achieve a sense goal: to become so comfortable with 162 NECC 1980

of purpose and direction despite the Appendix A

shrinking you may be suffering assumptions re coo PutoXfiMAd

elsewhere in the system. C81911th18irL-inaLLIIctiall

6) Don't wait for the perfect computer; 1)The influence of computing on human

current machines despite their costs life and on the entire planet's

and limitations are a perfectly good development will continue to

introductory device for teacher increase exponentially for the

training and pupil experimentation. foreseeable future. A great deal must be learned now if 2) All pupils should be educated to cope better machines are to be well with this growing influence; as employed when they do become many pupils as possible should be available. educated to master computing and to

7) Don't limit yourself to one type of use it creatively.

machine; get experience with 3) Such educating toward computing should several.As cheaper, more powerful be general rather than specfic ones bow in, you will have a better since hardware and software changes experiential background by which to will continue to occur at such a judge them. rapid rate that specific training

8) Find out what you can do with the will rapidly become obsolete.

hardware as it stands, then do it. 4) Any human being of normal intelligence Don't always wait for expert can understari the fundamental software; both teachers and principles of computers and children can learn powerful things computing, can operate computers, from exploring a machine's basic and can learn to write simple capabilities to plot, edit, speak, computer programs. delete, and so forth.

5) Pupils should begin learning about

computers and computing from the

time they first enter school and

should continue such learning Integrating Computing into K-12Curriculum 163

throughout the grades. 10) Computer hardware will become

increasingly less expensive, and 6) Pupils should also use computers as an both hardware and software will aid to other studies, in the become increasingly available. humanitits and social sciences and

the fine arts, as well as in 11) Computers can serve as tutors,

mathematics and natural science. teaching us skills and information

according to programs written by 7) Pupils learn to use computers by using others. Computers can also serve computers.The use of computers as tools, analyzing data, should be as thorougly integrated performing calculations, and so into the school program as the use forth. And they can serve as of pencils or chalk. tutuees as we teach them to solve

8) Computers are both an object and an problems, compose music, etc. Bach

instrument of learning. use has its place, but the most

Traditionally we study writing to important is the last.

learn how to write, and we use 12) Programming computers requires logic writing to refine thought and and precision. Pupils who learn to express feeling. so too should we program computers practice logic study computers to learn how to use and precision. them and use computers to extena

and refine our thinking. 13) The study of computers and computing

does not run counter to the spirit 9) If pupils are to learn about computers of humane studies and the exercise and computing, their teachers must of free, creative intelligence. On learn about computers and the contrary, it extends and computing. If all pupils are to deepens them by extending the power learn about computers and of the human mind and may foster a computing, all teachers must learn much needed dimension to our about computers and computing. intelligence. 164 NECC 1980

Appendix Et 3. Throughout the grades interested

guriculum Pupils should have opportunities to extend and enrich their knowledge 1. From kindergarten through high school, of computer programming. In the at increasing levels of elementary schols these sophistication, pupils should learn opportunities might take the form the fundamental principles of of after-school workshops or computers and computing. Some special individual or group Pupils should learn to write simple projects; in the junior high computer programs in elementary school. of projects( mini-courses, schools all pupils should learn to or club activities; in the high write simple computer programs school, of specialized elective before the end of their junior high courses, projects, or club school years. In grades 9 or 10 activities. all pupils should complete a (Sow can we make this item more one-semester course in computing, specific?) computer science, and the social

issues raised by computer

technology. Other high school 4. Interested high school pupils should

computer studies should be oriented also have the opportunity to study

towards computer skills which computer science, computer

students will need in college, such technology, and data processing.

as computer applications in

numerical analysis and statistics.

2. From kindergarten through high school

pupils should use increasingly

sophisticated calculating and

computing devices as an aid in

studying mathematics, science, and

other subjects.

1 '70 Integrating Computing into K-12 Curriculum 165

Appendix C Computers used in science courses.

Curriculum: Pour-Year Plea Computer applications in math developed. 1978-79 Independent study projects available. Elementary: None. Computer Club active. Junior High: Computer club plus? 1980-81

High School: Computer Center open. Elementary: Computers used in science courses. Mini - courses for 5th /6th graders; most have hands-on time 1st semester: Introductory with desk-tops during Computer Course. both years. Increased/experimental Computers & mathematical activity at primary application course 442 grades. Units for - 1/2 credit, 1 primary grades included semester course. in social studies, science curriculum Computer used in 9th grade math guides.

Computer Club makes much use of Junior High: center. Units on computers in social 1979-80 studies technology units. Elementary: Computer units in math courses both 1. Three week mini-course for years. intermediate students rotates among five Active Computer Club. elementary schools. High School: 2. Consultant and 2-4 primary teachers work with K-3 All before continues. students. Increased computer use in social Junior Highs studies, business, and science courses. a. Each 8th grader has an 1981-82 equivalent of one week mini-course on Elementary: computers (hands-on). Nini-courses for all 5 -6 graders, b. Each 8th grader has two week all have opportunity computer/literacy Or hands-on computer course in spring as work. part of social studies curriculum. Activities for primary grades included in social High School: studies, science, and Math guides. Introductory programming (2 semesters). Junior High:

Advanced programming courses Continuation of 1980-81. developed. High School:

7G 166 NECC 1980

Continuation of 1980-81. Expanded February). use of computers in non-math departments. e. Course on graphics and music.

f. Guest speakers - Computer tzpi)*hrerehitirL PrpnitA literacy.

g. Visits to other schools on an First Semester. expanded basis.

a. Consultant meets with staff in h. Social studies, science, and various formats. business faculty gain experience with b. Guest speakers meet with staff. simulation materials through BOCES course. c. Committee reads, discusses, meets, and argues. 1. Send one or two to computer Engages others not on repair school. committee in similar conversations. j. Course on programming for CAI (2nd semester). Second Semester. 1980-81

a. Consultant continues to meet a. Consultant continues support with staff. group on computers.

b. Consultant teaches literacy b. One or two faculty take course in Scarsdale. advanced college courses. Faculty visits other schools to examine use of c. Course in FORTRAN, etc., computers. offered to all with previous experience. Issues:What do we want kids to knows d. Introductory BASIC course repeated for those with What learnings are appropriate for -no computer experience. each age? e. Release-time workshops for How do we best teach about those with no computer computers? experience.

How do we go about organizing f. LOGO on the APPLE course for ourselves? elementary school teachers. How do we involve the entire staff? 1979-80 g. CAI programming course.

a. Consultant advises computer =78-79LI committee and others. 19

b. Advanced seminar for those who 1. Prepare statement of had first course. aspirations.

c. Four session course on BASIC 2. Educate ourselves. for novices and literacy graduates 3. Design structure to involves (Sept./Oct.). entire district.

d. Follow-up mini-courses on 4. Plan and implement system-wide educational in-service program. applications & programming led by in-house staff (Nov. - 17 Integrating Computing into K-12 Curriculum 167

S. Plan 1979-80 computer budget 2. Continued expansion of (16,000) curriculum off4rings.

6. Purchase 1979-80 computers and 3. Plan 1981-82 computer budget. software. 4. Make 1981-82 purchases. 7. Design courses at high school 1981-82 level. 1. High school curriculum in 8. Plan 3-4 yr. purchasing place. program for computers. 2. Junior high curriculum in 9. Design a mini-course for place. elementary students. 197v-80" 3. Elementary curriculum in place.

1. Continue system-wide general 4. Write computer scope and education. sequence.

2. Provide follow-through in-service for teachers.

3. Plan 1900-8i budget (21,000 for hardware, software).

4. Make 1980-81 purchases.

S. Design courses at high school level.

6. Pilot units at juor high level during 2nd semester. . -

7. Pilot mini-courses at elementary school upper grade level. experiment at primary level; expand primary level computer use, teach technology unit.

8. Develop maintenance procedures.

9. Integrate computers into expanded Sth grade technology unit.

10. Integrate calculators, computers into math curriculum.

11. Search for and write grant for computers.

12. examine software. 1980-81

1. Focus in-service on advanced courses.

1I*;*fc tj 168 NECC 1980

INTEGRATING COMPUTING INTO K-12 CURRICULUM

Beverly Hunter Catherine E. Morgan Human Resources Research Organization Dept. of Instructional Planning and 300 North Washington Street Development Alexandria, Virginia 22314 Montgomery County Public Schools (703) 549-3611 850 Hungerford Drive Rockville, Maryland 20850 ABSTRACT (301) 279-3321 aliFFically, schools and school districts have employed a wide variety of ABSTRACT strategies to integrate computer-related Ugger-based instruction in the Mont- activities into the curriculum. Several, gomery County Public Schools, a large such strategies are reviewed, based upon public school system, includes mmputer- a national sample. The need for a compre- assisted and computer-managed instruc- hensive computer literacy curriculum, tion, computer literacy, computer math- K-12, is described. A plan to infuse ematics, and problem solving. The inter- computer literacy objectives and activ- active use of computers for children ities into the traditional curriculum is begins as early as third grade while the set forth. Implications of such a plan management system in mathematics monitors on teacher training and curricular mate- individual student progress from kinder- rials me discussed. garter. Invited Session

FUNDING ACADEMIC COMPUTING PROGRAMS

Sheldon P. Gordon Suffolk County Community College Belden, New York 11784

Lawrence Oliver National Science Foundation Washington, D.C. 20550

ABSTRACT At most institutions, the single most computer oriented activities. These would critical problem in the development, imple- include CAUSE (Comprehensive Assistance to mentation, maintenance, or expansion of any Undergraduate Science Education), LOCI academic computing program is lack of money. (Local Course 'mprovement), ISEP (Instruc- This presentation will focus on a variety tional Scientific Equipment Program), and of possible solutions to this problem.A MISIF (Minority Institution Science number of alternate sources of funding, Improvement Prograzi",. including grants from private foundations, In addition, the presentation will grants from government agencies, and include a discussion of some of the contributions from local business and important do's and don'ts connected with industry, will be discussed.Primary proposal writing, and will relate them to attention will be devoted to the most the process of gre.nt review as conducted likely source of outside funding, namely by the NSF, government grants, especially those availabl through the National Science Foundation. In particular, details will be provided about the respective objectives and limitations on those NSF programs that would most likely sponsor various types of

169 1 S Tutorial

TECHNIQUES FOR INSTRUCTIONAL SOFTWARE DEVELOPMENT USING MICROCOMUTERS

Kevin Hausmann Minnesota Educational Computing Consortium 2520 Broadway Drive St. Paul, Minnesota 55113

ABSTRACT 7%saWocomputers become more and more 1) Adequate spacing of material on the available and easy to use, more and more screen. people will be designing and writing soft- 2) Efficient movement between frames. ware for them. Before a project is begun, 3) Effective presentation techniques. however, there are several obvious but often 4) Good ways to ask questi,ns. gotten considerations: 5) Effective feedback for answers. 1) Is this a reasonable application for 6) Adequate arS appropriate use of a microcomputer? graphics and other special features 2) What are the limitations of my of the microcomputer. equipment? After an application is developed and 3) What modes of interaction should written, it is very important to adequately be used? test the program. Many times this is best 4) What support materials may be done by watching someone use the program required? who is totally unfamiliar with it. Once the analysis phase is completed, More detailed written material of this the design and layout stage may begin. nature is available from the Minnesota Design should center primarily on the Educational Computing Consortium, 2520 organizational content and division of Broadway Drive, St. Paul, Minnesota 55113. the material. Also at the design stage, Ask for the Apple Authoring Guidelines. one should consider the mode of presentation of the material. So far, all the time spent on the project has not involved any time programming the microcomputer. After the design is planned out, one enters the ithplementation stage. Briefly, some items to consider here includes

170 Mathematics

A NETHOD FOR EXPERIMENTING WITH CALCULUS USING. COMPUTER-ASSISTED INSTRUCTION Frank D. Anger L Rita V. Rodriguez Department of Mathematics University of Puerto Rico Rio Piedras, Puerto Rico00931 (809) 764-0000

INTRODUCTION Second, the project introduced or The project which is the object of advanced sweeping changes in the curricu- this report was carried out in the (a- lum. Rather than simply paste on a little cuity of Natural Science at the Univer- computer-assisted instruction (CAI), we sity of Puerto Rico, Rio Piedras campus, took into account the full importance of during the years v.976 to 1979. The Na- the computer to modern science and in tional Scienc. ration sponsored the particular to the working scientist. Thus, project through its Minority Institations programming was made mandatory for all Science Improvement Program (MISIP), majors in the faculty, CAI was introduced making possible a wnole new aspect of in all basic science courses, and inter- instruction and learning for 1600stu- active computer utilities were madeavail- dents. Although the ptssent paper will mble.for students of certain advanced concentrate on the computer- assisted laboratories. In the calculus courses, instruction modules developed for the greater emphasis was placed on the use of introductory calculus course that we the calculator and on the numericalmethade believe are innovative themselves, we of calculus. most begin by describing briefly a few Finally, the project created the Na- points which make tnis project, as a tural Science Academic Computer Center v121e, unique. headed by a member of the faculty and First of all, as the composition staffed exclusively by students. Rather of the student body is that of the than implanting a ready-made system to Spanish-speaking culture of Puerto Rico, which students have recourse, we thereby the various materials of the project created an organic exterugion of the stu- were produced in Spanish. In addition, dents' environment, providingher imme- experience has shown that although UPR diate incentives for probing deeper into mantains a selective admissions policy, the uses and operation of computers. This entering students are frequently defi- last area has been one of the high points cient in analytic skills and in expe- of Ale project. (4) rience with methods of scientific inqciry. Inadequate or, frequently, THE MATHEMATICS COMPONENT incorrect preparation may lead a studen4.. The project developed in an orderly to frustration or failure when faced series of stages, beginning with the with laboratory or problem-solving si- training of staff, the search for appropriate tuations. hardware, and various necessary

171 172 NECC 1980

administrative measures. Although every one tutorial. Thus, we sought three basic part of the project seemed to be fraught things from each module: with unexpected difficulties, the faculty 1. It should perform some graphic as a whole responded more positively than or numerical simulation of some funda- expected, and their cooperation along with mental process, concept, or technique. that of various student assistants kept 2. It should be highly interactive things moving forward at all times. In and open ended, allowing the student some the Mathematics Department, the necessary freedom to experiment with the concept changes and investment of time were the under investigation. greatest, since to 'his department cell 3. It should minimize the difficul- the full weight of training professors ties faced by the student using the com- and implementing the new programming course puter. requirement for the whole faculty. We soon These objectives have much more ob- realized that the applications in the dif- vious implementations in the experimental ferent discip'ines involved were better sciences than they have in mathematics, handled by U..* new programmingcourses and we feel that there is still much room rather than by the old course (see Figures for exploration and development in this 1.a and 1.b). The pre-medical and biolog- direction. Nonetheless it is herein that ical science students were better served we feel much of the novelty and value of with a programming course that would inour modules lies. clude data gathering, statistics, and Math 211 examples in their own discipline. There Calculus I are many packaged programs that, with some Math 103 Math 104 3 credits programming knowledge, the student may Precalculus I Precalculus II easily adapt to his needs.On the other 4 credits 4 credits Math 107 hand, the physics, chemistry,and mathe- Programmine matics majors were to be taught more Figure 1(a) Previous Basic Matikaatics 3 credits sophisticated programming, also wit% exam- Sequence ples from these discipliner. /hdt 100 Math 200 Math 21: Calculus for Programming To train more of the faculty members Remedial Math 105 Biology and Math 1 Precalcuks with Statistics in PL/1, programming seminars were given. Life Science 3 credits We chose the PL/I language in particular 3 5 credits 5 credits because we believe that it is extensive credits fiaUt 205 Math 218 enough that the student who knows it may Fiigu a 1(b) New Programming teach himself whatever langUage the ma- Basic Mathematics Calculus I for 3 credits chine he has at hand uses. Parallel te Physical Sequence Sciences and this, the calculus course was divided Math 5 credits into two courses; the first, as in the programming course, designed to serve the Before the actual structure and con- pre-medical and biological science stu- tent of the individual modules are dis- dents and the second to serve physics, cussed, it will be instructive to look at chemistry, and mathematics majors (see some of the story of their development. Fig. 1.a and 1.b).These divisions re- At the start of the project we wrote to spond to long-standing curricular tensions numerous publishers and universities in in a faculty in which about two thirds of order to obtain Information on texts and the students will go no further inmathe.. already existing CAI for calculus rouses. matics than the first semester of calcu- Although we then found some interesting lus while the other third will need at supplementary texts, it soon became ap- least two or three semesters of more parent that very little had been done for advanced mathematics. calculus like the interactive programs that we envisioned. Algebra, statistics, THE COMPUTER-ASSISTED INSTRUCTION FOR and linear algebra seemed to be the areas CALCULUS that had attracted the most effort for a The major reason for the implementa- variety of reasons running from the size tion of CAI at the University of Puerto of the student populace to the appropri- Rico was to combat a lack of analytical ateness of the material for programming. skills and to overcome, to whatever extent We found the nearest thing to our aspira- possible in such a limited program, the tions at Georgia Institute of Technology tendency toward rote learning and memori- where Dr. J. C. Currie and a small group zation as the basis of learning.We de- had developed some rather brief programs sized, moreover, to avoid doing with the for illustrating limits, derivatives, ap- computer what could be done equally well proximate areas, and other topics along or better by more traditional methods, with a driver program which we totally with the possible exception of the one-to- failed to appreciate at the time. (6) Far

13 lifm=mmnimmIllmsleImmnEllmamsimilllEmMmwEIMMIIngi!!!!MO=111 Mathematics 173

more impressive things had been done for the time consuming operations often asso- logic at Ohio State University, for sta- ciated with the beginning of the semester. tistics at University of Akron, and for The structure of this program is shown in non-mathematical subjects in many other Fig. 3. (Of course the security with this places; we wanted to see something similar s;.tem is minimal, but in our environment done for calculus. this is not a problem.) This same program As we learned more about the diffi- maintains data on the total number of uses culties of compatibility and transferabi- of each module by each student and the lity and thought more about the problems total number of minutes of terminal time of translating whatever we found into of each student. Three separate report Spanish for our students, we leaned to- programs have been written to produce dif- wards developing our own programs out of ferent usage reports for the teachers of the many materials and ideas which we had each section and for global evaluation. collected during the first year of the No data are kept on right and wrong res- project. We had finally settled on buying ponses: it will become clear further on a Hewlett-Packard 2000 System for its re- that such information is either not ap- putation for fast response with a large -plicable or irrelevant for the majority number of users, to give us independence of response situations presented. from the main competes center with its The general form which most of the many large administrative jobs, and be- modules follow is given in Fig. 4. It cause it supported Coursewriting Facility, must be remarked here that these modules an approximate implementation of IBM's are nut in any way conceived of as re- Coursewriter with available BASIC numerical placing the textbook or the lecture. :hey functions. We had experimented with the are strictly supplementary and each one latter language on UPR's IBM 370/145 com- presupposes that the student has already puter and believed that it was a rather been introduced to the concept or tech- powerful tool providing the necessary nique in his regular class work. Thus,4m structure for developing good CAI. We the first part of the module there are' were eventually to come to grief with usually a few questions to find out if the the Coursewriting Facility, but we remain student is at all familiar with the mate- pleased with the rest of our system; rial, and if not, to recommend that he after many months of work in this lan- learn more about it before gotng on. The guage we found ourselves unwillingly examples which the computer presents are forced to convert everything to BASIC, as much for the sake of familiarizing the which is now the language in which all student with how the computer output looks our programs are written. and how to later enter his own data as they We arrive, then, at the modules are for illustrating the material being themselves. It was for the calculus studied. That is to say, it is expected course of the hard sciences and methane« that the student will continue making up tics (Math. 20S in Fig. 1) that the CAI his own examples and investigating the modules were produced. The various pro- computer's responses and will do most of grams which make up the entire CAI system his learning during this latter exchange. in calculus are depicted in Fig. 2. Those students who stop after the prepared Within the system, the student has examples are not really likely to ;lave complete freedom to choose among the gained much. Fig. S presents the eight different modules at any time, although modules and their content essentially as he is informed in class when it would be the manager program presents that menu to most appropriate to study each module. the students, except of course, it appears The manager program, which interfaces in Spanish on the screen. between the student and the individual lessons, also allows the students to register themselves, avoiding one of

Manager Pram TheMIModules Report, Facilities Registration S sa-on an I. INTRO 2. NUMBER]ElE Rised Students by SectionI Securit 1 of Students 4. SECANT S. GRAPH 6. CURVE Amber of SuccessfulCompletions! Offer Metes and Branching of tea tbdula by Student 7. NEWTON ITotal. Usage Time for eachtbduS1 by Student

Figure 2. The Calculus CAI Programs 174 NECC 1980

Present of /Modules 1----- Select Module I

Figure 3. Manager Program Flowchart

1. INTRODUCTION

2. 9pESTIONS. Does the student know enough to benefit from the module? If not, he may be asked to return to the text before continuing.

3. INTERACTIVE EXAMPLES. The student participates in working through an cx..mple selected by the program.

4. EXPERIMENTATION. The student is given the opportunity to make up and enter his own examples in the

sail format as in 3 . The computer provides the necessary data and prompting.The option to rerminate is offered before each example.

ob.a.mee dm. el re.wwer...mr Fi3ore 4.CFneal Scheme for Modeles Mathematics 175

1. INTRO Introduces the terminal, the keyboard, and its traits.

2. 111142814 Presents real numbers, round-off, and how to read mathematical tables.

3. LIMIT : Studies Limits of rational functions with values that make the denominator equal to zero (0) by the use of a table.

4. MAW: Illustrates how the slope of the secant tends toward the slope of the tangent.

5. GRAF : Calculates values of a given function and its derivrtivesinorderto help plot the graph.

6. CURVE : Evaluates the function,finds the inter- cepts, critical points, and points of inflection of a polinoeisl.

7. NEWTON: Finds tne roots of a given polinomial using Newton's method.

8. AREA : Calculates the approximate integral by various methods.

?IMRE 5.The Eight Calculus CAI Modules

Table 1.Calculus *Wale Usage Number of X of Modules Average tamdwar of Total Students Students Used per Sections Students Total participatingparticipating Student Minutes

,1Pall 5 150 37 25 % 3.3 65 78-79 Spring 78-79 9 270 123 441 5.6 84

Pall 200 67 34 2 3.1 68 79-80 7

Sable 2. Average of Responses to Module Questionnaire', 1978-79

relevantrelevant to to high recommend long clear interesting helpful coursestudies level fast it

3.5 6 3.6

2 AA

tart confusing boring not irrel- irrele- loso slow don't helpful elevent vane to level recommend to course studies it

is 6 178 NECC 1980

The first two modules in fact have nothing sequence of steps until the student names to do with calculus. They are there to the root being approximated with suffi- minimize the shock of dealing with the cient accuracy or until it becomes clear computer for the first time and increase that the process is not converging. (As the probability that the student will be you may expect, the only method the pro- able to successfully use the modules and gram has for determining the root is the interpret the results. Next comes a mo- senses Newton's (Method, so that in some dule on limits and then on the derivative, untidy cases the coMpeter gets it wrong, both centered around investigating tables leading perhaps to some confusion. This of values of the appropriate expressions sort of thing is, however, a constant in the neighborhood of a chosen point. source of difficulty in many of the mod- There are no epsilons or deltas here, and ules and in fact in all numerical methods. it is hyped that the student will get more Some care has been taken in the program- of a feel for how varying values approach ming to avoid some of these pitfalls and a fixed number. The next two modules deal to give some warnings in the text portions. with the use of the first and second de- The student manual, which offers various rivatives to describe certain properties forms of encouragement and hints, also of the graphs of functions and to draw points out same of the difficulties.) The those graphs. It is our continuing dis- student then has the option of trying for. appointment that these modules contain another root of the same polynomial, en- very few graphs. Despite the fact that tering another polynomial, or terminating we had proposed to buy some graphics ter- the module. minals and even a plotter, we were never able to do so. Several programs were RESULTS OF THE CAI IN CALCULUS developed to produce graphs with alphabetic Throughout the duratioa of the proj- characters but were never actually incor- ect, and continuing.now, certain evalu- porated into the fiaal modules. It is ative procedures were used to assure some still not clear whether a graphing routine kind of objective information about par- cannot make a straight line look like a ticipation, reaction, and effect of the straight line that will be very CAI. The first of these, participation, convincing for the students. The seventh was constantly monitored by the computer module illustrates the Newton- Raphson itself. Table 1 shows the percentage of Method for finding approximate roots to the enrolled students who actually used polynomials, while the last module does the modules to some degree or other, the approximate integration by three of the average number of modules run by each standard methods frequently treated in student, and the average time spent by introductory calculus courses. each student at the terminal during the To demonstrate the implementation of semester. The relatively low use is due the general structure of the modules as irincipally to three factors.Foremost shown in Fig. 4, we shall disease in more is the totally voluntary nature of the detail the seventh module: Newton.The computer laboratory. Any such voluntary introduction reminds the student of the activity cannot hope to attract more than objective and iterative nature of Newton's 75% of a class unless it guarantees better Method ant' requires him to recall the pre- grades.Our modules offer no such guar- cise form of the iterative formula, using antee. Secondly, there were considerable one of the very few multiple choice formats difficulties during the first semester of in these modules. Once he gets that 1978-79 due to the delay in completing straight, the sample function, f(x) e the final home for the computer center and the 4 x3 -5, is presented. The studeat isdi- resulting lack of terminals and comforta- rected to choose a starting value, x , and ble working conditions. At that time the computer then calculates, one 0at there also was no student manual, leaving a time on request, x,, until the the students more or less on their own in student decides *.hathAt 4 he is ready to attacking the modules. Finally,the pro- say what is vot ,eing approximated. fessors teaching the various sections of Upon entering his Guess, he is either told the calculus course change from semester it is not good enough and to look at more to semester, and there is a clear corre- values, or that it is right, or that it is lation between the enthusiasm of the pro- close enough, but that a better value fessor for the computer laboratory and the would be ,Once the .student has success- students' participation. fully named the root, the computer offers Student reaction was checked each him the ,..pportunity to enter a polynomial semester by a standard questionnaire dis- of his choice by entering, successively. tributed in class and in the computer the degree and the coefficients. center. It was originally supposed that The program then runs through the same a student would fill out one of these

1 L Mathematics 177

questionnaires for each module thathe (One fourth of the participating students used, but we soon decideC that we should ran five or more modules). One is tempted be content if we could elicit one overall to draw the conclusion that the students reaction from each participating student. who used the modules most are the hard- The results of these questionnaires are working students; not necessarily highly shown in Table 2. The reaction is clearly intelligent, but nonetheless successful. favorable, being most favorable on the To conclude this discussion of the issues of "interesting" and "would recom- evaluation of the project's results, we mend it." (Responses here were made on a would like again to recall that the ob- scale from 1 to 5, 1 in full agreement jectives of these CAI modules and the with the bottom description and 5 with basic philosophy used in constructing that of the top.) The form used also in- them are to overcome certain experimental cluded more direct questions over problems and analytical deficiencies in the stu- encountered in using the modules, and some dents' preparation (as well as to offer a of the students' comments led to correc- challenge to the more highly motivated tions and minor improvements. students), and hence they are not directed The actual effect of the CAI on stu- at the routine type of problem solving dentperformance and general understanding that is prevalent on examinations. It is obviously the mostdifficult to measure was, therefore, never expected that using of the three parameters, but perhaps the the modules would be directly and imme- most important. We have made two formal diately reflected in better test scores, attempts at measuring this. The first of but rather in a deeper, long-range appre- these was in the second semester of 1977-78, ciation of fundamental concepts and a when we set up non-participating control more enquiring approach to attacking new sections for comparison of results. Due material. Any correlation obtained be- to a plague of difficulties, many relating tween exam performance in the calculus to the problems which led to giving up the course and module use is therefore much Coursewriting Facility, it became impos- more likely due to pre-existing attitudes sible to gather any reasonable data. The of the student than to the contribution second attempt was made at the end of the the calculus CAI made to his or her knowl- first semester of 1978-79. At that tir edge. The calculus is neither a bag of a study was made comparing certain in- technical tricks, as itis frequently dices of calculus students who had actually taught to non-science majors, nor an ex- lama some of modules. These indices were ercise to mathematical logic, as it often their SAT scores, both achievement and appears in honors courses. It is a co- aptitude, their final grade in calculus, herent system of ideas elaborated for the and the number of modules used. In this modeling and analysis of certain clavier:sof small sample, no actual statistical cor- phenomena; experience with some of these relation could be established, but cer- phenomena and with the way calculus pur- tain tendencies were apparent. Porexample ports to capture these processes is neces- (see Table 3), all the students who used sary for anyone aspiring to scientific five or more modules obtained a grade of investigation or teaching. We hope that C or better in the course, while of the our modules make a start in the direction students who used five or moremodules, of providing some of, this experience. only one had obtained better than 700 on the SAT math achievement. DIFFICULTIES AND WARNINGS The creation and implementation of Table 3. Comparison of Module Usage successful computer-assisted Instruction Course Grade, and Previous Aptitude is a major undertaking involving many and Achievement Test Results 0 factors, any one of which is capable of Pail Semester 1978-79 nullifying many months of serious effort. 0 Our project, which we consider to have 274 4 ot-acio (Figores in percentof been reasonably successful overall, did partLcipating students.) 1108 8 01-700t not escape from its share of mistakes and 11 15 11 0-600X frustrations, and it still faces the difficulties of maintenance and imbrovement. 0A 4 27 15 11 4 44" The initial challenge is the selection of e 8 4 8 4 11 4 an adequate system of hardware and soft- I ware, unless one is locked into a pre- c 4 4 11---0...8----4 existing system. The choices today are o u 4 8 4 111 4 far wider than they were three years ago,

11 4 111 1 Number and serious consideration must be given to Aetitede 0400 601-700 701-400 1-23-4 5-67-8. microcomputer systems as well as to minis bcores doulfes and main frames. Although our mini

1 S 1Imammelmvir. 178 NECC 1980

computer, the Hewlett-Packard 2000 System, improper scheduling of student usage, poor has functioned well in our environment, response time when the computer is busy, its upkeep may be too great for a smaller and,inevitably, equipment failure. While institution, or capacity too limited for it is idle to suppose that one can avoid a more ambitious project.The lack of all of the problems and delays that typ- special symbols on our terminals (H-P ically arise, careful consideration can 2640B), such as integration signs or even avoid many of the difficulties. After lower-case letters, puts undue strains on all, you can not change your computer sys- programmers and users alike. The mathe- tem with the same frequency and ease with matical symbols as they appear in texts which you have been changing vourcalculus and are used by the teacher on the board text over the years!We hope that as the are important for a quick understanding advice and experience of a large number of the material and for adequate rein- of generous and helpful people helped forcement. For example, INT X2 DX needs guide us through the vagaries of this much more attention than project, so some of our experience may does the usual expression,jx2dx. Related contribute to future endeavors of this to this point is the ability of the ter- nature:- minal to recall previous pages of material (terminal memory), or careful programming REFERENCES to imitate this ability. Although all our 1- M.J. Christensen, MATHDOC; A Control terminals now have memory, at the outset System fox Computer-Assisted Calculus, of the project they did not, and we found Proceeding of Conference on Computers this situation much less flexible and more in Undergraduate Curricula, Denver, demanding. June 1978. Problems with software can be more 2- W.S. Dorn, G.G. Bitter, D.L. Hector, insidious. It is not easy to find out Computer Applications for Calculus, from salesmen exactly what a computer or Boston; Prindle, Weber and Schmidt, a software package is capable of doing, Inc., 1972. particularly in the area of educational 3- G. McCarty, Calculator Calculus, Palo applications. We began, as mentioned Altos Page-Ficklin Publishing Co., above, with iBM's Coursewriter which is 1975. an especially designed systemfor CAI 4- R.G. Selsby and M. G6mez Rodriquez, and its management. We soon discovered, "On- the -Job Training of Students in however, that it is incapable of doing Computer Science," Proceedings of any kind of calculations other than in- Minority institution Curriculum Ex- teger arithmetic. This can be overcome change Conference, Concord, North by begging, buying, or writing assembler- Carolina, January 1979. - language functions to increase the power 5- D.A. Smith, Interfaces Calculus and of Coursewriter, but this solution we the Computer, Roston: Houghton Mif- found to be painful and restrictive. H-P, flin Company, 1976. on the other hand, allows these functions 6- K.D. Stroyan, Manual for a Computation to be added to their Coursewriting Fa- Laboratory in Infinitesimal Calculus cility via simple BASIC programs. To a and Linear Algebra, Mathematical large extent this is what sold us on the Sciences, University of Iowa, 1976. system we bought. Several months later we gave up trying with the Coursewriting Facility and converted to BASIC. Com- plications in the compiling procedure, bugs in the management facility, and system crashes occasioned, at best guess, by the many needed disk accesses finally wore down the patience of even the most stalwart among us. Of course, BASiC is an entirely adequate language for most CA/ ( consider its transfer.- ability, by far the majority of the available cAi packages are in BASIC), but it does not encourage the same kind of careful organization and answer processing as do Coursewriter and other CAI languages. Some of the other components that can, and hence will, go wrong are: Inade- quate physical plant, too few terminals, Mathematics 179

COMPUTER APPLICATIONS IN A FINITE MATHEMATICS COURSE by R. Abernethy, G. Piegari, and A.L. Thorsen Mathematics Department Virginia Military Institute 703-463-6335

INTRODUCTION the final grade in the course). No comput- The course described in this paper super work was included on hour tests or the plements a standard course in finite math- final examination. ematics with computer applications. It The computer assignments were executed was developed with the following goals in on an HP2000 computer through 8 time-shar- mind. First, since the course is offered ing terminals in three locations on campus. to students in engineering or mathematics However, a group this size could easily cure^ula, it was the authors' hope that have been accommodated with as few as three the computer applications would increase terminals (perhaps fewer with proper sched- student interest and motivation as well as uling). To provide help on the computer head off the common complaints about the assignments, each of the three instructors irrelevancy of abstract mathematics was available for a one-hour scheduled courses. Second, it was felt that the help-session each week. Each session was computer assignments would lead to an in- open to all students in the three sections crease in the student's proficiency in and was conducted in a computer terminal programming, which would prove benefi- room where five terminals were available. cial in his later mathematics and engi- One restriction on the course, placed neering courses. Finally, it was hoped by our department, was that none of the that the student's grasp of certain top- core mathematics content of the course be ics, in particular probability, would be deleted or diluted in order to accommodate enhanced by the computer assignments. the computer work.We believe that we sat- Three sections of the course were isfied this restriction, even though cer- taught to a total of 51 students in the tain introductory topics from chapters 1 spring semester of their freshman year. and 2 of the course text (1) were omitted Each of the students had received, in the from the syllabus to provide the additional fall semester, at least 4 hours (contact time required for the computer assignments. hours, not credit or semester hours) of The time gained proved more than adequate, instruction in BASIC. However, most were and in addition to the computer work we far from being competent programmers, and were able to give a reasonable treatment of the authors are convinced that our course linear programming, which was not in the could be made self-contained by beginning original course syllabus.The major topics it with 4 or 5 lectures on BASIC. We in the course were set theory, discrete started our course with a diagnostic test probability (including Markov Chains and on BASIC and followed it with a one-day stochastic processes), matrix algebra, sys- review of BASIC. We resolved to discuss tems of linear equations (including a programming only as specific problems a- Leontief model in economics), and linear rose in connection with computer assign- programming. The text material was chapters ments and spent no more than an average 3,4,5, and 6 (1), but there are numerous of 15 minutes class time per week on textbooks which offer similar coverage. programming. Eight computer assignments, The course was evaluated in two ways. timed to correspond to the material being A student evaluation form was completed by covered in class, were given. Each aseach student at the end of the course, signment was to be completed by its as- which provided some insight into whether signed deadline, which was between one the goal of increased interest and motiva- and two weeks. These assignments, in to- tion had been achieved. More formally, we tal, were counted as an hour test (15% of compared the performance of the students in 180 NECC 1980

the computer-supplemented course to that of correct probabilities of the various out- students in the standard course.A lame comes. Varying the number of trials was part of the final examination which was intended to convince the student that the given in our course had been given to sev- larger the number of trials, the closer eral earlier sections and one concurrent the approximating probabilities reflect section of the unsupplemented finite math- the true probabilities obtained by analyt- ematics course. To test our hope that ical methods in the classroom. comprehension of some of the mathematical The third computer program was a simu- topics would be enhanced by the computer lation of the classical Buffon's Needle work, we compared the scores of the stu- Problem. It is both easy enough to simu- dents in the standard course with the late on the computer and difficult enough scores of the students in the computer- to defy solution by the analytical methods supplemented course for the relevant sec- studied in finite mathematics (although it tions of the exam. Results of these evalu- can be solved by integral calculus).The ations are discussed in section four of student is thus introduced to the main this paper, following a discussion of some value of simulation--solving problems that of the computer assignments in sections two cannot be solved by other means. In and three. Buffon's problem, a needle one unit in length is dropped on an infinite grid of COMPUTER ASSIGNMENTS IN PROBABILITY vertical lines. The probability that the The first five of the eight computer as- needle touches a line is the desired quan- signments were related primarily to proba- tity. For our experiment the lines were bility theory. Pour of these five were two units apart. In order to accurately simulations of probability experiments. simulate the experiment two independent One assignment was a program to compute random numbers must be generated, one to combinations and permutations and use obtain the angle the needle makes with a these results, in turn, to compute the prob- horizontal line and one to obtain the lo- abilities of certain events. cation of the center of the needle between The first program was simulation of one two lines. The test condition to determine hundred tosses, one thousand tosses, and a hit or miss amounted to comparing the then ten thousand tosses of a fair coin. projection of the needle on the horizontal It was in this assignment that the stu- line to the distances of the center of the dents were introduced to the random number needle from the adjacent vertical lines generator and its vital importance in sim- (5). The student was required to compute ulating experiments. Also, we indicated the approximate probability that the nee- the need for a test condition to translate dle hits a line for 100, then 10,000 simu- the value of the random number generated lations of the experiment. Incidentally, into an outcome for the experiment.For since the exact probability is 1/11, we example, since the random number generated asked the student to compute the recipro- is greater than or equal to zero and less cal of his approximate probability to see than one, declaring the outcome as heads if he would recognize this answer as an if the number is less than 0.5 is a simple approximation of t. and reasonable way to simulate the tossing Assignment four involved programming of a fair coin. algorithms to compute C(n,r) (the number The second assignment used the random of combinations of n objects taken r at a number generator introduced in assignment time) and P(n,r) (the number of permuta- one for a more complicated simulation. tions of n objects taken r at a time) and The experiment was rolling a pair of un- using these algorithms to compute proba- biased dice. The student was reminded that bilities, for example the probability of the outcome of one die is independent of getting four aces in a seven-card poker the outcome of the other, and the outcomes hand. One by-product of this assignment 2 through 12 must be generated in conformi- is the lesson that computers have practi- ty with their natural frequencies.The use cal limitations. The largest factorial of the integer function INT was introduced. that could be computed on our computer was Thc, outcome of one die can be simulated by 331. Therefore, the student was forced to generating a random number, multiplying it develop a more sophisticated algorithm to by six and adding one, and then taking the compute C(48,3)/C(52,7) (the probability greatest integer (INT) of this result. of four aces in a seven-card hand) than The outcome of each die is generated inde- simply computing each of the factorials pendently, and the two numbers are added first and then performing the division. together to obtain the outcome of the ex- The fifth assignment had three parts: periment. Assignment two consisted of (1) the simulation of 3000 World Series running 100, 1000, and then 10,000 trials between two evenly matched baseball teams; of the experiment and approximating the (2) the simulation of 3000 World Series Mathematics 181

in which one team had a .6 probability of Hence a quantity, which may actually be winning any single game; and (3) the simu- equal to zero, may be calculated by the lation of 3000 World Series if it were computer to be, for example 0.0000021. changed to a best of nine game series, Thus, in the subroutine to test for zero rather than r best of seven. In the first entries in BRAn, the students were forced part, where the teams are evenly matched, to structure their conditional statements the length of each series was tabulated, to in such a way as to ignore any difference illustrate the non-intuitive fact that a in compaAd quantities which are less than six-game world series is as likely as a the suspected round-off error introduced seven-game series. Since the probabilities by the computer. In addition to a loop to of having a six-game series and a seven- calculate An, assignment seven also re- games series are equal, about half the quired a double loop (one for the row in- class can be expected to conclude that a dex and one for the column index) to deter- save: -game series is most probable and half nine if An had any zero entries. The tran- that a six-game series is most prothle. sition matrices in the assignment were 3x3 It should be noted that a slight deviation matrices. Thus the students knew thgt away from exactly evenly matched teams their program need only test for makes a six-game series the most likely. zeroes because in class we gave the theorem The second part of the fifth assignment was stating that if an nxn transition matrix intended to illustrate the fact that if one is regular, then at least one of the first of the teams has an advantage over t...) other (n-1)4+1 powers of A contains no zero team in each of the games, then itz advan- entries. tage ie magnified when the trials are For the first part of the final computer grouped into a contest. The probability assignment the students were required to that a team with a .6 probability of win- write a program to solve two problems based ning each game wins the series is about .7. On our classroom discussion of Leontief The third part of the assignment illustrat- models. This prugram involved solving a ed that as the number of trials grouped in system of linear equations using the INV a contest are increased, the initial advan- function to find the inverse of the coef- tage is further magnified, in this cane ficient matrix. from about .7 to abut .73.Although each It has been noted that one of our ob- of these parts can be solved analytically, jectives was to increase the student's the solutions are complicated and usually proficiency in programming. Toward this beyond the objectives of a finite mathe- end, each of the above computer assignments matics course; so, the student is again required the students to actually write a impressed with the primary value of simu- program- -not merely supply data to a pro- lation. (Similar problems are ..iscussed in grew called from the system library. How- Kemeny (2), pages 143-144 and section 1.5 eVer, the authors felt that students should and Kemeny (3), pages 161-167.) be given an opportunity to realize the importance of the system library; hence, they COMPUTER ASSIGNMENTS USING MATRICES introduced the system library in the second Assignments six through eight were de- mitt of assignment eight.The problem signed to introduce the students to the ma- supplemented our class work on linear trix (MAT) commands in BASIC and to the sys- programming and the simplex method. In the tem library. Assignment six was a set a! assignment the students were asked to max- routine exercises that required the use of imize an objective function subject to five the commands for the matrix operations of constraints using a program (LINPROfrom addition, multiplication, scalar multipicce- our system library.The instructors felt tion, multiplicative inversion, and trans- that the introduction to this powerful fac- position. Additionally, the students were et of computer use was an appropriate end given a 2x2 matrix B and asked to use a loop to ;.he computer segment of the course. to determine B4. Assignment seven paralleled our class EVALUATION OP THE COURSE discussion of Markov Chains, and its purpose The results of the student evaluation was to introduce the students to subroutines, given at the end of the course were general- round-off error, and a double loop. For the ly quite positive. Most students sewed not project, the students were given two 3x3 only to enjoy the course, but to feel that transition matrices and asked to write a pro- it was useful. For example, a question on gram to test each matrix for regularity. A the overall rating of the course had an transition matrix A is regular if An has all average of 8.3 on a 0-to-10 scale. The positive entries for some nonnegative integer average response to the question; "Would n. The students were asked to recall that in you recommend this course to others?" performing calculations with a computer, (again on a 0-te-10 s ^ale , where 0 = not round-off error is almost always present. recommend at all, and 10 = highly recommend) 182 NECC 1980

was 8.1. in addition, many responses in were exposed to each of the three instruc- the comment section of the evaluation form tors, while those in group R were taught indicated that the course was well-received. by only one of these instructors. in ad- A statistical evaluation of the course dition, the class size for group C averaged was done by comparing scores of students in 17 students, while the average class size the regular course (group R) on 21 problems of group R was 24 students. However, the common to both final exams.We proposed authors believe that the introduction of average scores on the mathematics portion the computer into the course was an impor- of the SAT as a possible measure of how tant positive factor contributing to the groups C and R compared in ability. Imme- significant difference in performance noted diately a problem arose because the two above. groups were significantly different when Finally, we remark that when the stu- this variable was measured. Two comparable dents with Math SAT scores of over 600 groups were acquired by deleting all stu- were compared, the students from group C dents with Math SAT scores over 600 in each again performed better on the 21 exam group of students. We were left with 24 questions than the students from group R. students to group C with an average Math However, the difference was not as gag- SAT score of 531, and 29 students in group nificant. This result is consistent with R with an average Math SAT score of 533. the authors' feeling in the beginning of Students in both groups C and R were engi- the project that the better students were neering and mathematics majors. likely to perform well no matter how the The results of the 21 examination ques- course was designed and that the greatest tions are given below. effect of the computer supplements would Grou C Grou be on the average students. N er ost ents 29 SUMMARY AND RECOMMEHDATiONS Average Math SAT 531 533 it was our intention to add a computer Total number of 504 609 supplement to a course in finite mathe- questions matics without compromising the traditional approach to the mathematical content in the Number of correct 330 365 course. In short, we did not want our stu- responses dents to become computer experts while Sample success PC r .655 .599 sacrificing their ability to solve problems probability with paper and pencil. it is obvious, however, that one cannot introduce a computer supplement into a finite mathematics course From the table above we get the difference without making some adjustments. Since we of the sample proportions to be Pc-Ae.056. were restricted to the same 45 contact hours Let PC be the success probability for the as In the standard course, we had to eliminate certain introductory topics mentioned grow, C population and PR the success prob- at the beginning of this paper; however, we ability for the group R population. Test- did not reduce the amount of time devoted ing the null hypothesis Pc 4 PR against the to probability and matrices. We eliminated enough material to give us an additional alternate hypothesis Pc > PR, we found that eight contact hours, and we found this ad- the probability that fc fR > .056 is less ditional time sufficient for the discussion of computing in general and the computer than .028 (c.f. chap. 8,(4)). Therefore, assignments in particular. If, on the at the 5% level of significance, we rejected other hand, one wished to add a computer the zAll hypothesis and accepted the alter.- supplement without sacrificing any of the nate hypothesis that the performance of.stu- standard course material, the best approach dents in the computer-supplemented course would probably be to increase the contact would exceed that of students in the regular hours per week from 3 to 4. The extra hour course on the 21 exam questions. We point per week mould be more than adequate to out that the one-sided test was suggested by accommodate the computer supplement. the restriction that the coverage of the We also believe it is important for the central mathematics topics in the new course students and the teacher to maintain a be at least equal to the ooverage given in sense of perspective about the goals of the regular course. Consequently, it was this kind of course. it is too easy to expected that the value of Pc would be at drift unintentionally into a computer pro- least that of PR . gramming seminar to the detriment of the 1Aere may be other faotors which con- traditional theoretical and problem-solving tributed to the superior performance of aspects of a finite mathematics course. group C. For example, those in group C The student must be kept mindful of the 193 . Mathematics 183

main thrust of the course, even if, as is often the case, he is more interested in computers than in mathematics. Keeping the course properly balanced becomes a question of the instructor's resolve, but he can help himself (as we did) by keeping the computer part of the course out of the hour tests and final examination. Finally, we believe that when computer assignments are returned to the student, the student's program should be accompanied by the instructor's solution. The solution can be typed in capitals to resemble the actual print-out obtained at a teletype- writer terminal. Besides showing the student the correct answer, an instructor's solution provides the student with a valuable reference if he pursues computer pro - gramming after completion of the present course. Such a policy, of course, requires new or altered computer assignments for subsequent courses, but we consider the price small in relation to the benefits. We must admit, in summary, our satisfaction with the results of this course. It will likely be a permanent part of our curriculum, and we anticipate further refinements and additions to the computer assignments in future offerings of the course. To this end, we invite others who have had similar experiences with a computer-supplemented finite mathematics course to share their views with us. REFERENCES 1. Campbell, M.G. and Spencer, R.E., Finite Mathematics, New Yorks Macmillan; 9 2. Kemeny, J.G., et. al., Finite Mathematical Structures7-lagrewood Cliffs: PrenrainTal, 1959. 3. Kemeny, J.G. and Snell, J.L., Finite Markin, Chains, Princeton: Van Nostris,Ur 4. Mendenhall, W., Introduction to Probability and Stat fisFair,3ra edition, Belmont,Ca1-7Mrtias Wadsworth, 1971. 5. Mosteller, F., Fifty Challen in Problems in Probability, Read n , FIREWEVarais Addison- Wesley, 1965.

1 94 184 NECC 1980

A COMPUTER-ASSISTED COURSE IN BIOMATHEMATICS Pui-Kei Wong Mathematics Department Michigan State university East Lansing, Michigan 46824 517-353-6880

Applications of mathematics to prob- a one-quarter, four-credit course with lems in the physical sciences and engin- calculus as prerequisite.At Michigan eering are well known and have long been State we have a two-term, ten quarter- an integral part of the undergraduate cur- credit sequence in calculus for the bio- riculum of mathematics departments. touring logical and social sciences. A student the past two decades courses in numerical having completed this abbreviated calculus methods and elementary finite mathematics sequence is usually adequately prepared have also been added by many colleges and to enter MTH-481. No prior knowledge of universities, reflecting the growing im- computer programming is aesumed. portance of mathematics in economics, The major objective of this course is management. and social sciences as well to provide the student with a sound intro- as in modern technology.We need only duction to mathematical methods and deter- cite the contributions of such individuals ministic models in biology, especially in as Kenneth Arrow,' George Dantzig, Nasally the use of computer simulations to analyze Leaatiff. Paul Samuelson. and John model behavior. To this end a number of von Neumann in this connection. problems from biology are introduced and On the other hand, applications of studied in some depth. We start with the mathematics to the biological and life underlying biological description, hypo- sciences, though no less important, have thesize, and then derive a working mathe- not found their proper pla'ce in the under- matical model. Once constructed, the model graduate curriculum. It was J.B.S. Haldane can be manipulated and analyzed to reveal who observed in 1928 thatt "The permea- its possible behavior. Modifications on tion of biology by mathematics is only the basic model are made, and the system beginning, it will continue and (grow) is then studied again under various initial into a new branch of applied mathematics." conditions, external inputs, and pertur- In the decades since, we have witnessed a bations.All this can be done quickly and phenomenal growth in the applications of economically by computer simulations. mathematics to the agricultural, biolo- Instead of numerical solutions. it is often gical, and life sciences. beginning with more desirable that the long-term trend or the work of Haldane, Lotka, Volterra. qualitative behavior and stability proper- and Wright and onto that of Hodgkin and ties of the model be known. In this case Huxley, who shared the Nobel prize in computer graphics provide a particularly physiology or medicine in 1963. And most effective tool in the analysis of model recently, Allan Cormack and Godfrey dynamics. Hounefield were awarded the 1979 Nobel In adlition to conventional class - prize in physiology or medicine for their -room lectures and homework, students are pioneering work in applying mathematics required to spend time each week in the and computer technology in the x-ray image computer laboratory doing modeling and reconstruction technique called Computer- simulation exercises. These exercises use ized Axial Tomography (1). instructional modules written specifically Recognizing the growing impact of for the course, and they run on Tektronix mathematics in the life sciences, the 4051 graphics computing systems. These Mathematics Department at Michigan State stand-alone microcomputers are capable of University introduced five years ago a high resolution graphics (780 lines by new course MTH-461 titled "Selected 1324 pixels per line) and are especially Mathematical Ideas in Biology." This is well suited to our needs. A Tektronix

195 Mathematics185

-4631 hardcopy unit provides a record of the 24. Simple epidemic models. graphic output of any desired simulation 25. Cell cycle analysis. run. 26. Models in neurophysiology. We also use computer graphics in the 27. Fishery dynamics and marine food chains. classroom. Demonstrations are done using The interactive computer-assisted a Tektronix 4025 graphics terminal and an instructional modules used in the course Advent 1000A television projection system. are designed for the student with no The 4025 is a raster scan device with 480 prior programming experience, and they are x 640 resolution and provides a composite stored on data tapes.An orientation video output to the Advent. It is driven lecture on the use of the Tektronix graph- by a Tektronix 4051 computer. The Advent ics computing system is given at the start has a seven-foot diagonal screen and is of each term, explaining how the student suitable for viewing by a class of up to can access the system and use these forty students. Dynamics of a model can instructional tapes. These modules range be displayed on the screen before a class from tutorials on vectors and matrices and discussed very effectively this way. through population growth, fitting data to A Tektronix 4662 digital plotter is also logistic growth curve, differential equa- used to make high quality multicolored tions, and modeling using compartmental transparencies for use on ordinary over- analysis and the method of peeling. hied projectors. Students have found using interactive com- The mathematicg treated in MT8 -481 puter graphics particulty helpful and include matrix algebra, difference, and have later adapted and modified some of differential equations, and applications the programs written for this course far are drawn from various areas of biology. their own research use. Special emphasis is given to the qualita- To illustrate the use of computer sim- tive behavior and stability of nonlinear ulation in teaching mathematic modeling, equations. The course will typically cover let us consider the problem of growth of the first twenty-two items of the outline human populations.First the problem is below plus occasional substitution from analyzed by the very simple difference the remaining ones. equation

MTH 481 Syllabus (1) k 0, 1, 2, xk+1 Ark 1. Algebra of vectors and matrices. 2. Linear equations and determinants. Here ick is the population size or density 3. Eigenvectors and eigenvalues. at the kth census and A is the intrinsic 4. Methods of least squares. growth rate, which is defined as the dif- S. Discrete time single species population ference between the crude birth rateB models. and crude death rate D. In this case the Linear difference equations. 6. population is treated as a homogeneous 7. Depletion of nonrenewable resources. collection of individuals so age structure 8. Discrete time predator-prey models. and sexual differences are ignored. Equa- 9. Populations with age structure (Leslie tion 1 admits the closed form solution model). 10. Harvesting and exploitation of renew- xk = Akx0 k = able resources. (2) 11. Population genetics and difference wherex0is the initial population if A equations. 12. Asymptotic behavior and stability is constant. The future behavior of the (period three implies chaos). population is therefore completely deter- 13. Analysis of growth data. mined by the magnitude of A.However 14. Logistic, Gompertz, and other continu- real populations can rarely be described ous time 9opulation models. realistically by such a simple model in 15. Geometrical analysis of differential which the growth rate is constant over equations. long periods of time. 16. Models of photosynthesis. The first modification we make then 17. Linear differential equations and is to allow the intrinsic growth rate A systems. to be time- or population-dependent. 18. Compartmental analysis. Closed form solutions to Equation 1 are in 19. Lotka - Volterra and other models of general no longer possible, but the prob- competition and interaction. lem is easily handled by the computer. In 20. Volterra-Gauss principle. the Chi module for this course, actual 21. Theory of chemostat. census data is used to construct population 22. Stability and limit cycles. projections. Figure 1 shows several dif- 23. Enzyme kinetics. ferent functions used to fit the crude birth rate data for Costa Rica. The 1 u 186 NECC 1980

student is asked to select one of these, to include insects, trees, and other popu- and the computer will then calculate and lations that are more naturally grouped display the corresponding population trend into developmental stages or size classes for a specified number of years into the are also included at this stage. future (see Figure 2). At the end of each All the CA/ material was written in simulation run the student has the choice TEK-BASIC specifically for interactive of changing the parameters or entering an use and has been revised and tested over entirely different birth rate function and the past five years.Although much of it running the problem again. In this case was designed with teaching mathematical the same death rate function is used for modeling to biologists in mind, some of all the simulation runs. it can be and has been adapted and modi- At the next level of complexity, a fied for use in other courses as well. population is studied by separating it in- The equipment was acquired in part with to sex and age classes using the basic funds from the National Science Foundation model of Leslie (2). For human populations under an Instructional Scientific Equip- where the ratio of males to females is ment Grant. essentially constant over long time periods, it is customary to consider only the fe- REFERENal males and divide them into five-year age [1) L.A. Shepp and S.D. Xruskal, *Com- classes and take census every five years. puterized Tomography: The New The basic equation is still Equation 1, Medical X-ray Technology," American but xk is now a population vector with Mathematical monthly, 85 (1978), 420- 439. ncomponents, and Ais the projection [21 P.H. Leslie, "On the Use of Matrices matrix. in Certain Population Mathematics," Blometrika, 33 (1945), 183-212. f f f 1 2 f 3 n sl 0 0 . . 0

O s2 0 . 0 A

O 0 0 an_ Here fj is the average number of female offspring born to a member of the jth class every five years, and si is the probability that'a member of the jth class will survive and advance the (j + 1)st class at the end of five years. If we restrict our attention to females of ages O - 59, there will be twelve age classes andAwill be a 12 x 12matrix. if we assume constant fecundities and survival probabilities, the population for subsequent years can be calculated using Equation 2 as in the zimple scalar case. Figures 3 - 6 are the computer-generated population profiles for American females based on the 1964 data for fecundities f1,...,f12 end survival probabilities 81"*"1112' In the final stage the effects of population-dependent .A as well as harvesting ere added to the'Laslie model and studied. Figure 7 shows the menuselec- tion for this model; and Figures 8 and 9 are the output of two different simulation runs for a hypothetical population having three age classes. Extensions of the model solletuetWIYI £81

'10Til sonyo HU= 31Y$ VOA Y1800 05108-101004/M IMAM 31Y00 'ON AO ON1VZO aid 0001 NOX/V111d0d MOW= 1N3S3Sd3d mem MO 3NM1 1 SC 18Y31 ma-lams 114 01 MO ld 8d t ed 3UY 33UN1 MAXON 1YZ1N3N0dX3 Slid NOXIM N07.103POdd 3ASOO Od nu *IVA 01. ,a$n .41) 'Id 84 $0 . Q . 1 % \ 41th s,s 's s Od -....*:"..0...... f=p40.001000,0400=1itimloor. \ r ,.,... _ ...... id

1 % S Sw.SSS

0001 OLO1 0001 0106 0006

S I 188 NECC 1900

Pig. 2 POPULATION or COSTA RICA, I SWUM BASED ON PI PROJECTION RIGHT VERTICAL SCALE POPULATION IN MILLIONS CSOLIO CURVE) LEFT VERTICAL SCALE RN ES PER IOW POPULATION DIAMONDS REPRESENT ACTUAL CENSUS DATA

e OM

.,s `44z .e* SO ...' MI 0'601014 RATE . 117RTS4 RATE o'de 414, . s4...,, ...... diPoob wmi OPULATZON SS 2 .1 . DEATH RATE 4.1400 , 010 107e 7. .."6"."../INIA4.04401.1/401/0101/..04.41.20200 L 0.11..4.

19.9

...... ^....,.-..-1.-...... ,-.+ -...... -.-.....- - . Im. Mithon008 110

AMERICAN FEMALES , AOC* MOO POPULATZGN PROI'ZLI AT T 1004

12 $$ Is 0 e 7 e 6 4 2 $

$ 0 0 I 2 $5 HORIZONTAL SCALE I'M PeRCeNTO FZO . $

AMERICAN rEmAms, AOE* 40-60 POPULATION PRcruE AT 1 i 000

$2 $$ I* s 7 e 4 a

9 0 a 12 16 HORIZONTAL SCALE IN PERCENT, FZO. 4

200 190 NECC19130

AMERICAN FEMALES, ACES 8.-68 POPULATION PROFILE AT T 2014

..r./1.104/..,..1 * 0 0 12 16 HORIZONTAL SCALE IN PERCENTS FIG. 6 AMERICAN FEMALES, AGES 8-68 POPULATION PROFILE AT T 2839

H2' 10 I

0 1 a_7 r. Ll 4 I S I i s 1

0 111 12 ite HORIZONTAL SCALE IN PERCENTS MS. 0

2u Mathematics 191

Pig. 7

sow MENU *ow

t. CONSTANT FECUNDITY AND SURVIVAL PROBABILITY

2. POPULATION DEPENDENT FECUNDITY, CONSTANT SURVIVAL PROBABILITY

3. .CONSTANT FECUNDITY, POPULATION DEPENDENT SURVIVAL PROBABILITY

4. POPULATION DEPENDENT FECUNDITY AND SURVIVAL PROBABILITY

S. HARVESTING

SELECT BY NUMBER THE ITEM YOU WANT: 3

202 I 92 *CC 1990

FEC., POPDEPENDENT SURV. PROS. MUM MODEL, COM. 299 D2 00 2.6 FNDCW) 1/04EXPCW/D1D2)), Di 0.5 P2 01 12 St 0.893933333 32 F1 se 9 11 3149)' 8 WO) 4 ICS) a 1 PLOT OF TOTAL POPULATIONWCK)

70 89 19 20 05 40 59 80 no. s

.2(13 Mathematics 193

LESLIE'S MODEL: coNsar. FEC., POP. DEPENDENT WRY. PROD. FNDCW) m 1/C1.EXPCW/01-02)),DI on 180 02 m S F1 so 9 F2 m 12 $1 - 9.93333413$33 $2 m 0.5 XCS) m 6 WO) so 4 ZOO - 1 WOO - 11 PLOT OF TOTAL POFULATION WCK)

alm

1999 .., 880 I 998 790 \t 999 500 480

see -I- MO 189 \iIi

a a 19 28 30 40 SO 90 79 80 FIG. 9

204 194 NECC 1980

COMPUTER SYMBOLIC MATH David R. Stoutemyer Electrical Engineering Department University of Hawaii at Manoa Honolulu, Hawaii96822 (808) 948-8196

Most current calculators and mathemat- 80*X*Y4-4 112*Y+4 + 480*X*Yi3 ical computer programs are oriented toward -672*Y443 + 1080*X*Yi2 1512*Y+2 approximate numerical computations using +1080*X*Y- 1512*Y + 405*X - 567 an occult arithmetic called chopped non- Next, to solve the equation decimal fixed-precision floating-point. If this revelation causes discomfort, con- 2 (23-42s) =x3- a2x for x, the user types sider what Anston Householder, dean of SOLVE (2* (X+3-M2*X ) x+3x+2*X, X) numerical analysts, said "I don't like and the interactive response is the to fly in airplanes, knowing they are de- solution set signed with floating-point arithmetic." This arithmetic has undeniable advan- 00.0, tages, but it also has decidedly bizarre XBIA, implications which are difficult to X = -A). convey even to graduate numerical analysis Next, to invert the matrix students.More important, it is not the arithmetic that any of us uses for manual computation. 0 ql, l0 the user types THE NATURE et COMPUTER SYMBOLIC MATH The good news is that computers also ([1,19 can do exact rational arithmetic.More- (0,Q]) -1; over, even small personal microcomputers and the response is can do the nonnumeric symbolic operations of algebra, trigonometry, analytic geo- (11, PAN metry, and calculus. In fact, use of such (0, 1/01). computers for these purposes does not even Then, to simplify the trigonometric- require a knowledge of computer program- logarithmic expression (tan a)cos a)+ ming, other than the widespread custom of 1/csc(a)+1D(e2y) 2 In a, the user using "a" to denote multiplication and types "+" to denote raising to a power, in order to unambiguously specify expressions using TAN(A)*COS(A) + 1/CSC(A) + LN(X4.2*Y) a one-dimensional format compatible with -2*LN(X); the limitations of standard keyboards and yielding the response displays. For example, here is a sample,_ interactive dialogue using the muMATH-79 T'm 2*S1141(A)+ LN(Y). symbolic math program on an inexpensive To sum the series Radio Shack ?RS-Seta computer, which is increasingly popular in schools: 2 j In order to expand the expression (a j + b ) (5x- 7)(2y +3)4 the user types the line J1 (5*X-7)*(2*Y+3)+4; the user types and a few seconds later the interactive SUM(A*.7+2+1344, J, 1, N) response is to get the response

205- Mathematics 195

(2*A*Nt3+3*A*N+2+A*N)/6 for convergence, or determine (B1(N+1)-B)/(11-1). the first few terms of a Taylor (The arguments of the sum function are series. respectively the summand, summation index, lower limit, and upper limit.) Finally, to EDUCATIONAL USES evaluate the integral Computer symbolic math has long been available to applied mathematicians who jlax2 x sin x21dx , have generous computing allowances on the the user types largest computers, but this tool has only recently become available on the inexpen- INT(A*Xt2 X*COS(Xt2), X) sive small computers most often available yielding to kindergarten through sophomore calculus classes.How can this increasing] A*Xt3/3 + SIN(142)/2. available tool support math education? In contrast, traditional programming 1. A symbolic math system can be used languages such as APL, BASIC, FORTRAN or bx a computer-aided instruction PASCAL provide built-in math facilities prbgram to provide far more flex- essentially only for limited-precision ible, intelligent, aild responsive arCtIonstia. automatic drill or examination The above examples have illustrated than is otherwise possible.For symbolic math capabilities relevant to example, symbolic math systems are grades 0 through 14. However, computer able to recognize the equivalence symbolic math also features exact rational of a wide variety of mathemat- arithmetic, which is more suitable than ically equivalent expressions, and floating -point for supporting math educa- a computer-aided math instruction tion at the kindergarten through sixth program can adaptively take alter- grades. For example, continuing the above nate courses of action depending dialogue, to simplify the expression upon the users' performance. Such 3030/12/401 the user types programs free teachers to do what 30+30 * 121(1/2)/401 they do best--provide warmth, understanding, individual high- and the response a few seconds later is level guidance, and assistance for (253410016192626953125/ unanticipated difficulties. 2. The ability to do numerous large 502114206731006316270912)*(3)+(1/2) examples encourages creative ex- There are none of the roadoff, taufile1444 plorations which can reveal pat- oroverflowproblems that beset traditional terns and thus suggest general mathematical programming languages. theorems.Students* inductive Although none of the above examples generalization powers can be exer- requires a knowledge of computer program- cised and developed to a far ming other than the convention of using greater degree than is otherwise *** for multiplication and "V* for possible. raising to a power, curiosity of needs 3. Built-in two facilities can allow not net by the built-in facilities causes students to see each step of a someusers to seek deeper involvement computation, rather than merely with the mathematical and programming the final result. techniques used to implement these sym- 4. A demonstration that an operation bolic math systems. Accordingly, sym- can be done automatically by com- bolic math systems generally provide a puter can encourage average and programming language for writing exten- poor students that the flashes of sions. Examples of slch extensions are: inspiration given only to brilliant 1. introducing new functions (such as students are unnecessary for that hyperbolic functions) and their operation. Hope for plodders is simplification properties; revealed. 5. Students who are more enthusiastic 2. extending the class of expressions which can be factored, differen- about computers than about math tiated, or otherwise operated upon; are provided with a new avenue for appreciation of math. Computer 3. extending the class of equations which can be solved; symbolic math vastly enhances the opportunity for beneficial 4. automating the sequence of steps mutual reinforcement and cross- necessary to do an inductive proof, determine a limit, test a series motivation between math and computers. 196 NECC 1980

6. Inspection of the underlying com- from 32,000 to 64,000 bytes of memory, puter symbolic math implementation depending on how many of the various math- programs can help students learn ecatical facilities are loaded simulta- the methods for accomplishing the r =sly. For information, write Micro- operations manually. ...ft, at 10800 N.E. 8th, Suite 819, 7. Programming extensions to the Bellvue, Washington 98004. built-in operations can reinforce FORMAC is available for medium to understanding of both the built-in large IBM 360 and 370 computers which have and new operations. In fact, an at least 140,000 bytes of memory.For instructor can withhold portions information, write Knut Bahr at GMD/IFV, of the algebra system implementa- D-6100, Darmstadt, Germany. tion and challenge the students REDUCE is currently available for IBM to implement them. 360war110, DEC POP 10 and 20, Univac 1100 series, CDC Cyber series, and AVAILABILITY Burroughs 6700 computers. REDUCE requires For educational use, a symbolic math a minimum of from 300,000 to 400,000 bytes system should be interactive, general pur- depending on how many of the various math- pose, and available for a modest fee on ematical facilities are loaded simulta- computers typically available to students. neously. For information, write Professor In order of increasing computer memory Anthony C. Hearn at the Computer Science requirements, here are four symbolic Department, University of Utah, Salt Lake math packages which meet these City, Utah 84112. requirements: PICOMATH-80th is a set of three small LITERATURE symbolic math demonstration programs Most of the relevant literature is written in BASIC that should run on vir- oriented toward research rather than edu- tually any computer. For information, cation, but a good way to get started in write The Soft Warehouse at Box 11174, this area is to join the Association for Honolulu, Hawaii 96828. Computing Machinery Special Interest Group muMATH -79 is currently available for on Symbolic and Algebraic Manipulation. manY57Mvarious brands of personal TheirA6WSIGSAMBulletinis a timely source microcomputers based on the Intel 8080, of information on such topics as Intel 8085, and Zilog Z80 microprocessor meetings, abstracts, and new systems. chips. Computer memory can be measured For information about joining, write in units called bytes, and muMATH requires the ACM at 1133 Avenue of the Americas, New York, NY 10036. Invited Sessions

COMPUTER-BASED RESOURCE SHARING: DIVERSITY AND OPPORTUNITY

Chaired by Donna Davis Mebane and Rodney Mebane EDUCOM P.O. Box 364 Princeton, NJ 08540 (609) 921-7575

ABSTRACT Computer-based resource sharing occurs Inter-university Consortium for in many places, by many people, in many Political and Social Research (ICPSR) forms, for many reasons. A professor who Merit asks a colleague to evaluate the pedagogic New England Regional Computing Program quality of a CAI tutorial, a student user (NERComP) of SPSS, a demographer working with census North Carolina Educational Computing tapes, a librarian with access to OCLC -- Service (NCECS) all of these people are engaged in some Research Libraries Information Network form of computer-based resource sharing. (REIN) This special session will report on a Profiles of these organizations will be major EDUCOM research project, funded by presented, and the nature of services NSF, to examine the diversity of sharing offered will be explored in depth. activity that takes place within the Specific RSO activities will be cited to higher education and research community illustrate the diversity of organiza- and the opportunity that exists for tional response to such questions as increased cooperation. these: In describing, explaining, and evaluat- What kinds of resources are available ing the sharing phenomenon, two scenarios and where do they originate? are apparent. One is of relatively small- What options do developers have? scale activity, very informal and with How ate resources packaged and information exchanged primarily by word of distributed? mouth. Many are neither aware of what Who are the current users of these others are doing with computers nor of the organizations' and what services do possibilities for shared use. The other they find most valuable? scenario is of an active attempt by What are the technical, economic, and various resource-sharing organizations organizational conditions of using (RS08) to promote the orderly exchange of shared resources? computing materials and experiences. How do the sharing organizations them- Focusing on the latter scenario, this selves adapt to a rapidly changing session will begin with the benefits of environment? sharing and the various pathways to The primary objective is to make praoti- sharing. It will then specifically tioners and leaders in the field more address interinstitutional resource aware of organizational options available exchange and nine representative organ- to meet their own diverse computing needs. izations that provide a link between Materials will be distributed describ- resource developers and a resource users ing the study and the participating and effect the shared use of hardware, organizations, and representatives of the software, machine-readable data bases, nine organizations have been invited to and other computing systems. The RSOs apply for space in the exhibit area. participating in the EDUCOM study include the following nonprofit organizations: CONDUIT ZOOM Georgia Information Dissemination Center (GIDC) Health Information Network (HEN)

197 208 198 NECC 1980

COMPUTERS AND INSTRUCTION: DEVELOPMENT, DIRECTIONS, AND ALTERNATIVES

Chaired by William Grusner Addison-Wesley Publishing Co., Inc. Reading, MA 01867 (617) 944-3700

ABSTRACT PLATO: COMPUTERS AND INSTRUCTION, A Tlifrsession will examine the world of LARGE-SCALE SYSTEM computers in the instructional environment. Our three panelists have used micros, minis, Robert Hart and mainframes in various educational University of Illinois situations and will present papers as a basis for discussion and examination of ABSTRACT computers and instruction. The Language Learning,Laboratory at the During the session directions and alter- University of Illinois makes available an natives for new uses of computers will be 80 terminal PLATO site for humanities usage. explored and the existing systems examined. During the past seven years, the Language Participants and attendees are encouraged Learning Laboratory has supported PLATO to exchange information and ideas. It is materials development for French, German, hoped everyone involved will leave with Spanish, Russian, Swedish, Swahili, Hindi, fesh insight into the creative use c! Hebrew, Chinese, Japanese, Latin, Classical ..omputers for educational purposes. Civilizations, and E.S.L.The relative independence of development projects has led to a number of models for incorporating PLATO in classroom activity, ranging from CREATIVE LEARNING WITH COMPUTERS - -THEORIZE totally PLATO-centered to voluntary and OR PERISH supplementary usage. Research now in progress seems to reveal a relatively Margot Critchfield stable pattern across a wide variety of Pittsburgh, PA situations: (1) There are two overlapping but distinct ABSTRACT populations of users, one PLATO-receptive, *Thperience of the dyed-in-the-wool the other non-receptive. computer hobbyist provides educators one (2) Both instructor and student dissatis- point of reference for our warm enthusiasm faction with current materials centers on for the computer. Our observation of response analysis and feedbadk character- students who learn to program or who play istics, which are perceived to be relatively complex games on interactive microcomputers inaccurate, inflexible, and unable to deal provides another point of reference.What with meaning as opposed to fdrm. do these vastly different kinds of learners have in common? A theory of learning is needed to tie these disparate computer users together. Such a theory must be MICROCOMPUTERS IN ELEMENTARY AND SECONDARY applied to the design of educational EDUCATION: WHERE WE'VE BEEN, WHERE WE'RE experiences that include computers in order GOING. to ensure computersl.success and continued growth. Dan Isaacson This paper will attempt to clarify the University of Oregon definition of creative learning and to state some of the explicit values of formal ABSTRACT educatita as they relate to computer use. 4*=Bducational research says about the The possible contribution of some current use of computers in teaching, what's learning theories to a set of principles happening now, what the future holds, and for creative learning with computers will be what's holding us back will be discussed. discussed. C'9 Tutorial

VIDEODISC TUTORIAL

Bobby R. Brown and Joan Sustik Weeg Computing Center University of Iowa Iowa City, Iowa 52242 (319) 353-3170

ABSTRACT 'mss tutorial is for individuals who have little or no experience using a videodisc An instruction.The tutorial will consist of a presentation of the performance characteristics of videodisc as they relate to .tnstruction. A variety of applications c...-vnt and potential will be discussed. Various approaches to the development of materials and mastering of discs for videodisc applications will be presented. The tutorial will also include a demon - stration of an operational intelligent videodisc system. Handouts will be provided to the _participants.

199 210 Testing/Placement

MICROCOMPUTER-ASSISTED STUDY AND TESTING SYSTEM (MASTS)

Hugh Garraway The University of Texas at Austin EDB 43GB AlPtin, Texas 78712 (512) 471-4014

INTRODUCTION 1) Time and expense involved in custom Computer-managed instruction (CMI) has designing CMI systems for individual ap- proven to be an effective teaching/learn- plications. ing strategy. Studies using large com- 2) Difficulty with adapting existing puter-based instruction (CBI) systems such CMI programs to run on computer systems as PLATO (Nievergelt, Jurg and others, other than those for which they were de- 1978) and TICCIT (Reigeluth, 1978) have veloped. shown that in addition to facilitating 3) Knowledge of programming necessary learning, students enjoy CMI. The mili- for an instructor to create or modify a tary is one of the largest implementers program. of CMI. The Navy is using several TICCIT 4) High initial cost for setting up systems in a CMI application that is ef- hardware or adding to existing systems. fective in terms of both cost and in- 5) Communication expenses and problems, struction. The Air Force is using a de- and down time associated with time-sharing dicated CMI system, the Advanced In- systems: structional System (AIS), developed with Recommendations for improving the cost the McDonnell Douglas Corporation. effectiveness and efficiency of future The Teaching Information Processing CMI systems have included the use of System MIPS) and the Program for Learn- stand-alone microcomputer systems and ad- ins; in Accordance with Needs (PLAN) are vanced authoring languages that allow an two CMI systems available for use in instructor to specify or choose instruct- all levels of education. Both of these ional and testing strategies and enter systems are batch-oriented and are in- related content without having to learn tended to be used on mainframe compu- a complicated programming language. ters. TIPS and PLAN are designed to be used mainly with large numbers of stu- PROTECT OBJECTIVE dents. The objective of this project was to de- Successful medium- and small-scale sign, produce, implement, and evaluate a applications of CMI (Bork, 1977; Brock- CMI system with the following specifi- osier, 1977) have helped define some cations: common problems in using CMI including' 1) The system will be based on inex-

200 211 Testing/Placement 201

AUTHOR DISK STUDENTDISX

AUTHOR PROGRAM STUDENT PROGRAM

Creates files for assignment/test mod- Executes essignment/tests, updates ules, edits fi:es, maintains index. "STUDEX" file.

.00 401 mks mk. INDEX FILE Holds execution sequence.

Collects data from completed student disks. STUD= TILE

Holds student data, past scores, sequence pointer, and re-entry status.

Displays results from completed nodule. ASSIGNMENT/TEST TILES

Holds couplets text and data for 1 sod- uls in sock file.

?JOUR! 1

MASTS INSTRUCTIONAL ALGORITHM

REPEAT THROUGH LAST REMEDIAL NODULE X, N

YES I NEXT REME- DIAL OR SLOCX 4 SIX INSTRUCTOR

YES .111=111M OR. Vila .00 GIVE NEXT MAIN ASSIG% KENT (X+1) 0

SANE AS ABOVE THROUGH ALL MAIN AND RUMEDIAT. RDE14ENENT/TDst NODULES I

YEE

?IMRE 2

212 202 NECC 1980

pensive microcomputers with disk storage now interact with the authoring program ability. to create the CMI unit. A complete CMI 2) The system will be interactive. unit may consist of numerous assignments 3) The CMI program(s) will be con- and tests, but MASTS does not require tained on individual student disks so that that the complete unit be entered in a they may be used on any compatible micro- single authoring session.The smallest computer.' amount of information that may be enter- 4) An interactive, intelligent auth- ed during a session is one module con- oring program will allow an author to taining an assignment, the test based on create CMI modules without having to the assignment, and the data such as com- master a programming language. petency level for advancing, number of 5) The system will allow an author test questions to give, whether questions to gather student data from individual are to be given in random or fixed order, disks for storage on a single data and what remedial strategy is to be taken disk. upon the student's failure to reach the 6) The system will perform simple established competency level. statistical computations and display Each assignment/test module is given a group and individual results as raw two-number label X, Y, which determines score, mean, standard deviation, and its position within the instructional al- s-score. gorithm. X represents the main assign- 7) The collected data will be ment/test. Y represents a remedial as- stored in a format that can be easily signment/test for the main assignment- transferred to other statistical anal- test X. For example, the first assign- ysis packages. ment in a MASTS unit would have the label 1,0. I: the author-specified competency MASTS level were met for test 1,0, then assign- The result of this project is the Micro- ment 2,0 would be made. If the compe- computer-Assisted Study and Testing System tency level were not met, the MASTS would (MASTS). MASTS is contained on two disks execute the author-specified remedial (Figure 1). The author disk contains the strategy. The student could then repeat authoring program which creates the data assignment 1,0, be given a remedial as- and index files to be stored on the stu- signment Test 1,1, or be blocked from fur- dent disk. The student disk contains ther assignments pending a visit to the the student program which executes ac- instructor, at which time the block can cording to the data stored by the author- be removed from the student's disk. When ing program.Copies of the student disk a remedial test is passed, the student is are distributed to each student in a CMI advanced to the next main assignment., course. The progress for collecting and When a remedial test is failed, the op- manipulating student data from the indi- tions are the same as outlined for 1,0. vidual student disks are included on the Thus, a remedial assignment/test pre- author disk. The heart of MASTS is a scribed for failing 1,1, would be 1,2. two-dimensional variable instructional Passing 1,2 would advance the student to algorithm (Figure 21 with parameters and 2,0. If the last remedial test fora main branching or looping conditions set by assignment is failed MN, the student the authoring program. This algorithm is automatically blocked and told to see is assembled by the author program through the instructor.When a student has suc- interactions with the author.To assem- cessfully completed all of the main as- ble an algorithm, the author need only signments on a disk, he is instructed to be familiar with the options available turn in the disk to the instructor. Any in the algorithm model (Figure 2), as the number of main assignments (X) and any authoring program is menu-driven and all number of remedial assignments tY of X) options at any point in an authoring may be created within the file boundaries session are displayed on the monitor. of the microcomputer disk-operating system. AIOTHORIN0.SYSTEM To begin a MASTS authoring session', the An instructor wishing to use MASTS author places the authoedisk in the receives an author disk, a student disk, microcomputer and turns on the power.The and a short booklet outlining the options author program automatically loads, runs, available in the instructional algorithm and presents the author with a master for creating a CMI module. After read- menu.The author may select one of sev- ing the booklet, the instructor plans the eral options. If he selects to create assignments and tests and determines the new assignment/test modules, he must put criterion levels to allow the student to a student disk into a second disk drive, progress or receive remedial assignments. and is so informed by the author program. The instructor or clerical assistant may The author will then be asked to enter

21 3 Testing/Placement 203

the 2C,Y number of the new ;nodule. After on each student's disk. this information is entered, the program When the author finishes entering in- will prompt the author to enter the assign- formation fields, the author program will ment, the text of which can be any length. place this information on the student At the end of the assignment, the author disk, along with other index information enters the letter "S* on a new line to sig- making the student disk complete. In- nal the end of the assignment. At this structions are then given to make copies point, the author may instruct the proof the master student disk for distribu- gram to execute the test immediately af- tion, which necessitates the program re- ter the assignment or allow the student to turning to the microcomputer's disk-oper- leave the computer to complete the assign- ating system. ment. It is also possible to run a pro- grain specified by the author at this STUDENT DISKS point, which allows the integration of Each student is given a copy of the other CBI programs with MASTS. master student disk. To use his disk, a The program will now prompt the author student places it in any compatible mi- to enter the test.The test is construct- crocomputer and turns the microcomputer ed through interaction with a menu-drive on. The program automatically loads and subroutine which allows any number of starts at the proper point in the instruc- (mixed) multiple choice, true-false; or tional algorithm. On the first session, key word questions to be entered. The the program gathers the information menu allows the author to select a ques- fields entered by the author and the first tion type or end the test. The process assignment is given. Then according to for entering the text for all types of the instructions entered by the author: questions is similar to that for enter- 1) the program stops and will not contin- ing the assignments. A multiple choice ue until the computer is turned on again, question may have any number of choices thus allowing the student to remove his since they will be presented one at a disk and carry out the assignment before time to the student. (After all choices returning for a test; 2) the program pro- have been shown, they will be repeated ceeds directly to the ti it on the_ assign - until the student picks one.) The key sent, in which case, the assignment text word question allows the author to specs- could be used as an instructional frame fy a number of keywords or character rather than as an assignment as such; or strings to be considered a correct answer 3) the program chains directly to another to the question. In the student program, program which could be placed on the stu- a key word question asks the student to dent disk by the author. type in a response. This response will When a student finishes a test, he is be searched by the program to determine given instructions which may be the next if it contains any of the key words. If main assignment, a remedial assignment, it does, it will be counted as correct. the same assignment again, directions to After entering a test question, the pro- see the instructor (a block is placed on gram returns to the menu which allows re-entry until the instructor unlocks the other questions or the end of the test to disk), or instructions to turn the disk be specified. At the end of the test, the in, if the student has finished the last author is prompted to enter the competency assignment. All scores for each test are level in percent for passing to the next stored in the studex file which also holds main assignment. He is then prompted to the information on a student's location enter his choice for a remedial strategy in the algorithm and his re-entry status. for students who do not pass. When this information has been entered, the program MASTS DATA COLLECTION AND DISPLAY will return to the master menu. From. the When students have completed their master menu, the author may add other mod- MASTS disks, the disks are turned in to ules as outlined above or print a master the instructor for data retrieval. One student disk when all modules for the of the options on the authoring program MASTS unit have been entered. When the master menu is collecting data from disks. latter option is selected from the master Each student disk is placed in the second menu, the program will prompt the author disk drive and is read by the author pro- to enter the information fields he wishes gram aollect subroutine. The information to gather on each student's disk. The fields and scores for each test are stored author simply types a one-line instruction, with those for all other students. When such as "Please enter your last name," for all student disks have been read, the in- each information field he wants.Any num- structor may use the result display option ber of fields may be entered, and they will from the master menu. This option causes be presented before the student's first the computer to chain to a data display assignment. Tim responses will be stored program that computes and displays

2`r4 204 NECC 1980

statistical information for each test and REFERENCES each student. If a particular test has been repeated, the statistics are computed Baker, Frank B.Computer Managed Instruc- for each attempt. this information may be tior. Englewood Cliffs, N.J.: Educe displayed on the monitor, printed on a trail Technology Publications, Inc., line printer, or both. The student data 1978. is stored in standard ASCII text files so Bork, Alfred. "Course Management System that simple programs may be written to for Physics III." Proceedings of the transfer student data to other data base Conference on Computers in the Under- formats used by more sophisticated statis- graduate Curricula,1977, 213-222. tical analysis packages. Brockmeier, Richard. "Report on a Highly APPLICATION AND EVALUATION Used Computer Aid for the Professor in MASTS has been used for two semesters His Grade and Record Keeping Tasks." in an upper-level media course for educa- Proceedings of the Conference on Com- tion majors at the University of Texas at puters in the Undergraduate Curricula, Austin. Most of the class time for this 1977, 93-100. course is spent teaching production pro- Hilgendorf, Allen.A Study of the Trans- cesses while MASTS is used to assign out- portability and Effectiveness of the side readings describing media applica- Un-Stout CMIS and Individualized In- tions. In the past, assignments were made structional System Based Upon Learning with a reading list, but many students did Styles.Menomonie, Wisc.: Univ-Stout, not finish the readings. Another approach 1976. (ERIC Document Reproduction Ser- used was to require students to write re- vice No. ED 131 305). action papers or short reports based on the reading assignments, but this approach McDonnell Douglas Corp.Advanced Instruc- reduced the number of readings that could tional System. Brochure distributed be assigned during a course. Using MASTS at AECT Convention, 1979. for these assignments allows the instruc- Nievergelt, Jurg, and others. Alterna- tor to set due dates for completion that tive Delivery Systems for the Computer- coincide with related in-class activities. Assisted Instruction Study Management These first applications of MASTS have System (CAISMS). Urbana, Ill.: Illi- served as formative field evaluations help- noisUniversity, February 197R. (ERIC ing to define needed changes and to catch Document Reproduction Service No. ED bugs as only trial by fire can do. An 149 785). added benefit from using MASTS has been that the computer anxieties of students, Reigeluth, Charles M. TICCIT to the Fu- most of whom have never worked with a com- ture: Advances in Instructional Theo- ry for CAI. Salt Lake City, Utah: puter, have been lowered (according to pre- and post-MASTS questionnaires). Brigham Young University, 1978. (ERIC Document Reproduction Service No. ED MASTS HARDWARE 153 611). A Radio Shack TRS -80 with 48K memory Roll and Pasen. "CMI Produces Better and two disk drive units is used for auth- Learning in an Introductory Psychology oring and data collection from student Course." Proceedings on the Confer- disks. The student programs require only ence on Computers in the Undergraduate one disk drive and 32K memory. Curricula, 1977, 229-238. CONCLUSIONS Van Metre, Nicholas H. "Problems in Re- The definite student enthusiasm and searching Computer-Managed Instruct- lack of major problems encountered in im- ion." Paper presented at the American plementing MASTS indicate an open door fox Educational Research Association, To- exploring the use of small scale CMI. ronto, Canada, March 1978. MASTS or a similar system could form the heart of a programmed text or workbook. Another application might be in individual learning centers, using all forms of instructional materials as possible assignment media.

215 Testing/Placement205

RIBYT -- A DATA BASE SYSTEM FOR FORMAL TESTING AND SELF-ASSESSMENT

F. Paul Fuhs

School of Bussness Virginia Commonwealth University 1015 Floyd Ave. Richmond. Va. 23284 804-257-1737

ABSTRACT The logical associations among the This paper describes the function and questions within a pool present adiffi- structure of adata base system called cult administrative problem. Questions RIBYT,which simultaneously controls and must simultaneously be ordered in a number associates questions in question pools for of ways, they may be grouped into sets by many courses ofinstruction. The data course topics, by textbooks,by page num- base stores questions created by both fac- bers within textbooks, bytests in which ulty andstudents and is usedfor kormal the questions were used. and by many other testing and student self-assessment. The categories. Question pools are also dif- structure of the data base allowsrapid ficult to manage because they are dynamics retrieval of multiple typesof sets based questions have different life spans, sub- on predetermined logicalassociations. ject to such factors as over-exposure in The data base also provides forthe col- usage, changes in textbooks, and obsoles- lection and association of feedback cence due to technological changes. New information notonly on student perform- questions must continually be added to ance, but also on the quality of the ques- these pools and sachnew question forms tions in the data base. This quality is multiple associationswith previously assessed through students' subjective com- existing questions. In addition, periodic ments about the questions they receive and rewording and pruningof existing ques- through statistical item analysis. Feed- tions are required and, unless the ques- back isalso provided to studentson how tions in a poolare already well organ- they are able toimprove the quality of ized. it is difficult to detect synonymous their question input, The structure of questions. the data base is easily transferableto Many traditional file systems have manydata base systems which currently attempted tosolve these associational exist. problems. Yet, frequently thesesystems are expensive, overly complicated, and INTRODUCTION difficult to use. They lack the capabil- Many educators administer one or more ity to make multiple associations among pools of objective testquestionsas the the data items. basis for examinations. These pools usually arein the form of multiple-choice, DATA BASE TECHNOLOGY fill-in-the-blank, or column-matching Data base technology over the past ten questions. Educators have a need to gen- yearshas been successfully applied to erate, organize, store, .modify, and many storage and retrieval problems espe- retrieve such test questions'. cially where multiple associations exist among dataelements. Suchassociations are logicallyrepresented as relations 1Publiehing companies like Harcourt, (2), hierarchical structures (8), andsim- Brace, Jovanovich (7) and SRA (10)have ple or complex networks (9). They are recognizedthisneed and are willing to physically implemented through such supply such pools in certaincourses to methods ashashi0g, tables, andlists. faculty who adopt their texts. Their The data base approach has penetrated pools, however, lack associational struc- business and government and is in the tureandare merelysequential files process of replacing many traditional file organized by text page number.

21 206 NECC 1980

systems. DBMS arecurrently available generation because they understand the evenon minicomputerslike IBM's Sys- merits of the self-testing. tem/38, Hewlett Packard and DEC machines. To improve the quality of student However, educational institutions have questions, faculty, whilereviewing the been slow to use data base systems except questions, optionally enter into the data for some purely administrative functions. basecomments directed to the student authors of thequestions. Thisfeedback FUNCTIONS OF RIBYT is a learning experience in test construc- In this paper, we describe the function, but even more so in course content. tions and structure ofthe data base sys- There are two other feedback mechanisms to tem called RIBYT at Virginia Commonwealth improve the quality of the questionsin University. This system applies data base the data base. First, from an inspection technology tothe associational problems of patternsof studentresponses faculty of using and maintainingpools of ques- may detect acceptable alternative respon- tionsforformal testingandstudent ses that were not previously considered. self- assessment2. Thesingle data base Secondly, students can enter into the data contains multiple pools of questions, one base complaintsconcerning anyof 'the pool foreach course using RIBYT. The self-assessment questions they have questions are of various types, including received. Asa result faculty and stu- multiple-choice questions with single or dents critique the self-assessment ques- multiple correct answers; fill-in-the- tions. Questions for formal examinations blank questions with multiple and alterna- areselectedfrom questions which have tive correct answers; and column-matching already passed through these filters. questions. Each pool is under the direct Student responses to test questions control of the faculty members responsible are used not only for student performance for the course. In addition to faculty, evaluation, but also for statistical item students insome coursesare allowed to analysis. Eachresponse entered into the create and addquestions into thedata data baseis associated base. Student questions, however, are con- witha unique question number within the data sidered unedited until reviewedby a fac- base, a course testnumber, thequestion number ulty member. Such questions, while physi- within the test, cally intermingled with edited questions, the student making the response, and the student's score for the are logically distinct. During the review response. process, student questions are accepted, The RIBYT data base system associates modified, or deleted3. The question pools two coursetopics with each question in grow at a much more rapid rate when stu- the data base. Topicsassigned to ques- dents are allowedto enter questions. tionsby students are consideredtempo- With a largepool of questions and the rary, until reviewed forediting b7 fac- security provided by a data base system, ulty. Each course canorganize its faculty allow student self-testing ina permanent topics hierarchically and nonthreatening spirit similar to the independently of anyother course. There self-assessment procedures afforded mem- is no limit tothe depth Students view either of anyof the bers of ACM (M. hierarchies. The topicsfor all courses faculty selected questions or random sets are physicallyin thesame data setsof ofquestions, and receive feedbackon the data base, but logically keptdis- their performance. The facultycontrol tinct. The hierarchical organization of the studentsare allowed which questions topics provides flexible retrieval possi- to view, but students control (within lim- bilities. Questions may be retrieved based its) the course the number of questions, on specified topics. Alternatively, they question topics, and text page ranges over may also be retrieved by topics that exist which questions may beselected for self- as children nodes in the hierarchical assessment. Students are generally enthu- topic structure. siastic about participating inquestion Retrieval, therefor, is made asgeneric or specific as desired. Faculty and student userscan view at any time the hierarchical structureof the topics which, similarto thequestions within any question pool, changesover 2Theself-assessment aspectof RIBYT is time. one of the data base's unique characteristics, from which it derives its name IMPLEMENTATION OF RIBYT ,(Review It Before YOU Test). RIBYT was designed under the assump- 'The review process is also used to modify tion thatfaculty and students using the or delete questions which have failed system would know little or nothing about statistical item analysis.

21 7 Testing/Placement207

data baseprocessing. Application pro- Before faculty or students store or grams written in COBOL and BASIC act as retrieve questions, information concerning the communication between the DBMS and the the users is stored in the data set called users, who access the system through batch ROSTER; course information inthe data and on-line devices. RIBYT is implemented set TEXTBOOK-MASTER, and system initiali- on a Hewlett Packard 3000 SeriesII com- zation information in the data set LAST- puter using the DBMS package, IMAGE, which MASTER. Each record in ROSTER contains a has been widely acclaimed (5). The design user's name, course and section number, a of RIBYT is not restricted to IMAGE, how- status (faculty or student), and a unique ever.With minor modifications RIBYT can user identification number.TEXTBOOK-MAS- beimplemented on manynetwork-oriented TER contains one recordfor each block of DBMS. CINCOM's TOTAL DBMS is but one twenty pages of each textbook. Each record example (3). contains the textbook title, anda data item BOOK-BLOCK composed of course and an ASSOCIATIONAL STRUCTURE OF RIBYT associatedtextbook numberand the page Figures 1 and 2 illustrate the 22 data block. Such structuringgives more rapid sets (files) of RIBYT. There are eight retrievalfor those questions based on master4 and 14 detail data sets. Master textbook page %webers. The data set data sets contain records stored by hash- LAST-MASTER containsone record. Stored ing. Each master record, composed of one here are the sequencenumbers of the last or more data items, can be linked to one questionentered into the data bare or more detail datasets. Master data (LAST-QUES), the lasttemporarytopic sets can be viewed as either independent (LAST-TEMP-TOPIC), the last permanent files or as sets of records where each topic (LAST-PERM-TOPIC), and a faculty record acts as a chain head to one or more password (FAC-CODE). Thelatter allows records ina detail data set. Within a faculty to perform such special functions detail data set records are linked by for- as dumping an entire question pool for a ward and backward pointers to form sets. A course. In addition to the securitypro- detail record may belongto one or more vided bythe operating system and the master data sets. Master data sets are of DBMS, FAC-CODE and ROSTER are used to con- two types, manual and automatic. Unlike trol access to the data base. SYSTEM- records in manualdata sets, automatic STACK is the other system data set, imple- data set records are stored or deleted by mentedas a stack that contains the DBMS itself. Theserecords act as information on the last50 interactive or Chain heads forrecords in detaildata batch jobsrun. Eachrecord contains a sets. With this data base architecture, time stamp, plus a user and program iden- networks of associations among records and tification number. This information aids data sets can becreated and then manipu- in backing out of system malfunctions and lated by the DBMS. Figures1 and 2 show detecting unauthorized access. for each data set its name, its type5, and A user, wishing to enter questions its associations with other data sets. into the data base, begins a session at a Within a detail dataset a chain of CRT by choosing from a menu of textbooks records can be further logically ordered associated with the user's courses. This by specifying one of the data items within menu usesinformation from the datasets the record as a sort item6. ROSTER, ALTERNATIVE-COURSE, and TEXTBOOK- MASTER. Then, for each question, users are prompted for background information on the question before entering the text of the question. This information will be entered into the data setQUESTION-INDEX. The user enters two course topic numbers, a textbookpage number, and the type of 4Aside fromtheterms "master" !_and question. The application program assigns "detail"which arepeculiar to Hewlett a permanent or atemporary number to each Packard,we will adhere to CODASYL(1) new topic. The namesand numbers of new terminology as closely.as possible. topics are entered into the data set TOP- 5In figures 1 and 2, B- Binary and IC-MASTER. If the user wishes to index a X=Alphanumeric. The numeric value following the alphanumeric specification "X" represents the number of bytes assigned to that data item. 6In figures 1and 2 these sort items are 'Mr readability the data.base, data sets, designated as"S" underthe appropriate and data items areshown in capital let- data items. ters. 208 NECC 1980

question on only a single topic, the sec- keeps thechoices and the lines of text ond topicbecomes a dummy. The program within each choice in logical order. The then adds to therecord in QUESTION-INDEX totalnumber of timeseachchoice is the DATE, the BOOK-BLOCK, a unique ques- selected as ananswerin eitherformal tion number (QUES -NUM), the version number testing or self-testing is later stored in of the question (QUES-VERS), and the the data item RESP-TOTAL. user's identification number. The status of thequestion (QUES-STATUS) is set to QUESTION RETRIEVAL 'edited" if the useris a faculty member, There are many selection options for 'otherwise it is -set- to "unedited."The question retrieval from RIBYT, aside from other data items inQUESTION-INDEX are retrieval for editing purposes. Question initialized to zero. retrievalis first divided intoformal After all data item values have been testingandself-testing. Facultycan coastructed fora recordin QUESTION-IN- restrict questions so thatthese are used DEX, as the recordis physically stored, onlyin formal testing. Each type of the DBMS creates multiple associations testing is further divided into three using hashing and linked lists.The ques- options. Underthe first option faculty tion's background information in QUESTION- specify byquestion number the actual INDEX islinked to its author in ROSTER, questions to be included in a test. Under the textbookreferenced inTEXTBOOK-MAS- the second option faculty specify that a TER, the question's topics in TOPIC-HEAD, single set of questions are to be randomly and itsuniquequestion number in QUES- selected foran entire class. under the TION-HEAD. third option faculty indicate that a dif- The text for a questionand its ferent set of questions are to be randomly answers are next entered into thedata selected foreach student. Faculty may setsQUESTIONS and ANSWER-CHOICES. For specify that questions areto be selected fill-in-the-blankquestions the textis based on one or more topics, page ranges, stored inQUESTIONS andthe answers are or any combination oftopics and page storedin ANSWER-CHOICES. Alternatively ranges. In self-testing, faculty may acceptable answers foreach blank are leavetopic andpage selection to each stored in ANSWER-CHOICES. For multiple - student's choice. choice questions, the stem of the question is storedin QUESTION and the choices in DATA SETS USED IN QUESTION RETRIEVAL ANSWER-CHOICES. Each choice is designated Three data seta areused tocontrol as correct or wrong (CORRECT-WRONG). One question retrieval. These data sets con- can not afford to assign to each question, tain specifications that areused by a choice, and answer a fixed length equal to retrieval program to select questions from the largestpossible number of bytes the the data basefor formaltesting and text might need, rather, data base tech- self-testing. The first data set, TESTS, nology isused tosolvethisproblem. has a primary key COURSE-TEST-NUN, which Text is considered as 40 or72 Character is a combination of the course designation blocks linked together. Records in QU2S- and a test number within the course. Each VERS-HEADarechain heads forsetsof record in TESTS represents a single test. character blocks in thedata setsQUES- Each record has a SECURITY-CODE to control TIONS and ANSWER-CHOICES.Within the data access to the test. The percentage that set QUESTIONS each textual block of char- each testrepresents towards the total acters is assigned asequential line num- coursegrade is stored inTEST-WEIGHT. ber (LINE-NUM), which is used as a sort The number of questions per test, the num- item tokeep the text incorrect logical ber of times atest maybe taken, the order on a chaine. Since any choice of a average difficulty factor per test ques- multiple - choice question may also have tion, the types of questions, the selec- more than one line of text, the sort item tion options, and the mix between vali- called SEQUENCE is composedof the alpha- dated and unvalidated questions are stored beticchoicedesignator (A,BiCiDtor E) in the data set TESTS. When faculty wish concatenated to a text line number within to designate the actual test questions, the claim. This system automatically the set ofquestion numbers is stored in the data set SPECIFIED-QUES by question number (VERB -NUM). Each number acts as a search,argumentto retrieve the text of one questl m,This retrieval is performed as follows. First thesystem usesthe ftecords in any of thedata sets are not value VERS -NUM in SPECIFIED-QUES to hash necessarily storedin the same orderin intothe data set QUES -VERS -HEAD. The _which they ;sere .input. retrieved record than points to setsof 22 Testing/Plaoment209

textual blocks, representinga question. correct or incorrect, thestudent's its choices, and answers in the data sets response, and a combined data item (USER- QUESTIONS and ANSWER-CHOICES. DATE-TIME) formed from the student identi- The data set PAGES-TOPICS is a multi- fication number and the time stamp. Each purpose data set. Each record in PAGES- response in the data set RESPONSES is on TOPICS stores either a textbook page range two link paths, one originating from the or one or twocourse topic identification data set QUES-VERS-HEAD, the other from numbers. Each record is linked to an the data set TEST-HEAD. In addition, individual test inthe data set TESTS. since it is desirable to have the respon- The value ofthe data item PAGE -TOPIC - sescontinually sorted by test question SWITCH indicates-whether a record contains number withina test and sincethere may page or topic information. For pages, the be morethan onetext block per student data items START and END store page ranges response, the two sortitems, TEST-QUES andthedata item BOOK-HIERstoresa NUM and LINE-NUM, are used asmajor and coursetextbook identificationnumber. minor sort fieldsfor the records on a For topics, two topic identification num- USER-DATE-TIME chain. berscan bestored perrecord;one in Besides receiving the traditional per- START and one inEND. BOOK-HIER is then formance feedback totheir responses (a used as a binary switchtoindicate list of correctanswers, questions, and whether only designated topics or their test scores), students may receive textual hierarchical childrenare to be included comments concerningparticular responses. in question retrieval. Question retrieval Faculty store these in thedata set RESP based on textbook pages is accomplished by FEEDBACK. After students receive feedback hashing into TEXTBOOK-MASTER and following on their tests, they may enterinto the appropriate chains into the data set QUES- data set STUDENT-GRIPES their reactions to TION-INDEX andfrom there usingthe com- any individualtest questions, anony- bined data items, QUES-NUM and QUES-VERS, mously, or by studentidentification num- as a key tohashinto QUES-VERS-HEAD. ber. Course grades can easily be deter- Then the program follows chains into QUES- mined based on the datasets TESTS and TIONS and ANSWER-CHOICES as described TEST-SCORES. above. Question retrisVal by topics is similarafter entry into QUESTION-INDEX STATISTICAL ITEM ANALYSIS from TOPIC-HEAD. Sinceall hierarchical Two measures of a question's quality twins of a given parent topic are organ- under statisticalitem analysis are item ized as a set in the data set TOPIC-INDEX, difficulty and item discrimination (4). A whosemembers are linked to a topic in difficultyindexis calculated concur- TOPIC-HEAD, sub-topics are found by first rently for every question in a test. The hashing into TOPIC-HEADand following a set ofstudent responses is retrieved by chain into TOPIC-INDEX. hashing intothe data set TESTS with the key COURSE-TESTNUM and obtaining all STORING AND RETRIEVING TEST RESPONSES AND chained recordsin the data set TEST- TEST SCORES SCORES. For eachrecord retrieved the Information concerning student respon- data item values of USER-NUM and DATE -TIME ses is stored insix data sets/ TEST- are used as a concatenatedkey to hash SCORES, TESTS, RESPONSES, STUDENT- GRIPES. into the data set TEST-HEAD. Alltest ANSWER-CHOICES, andRESP-FEEDBACK. The responses fora single student are set of all test scores for a given test is retrieved from the data item CORRECT-WRONG linked together within the data set TEST- of the data set RESPONSES through a chain SCORES, and this setis itself linked to originating in the retrieved TEST-HEAD the data set TESTS. Within the data set record, and this procedure is then TEST-SCORESis a chain associatingall repeated for each student. The retrieval tests for each student. Since RIBYT han- is rapid and passes over all other tests dles the possibility that the same student in the data base. The calculated diffi- takes the same test more than once, each culty indices are thenstorcd in the data test score is made unique by being associ- set QUESTION-INDEX. The calculations for ated with a time stamp DATE-TIME in TEST- statisticalitem discrimination require SCORES. thesame data setsas item difficulty The student responses to each test determination. question arestored in the dataset RESPONSES. Each record is composed of a CONCLUSION question identification number (VERS-NUM), The associational powerinherent in the question number within the test, the thedata basesystem RIBYT allows many text '.ine number within the response, a relationshipsto bemade and maintained code indicating whether the response was among courses, textbooks, coursa topics,

220 210 NECC 1980

students, faculty, tests, test questions, (9) Kroenke, David, Data Base Processing, testconstruction options, and student Chicago: SRA, 19777- responses to questions. RIBYT stores subjective and statisticalfeedback on the (10) SRA, "Mid -Term Blues? Computerized quality of the questionsin the data base Test Service Available," Data Proc- and aseociates this feedback with the test essing News (Spring 1979)71 questions. Thissystem supportsformal testing and self-assessment procedures and (11) Wong, J. W. and G. Scott Graham, salads the storage ofmany types of ques- "Self.-Assessment Procedure VI," CACM, tions, while imposing no limit on the size 22,8 (August 1979),449-454. of any question or answer,---The centrali- zationof the logical associationsinto one data base provides better security and easier maintenance than traditional file systems.

ACKNOWLEDGEMENTS I am grateful for the assistance of Dr. Kathleen cokrigan Fuhs and Mr. Paul Thompson for helping to specify and clarify some of the user requirements of RIBYT.

REFERENCES (Manual) LAST-MASTER (1) CODASYL, CODASYL Data Base Task Grp* Report, ACM, New York,

STACK LAST- LAST- LAST- FAC- (2) Cwdd, E. F., "A Relational MOdel QUES TEMP- PERM- CODE of Data for Large Shared Data TOPIC TOPIC Banks," CALM, 13,6, 1979.

(3) Datapro, TOTAL, Delran, N.J.: X4 B B B X8 Datapro Research Corp., March 1978, section M12-132-101 to M12-132-104. key (4) Downie, N.M. and R. N. Heath, ---- .. Basic Statistical Methods, 2nd. ia7,Wew York: Harper AN Row, 1965, p.228. (detail) SYSTEM-STACK (5) Gepner, Herbert I., "User Rat- ings of Software Packages," Datamation (December 1978), 183486. :.;ACK DATE- LAST- LAST- TIME USER PROGRAM (6) Hewlett Packard, Image, Data Base Management, System, ence Manual, Santa Clara, ca.t MgreETWalcard, 1978. X4 X12 X10 X4 (7) Hilgard, E. R., R.L. Atkinson, and R. C. Atkinson, Introduction

to Psychology, 7th ed., New York: I S Harcourt, Brace, and Jovanovich, 1979. FIGURE 1: SYSTEM CONTROL DATA SETS OF RIBYT (8) IBM, IIMB/VS, Application Pr ram- akaeference Manual, or er no. SR-20-1026, whit7-17017ins, New York, 1978. (Auttratic) (Manual) (Automatic) (Manual) (Manual) TOPIC-HEAO TEXTBOOK- QUESTION- ROSTER TOPIC-MASTER MASTER HEAD

TOPIC BOOBLOCKK LAST-FIRST- USER - TOPIC NUR TITLBOOKE NAME NAME COURSE \SECTIONSTATUS NUN ICOURSE \TOPIC NUN

X30 X12 X20 XI6 X6 ' X2 X2 XIO X6 X30 X4 - key key

1 _

(Detail) QUESTION-INDEX

TOPIC- TOPIC-DATEBOOKPAGE-QUES-QUES-DIFFI.QUES -.QUES-DISCRINI.,°"; BLOCKNUN NUR'MRSCULTYTYPE STATUSNATION" NMI NUN2 NUM

X4 X4 X6 X12 X4 X4 X2 X2 X2 X2 B -X10

,S S S S . . .

(Detail) (Detail) (Detail) TOPIC-INDEX QUES-FEEOBACK ALTERNATE-COURSE

TOPIC-SUB- SUB - VERS-LINE-FEE° USER - NUM TOPICBic. NUM NUMma NUNCOURSESECTION X4 X30 X4 alaX2 MUX10 X10 X6 X2 MINIM I

FIGURE 2: DATA SETS OF RIBYT

241 223 (Detail) QUESTIONS (Detail) ANSWER-CHOICES

VERS- LINE- NUN (Automatic) NUN TEXT VERS- SEQUENCE RESP- QtES-YERS4AD NUN CORRECT- TEXT TOTAL X6 X2 X72 WRONG VERS- X6 _. NUN X4 B. I---.. S- -X2 X40- S X6 (Detail) RESP-FEEDBACK (Detail) STUDENT- GRIPES VERS- LINE- NUN NUN FEEDBACK VERS- STUD- LINE- - HUH NI/4 NUN GRIPE

X6 X2 11. X72 X6 X10 X2 1 X72 S S (Detail)' RESPONSES (sqeal) TEGTu-HEAD

TEST- LNUINE- CORRECT- STUDENT- USER- USER-DATE- S. N WRONG RESPONSE DATE- TINE NUN T

X6 X6 X2 X2 X40 X22 X22 4 -s

Figure 2 (cent): DATA SETS OF RIBYT 2? (manual) TESTS

(from Roster)

,0,,,,I0U5E; SECURITY- HUM-OF-TEST- AVE- QUES- SELECT- VALID - TEST: CODE "'u"WEIGHT QUES TIMES OIFF TYPES OPTIONS MIX

- -.. . X12 X8 8 8 8_ 8 X2. X2 8

...._ - , key

. , -I, -

I ,

(dtai) (detail) (detail) SPECIFIC/lIVES PAGES-TOPICS TEST-SCORES --- _

COURSE- %MRS- COURSE- START ENO BOOK- PAGE- COURSE- CATE- SCORE USER- TEST-HUM NUM TEST-NUM HIER TOPIC- TEST-HUM TINE NUM SWITCH

X22 X6 X12 X4 X4 X2 X2 X12 X22 8 X10

. , . S

Figure 2 (cont): DATA SETS OF RIBYT

2 5 214 NECC 1980

COMPUTER MANAGED PLACEMENT

MATHEMATICS INSTRUCTION

FOR

HEALTH OCCUPATIONS STUDENTS

Thomas A. Boyle Peter T. Magnant Purdue University Indiana Vocational Technical College West Lafayette, Indiana 47907 1815 East Washington Street (317) 749-2256 Indianapolis, Indiana46202

.INTRODUCTION

The Indiana Vocational Technical 4. Neither the student nor the College at Indianapolis (IVTCIN) main- staff could readily get an impression of tains an instructional program to serve how the student stood with respect to the mathematics needs of students be- the whole program. ginning study in health occupations COMPUTER APPLICATION technologies. The program is based on in an effort to mitigate these Streeter and Alexander's Fundamentals problems, work was begun early in 1978 of Arithmetic (1), beginnlTifh-ngt to adapt test materials and procedures whole number operations, continuing which had been in use at another XVTC with fractions, decimals, and percents, institution (2). These materials con- and extending to ratios and signed sist of a programmed test format, which numbers. The materials are organised require. students to attempt equal num- in 15 modules, each with a pre -test, bers of test items in each of several a practice test, and a mastery test categories, together with computer used to guide individual students to scoring and data processing. This use the study materials appropriate to of a computer enables first the scoring their evident needs. of each student in each item category, Although the instructional with subsequent determination of score materials in the program proved statistics for each student group. As appropriate and were reasonably well will be shown in subsequent examples, received by students using them, this use of a computer yields informa- certain problems soon developed in' tion in detail which would be practi- guiding students to the modules speci- cally impossible to get from hand scor- fic to their needs. in particular, the ing. Previous work with the combina- following were noted: tion of test format and computer data 1. The handling of pre-tests on processing had demonstrated efficiency an individual basis was time consuming in the use of test administration time and otherwise a burden on testing and the possibility of rendering in- personnel. formation from each of several relative- 2. Because of the limited number ly independent item categories (3). of pre-test 1orms and the level of Figure 1 shows the computer scor- supervision for actual testing, it was ing output for a select group of possible for students to use the pre- students. The category of subtest test materials in ways which would fool scores appear in six columns following themselves regarding their mathematics the student names.A key at the bottom skills. serves to identify six subtests. For 3. Capable students were bothered this test some grouping of modules was by the need for working through several necessary, e.g., instructional modules tests in order either to find a' suitable 1-3 dealing with whole numbers are all starting module or to demonstrate represented in subtest one.The key competency and by-pass the program further identifies the SCR column as entirely. 2 7 Testing/Placement215

I V T C INDIANAPOLIS 4Z6/79 SUBTEST SCORES NAME 1 2 3 4 5 6 1-6 NF ERROR ITEMS

Bob W. 3 -4 2 1 -4 1 -1 37 51 32 18 50 4 5 14 26 61 72 76 82 71 79 8 .43 -.57 .33 .17 -.80 .17 -.03

Bev H. 5 -2 -2 -5 1 -4 -7 44.13_33 53 17 45 52.39 58 14 66 76 82.71 85 7 .83 -.22 =.29 -.63 .14 -.57 -.16

Sue D. 7 -4 -2 0 -1 -2 -2 48 7 24 21 53 28 18 50 57 40 14 66 76 82 71 6 1.00 -.40 -.29 0 -.13 -.22 -.04

Jo B. 7 -2 1 -4 0 -2 0 48 31 42 54 50 45 23 38 59 44 66 76 71 79 62 6 1.00 -.20 -.13 -.57 0 -.25 0

Pat W. 7 5 3 5 4 -1 23 42 18 56 52 5 59 19 66 76 81 74 62 83 80 78 1.00 .56 .43 .83 .67 -.14 .55

Pan W. 7 8 2 0 -2 0 15 48 31 23 58 30 26 83 80 77 85 81 74 66 76 65 7 1.00 .89 .22 0 -.25 0 .31

Joy D. 8 7 6 -1 2 2 24 48 7 24 52 30 26 67 77 85 75*77 79*73 65*79 6 1.00 .78 1.00 -.13 .29 .20 .50

Dee R. 8 8 5 -1 5 0 25 46 53 56 52 14 72 7S 77 85 81 70 65 66 76 65 6 1.00 .89 .71 -.14 .71 0 .54

Liz R. 7 9 6 -2 7 8 35 46 13 43 42 14 79 GG 85 1.00 1.00 .86 -.25 1.00 1.00 ./6

Kay B. 7 10 7 -7 8 8 33. 48 13 43 42 57 30 79 85 65 1.00 1.00 1.00 -.88 1.00 1.00 .69 SUBTEST SET ** 1 = WHOLE NUMBERS ** 2 = FRACTIONS ** 3 *DECIMALS ** 4 = PERCENT ** 5 = RATIO ** ** 6 = SIGNED NUMBERS ** ** SCR = SUM OP SMUT SCORES ** NF = NUMBER OF ITEMS ATTEMPTED ** RATIO SCORE = SCORE DIVIDED BY RELATED NP ** Figure 1. Facaimile*GlftthemfticeScaingPrcammOiftlat holding the sum of six subtest and the NP toe result obtained from dividing the raw column as showing the total number of score by the number of related items at- items attempted by each student. Numbers tempted. As should be evident, the com- at the far right identify the first ld putation of seven raw scores and seven mistakes made by each student. ratio 'corm would present a formidable TEST SCORING task for anyone attempting hand scoring of Subtext scores for each student are this type of test. presented in two numbers. The first, the APPROPRIATE PLACEMENT raw score for each subtext, is based on The students represented in Figure 1 one point for each correct response.A were selected to present a range of stu- penalty is applied for making errors in dent performance. Bob, the first student, sequence, this scoring yields an expected obtained a row of scores which could well score of zero for a student strategy of result from guessing. It appears likely pure guessing. If performance is worse that he made consistent errors in items than that from pure guessing, i.e., if a dealing with fractions and ratios, but he student consistently makes errors in a earned only three points in seven attempts subtest, then that subtest score becomes with whole-numbers, so we may be reason- negative. The second number, the ratio ably confident Bob should start at the score appearing below each raw score, is beginning of the instructional program. 218 NECC 1980

INDIANA VOCATIONAL TECHNICAL COLLEGE INDIANAPOLIS M - 31 MATHEMATICS SKILLS ASSESSMENT PROGRAM FOR Pat W. 4 679 80981 44 THIS PROGRAM IS AIMED AT HELPING YOU DEVELOP THE BASIC MATHEMATICAL SKILLS NEEDED TO DO AN EFFECTIVE am IN THE IVTC SPECIALTY YOU INTEND TO ENTER.THE PROGRAM HELPS YOU AND YOUR INSTRUCTOR FIRST BY NOTING THE PARTS OF THE IVTC MATHEMATICS SEQUENCE IN WHICH YOU MAY ALREADY HAVE ADEQUATE SKILL.NOTE IS ALSO MADE OF THE SECTIONS, OR MODULES IN WHICH YOU APPEAR TO NEED ADDITIONAL HELP AND PRACTICE.AS YOU KNOW, MOST PERSONS EN- TERING TECHNICAL EDUCATION NEED TO IMPROVE THEIR SKILLS IN SOME PARTS-OF-MATHEMATICS. THE PRIMARY PURPOSE OF THIS ASSESSMENT PROGRAM IS TO HELP YOU GET STARTED AT THE POINT WHERE YOU CAN DO YOURSELF THE MOST GOOD., ONCE YOU KNOW THE MATHEMATICAL MODULES YOU NEED TO WORK ON, YOU WILL FIND YOUR INSTRUCTOR AND THE LEARNING LABORATORY STAFF READY TO HELP. ,AND AFTER A PERIOD OF INSTRUCTION AND PRACTICE, YOU CAN CALL ON THE SKILLS ASSESSMENT PROGRAM TO CONFIRM THE PROGRESS YOU HAVE MADE. THE FOLLOWING MESSAGES SHOULD HELP - Pat W. YOU WORK SOMEWHAT FASTER THAN THE AVERAGE WHEN TAKING THE TEST.THIS IS A REASON- ABLE STRATEGY, OUT YOU MAY BE GUESSING ANSWERS TO ITEMS WHICH YOU COULD SOLVE IF YOU SPENT A LITTLE MORE TIME ON THEM. IT APPEARS YOU DO YOUR BEST WORK WITH THE MATHEMATICS IN MODULES 1 - 3.YOUR SCORE OF 100 INDICATES ADEQUATE SKILL WITH WHOLE NUMBERS.

JUDGING FROM YOUR RESPONSES, IT APPEARS LIKELY THAT YOU HAVE ADECIJATE SKILLS IN THE FOLLOWING TOPICS.

TOPIC SCORE IVTC MATH MODULES PERCENTS 83 TEN ALTHOUGH YOUR SCORES DO INDICATE SUFFICIENT SKILL, THERE MAY BE SOME ROOM FOR IM- PROVEMENT, AND YOU MAY BENEFIT FROM A QUICK REVIEW OF ONE OR MORE OP THE MODULES LISTED.

EVIDENTLY YOU HAVE MADE SOME PROGRESS IN LEARNING ABOUT FRACTIONS, HOWEVER YOU DO APPEAR TO NEED MORE PRACTICE WITH THE MATERIALS IN MODULES 4 - 7.

PROBABLY YOU CAN DO YOURSELF A LOT OF GOOD.BY GETTING TO WORK ON THE MATERIALS IN MODULES 8 - 9, EVIDENTLY YOU DO NOT HANDLE DECIMALS VERY WELL. NO DOUBT, YOU REALIZE THAT YOU KNOW SOMETHINGS ABOUT PATIOS, BUT YOU DO MISS SOME POINTS HERE AND PROBABLY WILL BENEFIT FROM ATTENTION TO IVTC MATH MODULI ELEVEN. EVIDENTLY YOU MADE MISTAKES ON THE FOLLOWING TEST ITEMS 18 56 525 59 19 66 76 91 74 62 83 80 78

Figure 2. Paeietrdle 14-3119it1anatics Sooting Prcgraft Oftpirt The next three students received in other subtests. very low scores inmost 'attests, yet Pan, Dee. And soy, the next students scores which are evidently adequate in listed, had scores characteristic of ade- the whole number subtest.Bev had the quate performance in two or three sub - lowest total score (-7) observed to date, tests. Recommendations for these students but did five out of six whole- number would lead to intermediate modules in the items correctly. Evidently these students instructional program. The scores for will benefit from starting with the in- PAM and Dee indicate some study is needed structional materials on fractions. pat beginning with module eight. Joy may will also benefit from starting with the well be advised to begin with module ten. instructional materials on fractions, The two students remaining, Lis and however, she obtained acceptable scores Kay, received practically perfect scores

229 TestingiPlacement 217

in all but one subtext. Evidently these SCORE STATISTICS girls made all but one of their mistakes During the period to July 1979, the on items in the percent subtext. May's M-31 test was administered to approxi- scores are especially noteworthy. She mately 1000 IVTCIN students. Prom the worked rapidly on the test, attempting 48 total group, some 763 were identified by items in the 30 minutes allowed. She instructional program.The response data got all items correct in five subtests from these students were grouped and test yet made errors in all eight of the per- score statistics obtained for students cent items she tried: the negative seven entering each of seven health occupations score results because the first error is programs at IVTCIA. The numbers of stu- not penalized. dents identified by_ instructional program INDIVIDUAL OUTPUT ranged from 31 for health career prepa- The computer program which produces ration to 455 entering the practical the test scores has been extended to nursing program. Numbers of students in yield a one-page message for each student groups may appear to change and the sums on which the student's ratio scores are of groups may be different from the total ranked, and the subtests on which he did tabulated in Figure 3.This is because best are identified. Appropriate mes- additional data were received and pro- sages are selected, including recommenda- cessed during the time the report was in tions for notion. Duplicate copies are preparation. produced for the student's file and for The score statistics were arranged instructional supervisors.A facsimile by ranking the mean total ratio scores of this output is presented as Figure 2. for the instructional groups.This

1NDIMMIKMIXAMLIECENVFICILLEGE 1101/MOLys 14-31 19111114=8 VMS=SUMMRRY AUGUST, 1979 INKRUCIFICING. 029421 DE m= maw maw exam Ian riss FFOGIOM NOS. PEACTICAS NOS. SOIREMOrtED

148 mem =ARM MEAN .93 75 .74 .49 .76 .71 .73 40.56 MO 71 STD DEN .09 .35 .33 .44 .39 .34 .22 8.62

2 28 MIID LAS TECH MEAN .93 .72 .75 .45 .72 .70 .71 40.07 MO 61 9P1) MY .15 .41 .29 .52 .37 .37 .25 9.06

3 46 RADEECGIC 'MICE toe .92 .66 .69 .47 .74 .66 .69 39.31 Ns 57 STD 1:43/ .13 .36 .34 .44 .39 .33 .21 7.84

4 42 SURIOCALIECH MEAN .95 .57 .68 .39 .67 .69 .65 39.25 MO 57 STD DEV .09 .48 .34 .46 .51 .39 .25 8.27

5 44 LICPRACEICRSE MEAN .92 .62 .66 .27 .59 .49 .60 38.65 N- 455 STD DOW .15 .41 .30 .48 .47 .44 .25 8.42

6 37minawAser MAN .93 .62 .66 .22 .45 .46 .56 35.64 MO 39 STDrev .12 .42 .32 .50 .53 .52 .26 8.27

7 93 MUMMIES:REMO MEAN .78 .33 .33 -.06 .21 .30 .31 37.45 MO31 EaDDEV .27 .54 .60 .53 .50 .38 10.87

wan .92 .62 .66 .32 .61 .55 .61 38.52 ETD DEV .15 .42 .32 .50 .47 .44 .26 8.69

115t9111 3. 14-31 =SE E5312791,101 DIMSWEID BY INVIRICYICti% GROUP

23 218 NECC 1960

ranking placed the respiratory theraphy students grout's, for example to see *tether the respiratory highest, and the health career preparation therapy students and the LPN's could be regarded students latest. The 0:Inplete statistics for all as caning from the same PoPoleriro- The analyses seven Student grape are presented in Figure 3. were me twice, once with and once without the Mean scores are presented for all six of the health career preparation group. The results are N-31 subtests, as well as for total score and presented in Table x la teats of P. values. number of item attempted. With the exception of the items-attepted mean, the scores represented are ratio mores (i.e., the raw score WRTHOUf WITH divided by the number of related items atteipted) ;MEM SCP MCP for each student. As can be seen the subtest means range fran 0.95, correeronding to a mean of 1 %tole Numbers 0.545 5.271* ninety-five percent, down to a negative 0.06. This latter score indicates average test per- 2 Fractions 2.099 4.473* formance just slightly worse than would be mc- rented for aetudett strategy of pure guessing. 3 Decimals 1.581 7.425* Figure 3 Indicates that, an the average, the 14-31 test is of reasonable difficulty for 4 Percents 5.246* 7.455* students entering health service instruction programs at IPDZIN.However, the difficulty is 5 Ratio 4.172* 7.413* not even for different groups. The respiratory therapy and medical laboratory technician 6 Signed mamba= 7.747* 8.135* trainees and the radiologic aid surgical technicians find the test quite easy. The statistics ghat that, especially in the first grouPs, many *Significant at 0.01 ac less students; handle the designated range of arithmetic skill adequately.Unquestionably, Table x.F-values for analysis of there are many of these students vim have little variance in subtext scares, to gain from further study of the skills =pre- with and without health sented. Among the students who do seen to need care preparation students. further related study, the need appears to be greatest for percent drills. These values show that, excluding the OCP Data for all students, including sane students, only marginal differences appear tested after mid-July, We that there is sane correspondence between the arltlenetic between groups wititxecex4 to ichole =bar, that have been =Allred by the entering students teactica, and decimal'ambtest scores.Signifi- ard the sequence of interactional reddest. On cant differences appear in the other subtast average, all groups score highest on scores. When the RCP students are included, the whole amber subtest. With the eocception sg=icant differences appear in all subtlest of the percent scores, there is a trend toward APPLICATICN OF MUMS laser scores in the subtests related to later Zang the ptvoedures discussed, students are instructional modules. The break in the se- able to review their am strengths and weekneeses quence of scores, at the percentsubtlest, may indicate either need for special effort if related to mathematics, both as individual students are to attain 80 percent performance applicants and in relation to the other Maim** or disPlY that the test items mere seeking admission into a program area.The sonbydifficult.t1te heck dose call at- Indiana Vocational Technical College at Indianap- tention to an advantage of the test format: olis helps students in their areas of deficiency evidently saw students oho have difficulty with sib they can develop the mathematical skills test item related to module 10 readily surpass necessary for admission aid eventual success. If the criterion BO- percent perfonerms in item scores are acceptable for program adedssicn, the related to module 11 and to signed ambers. Related Education Department can use the date as Here the test faint and amputee data proces- a basis to increase mathematic skill in light of sing enable identification of the difficult the health programs needs and objectives. materials and may obviate the need for subse- aben imbalance develops belmeen the ambers quent modules. of students seeking to enter different health SIGNIFIGINZE specialities, an 1ndividual's scores may wag* TIO3 statistical teats have been performed encouragement for consideration of other on 14-31 data obtained to date.One-way specialities also within the student's preterit or analyses of variance have been clone at each future range of mathematical developeent.For ooltent4Figure 3) of group subtest scores to example, sane students seeking to enter the check for significant differences between practical nursing program may beneficially

231 Testing/Placement219

ansider other instructimal program. in, or by-pas3 of, a sequence of asthmatics OICELBICHS instruction SWUM. Scam normative data have A ccmputer-tesed testing prooedure has been been obtained and ease significant differences adapted to menage the mathematics placement of seen to exist between scores of students Madames pursuinginetzuctia, in health mama- seeking inatructian insdifferent health time. The anaiattion enables more efficient specialities. use of scixol persamel and expedites pleasant

REFERIEWES 1. Streeter, Jame and Alexander, Gerald.Fundamentals of Aritbatic. New York: Raper and Aces Publishers, 1975. 2. Boyle, Timms A. and Shaver, Carroll G. Ccmwtar Amassment of Mathematics Skills for Students beginniza Postnday Technical Evalw. Seventh 0:inference on Omenzbers in the Maks- gradeate OnTicula. KrOaftn: state university of New York, 1.976. 3. Boyle, Thaw A. and Wright, Guy L. Itcarater-aanisted Evaluatica of Student Acheivement." Engineering klucatice, vol. 68, No. 3, Decenber 1977, m. 241-245.

232 Invited Sessions

IMPROVING UTILIZATION OP TWO-YEAR COLLEGE COMPUTER CENTERS

Robert L. Burrows Triton College 2000 Fifth Avenue River Grove, Illinois 60171 (312) 456-0300

ABSTRACT th=difiC11118i011will begin with an overview of the areas of responsibility of a computer-assisted learning agency and then elaborate on what can be done in these areas once one actually becomes an agency. Specific topic areas include the following: --Marketing approaches to attract teachers to the computer via such means as seminars, newsletters, and committees. --Training instructors in computers, using as specific examples the three courses now being offered to Triton faculty each semester: Introduction to Computers and Programming, Intro- duction to Statistical Analysis Using SPSS, and Introduction to Computer Graphics. --Working with faculty to obtain educational software including catalogs, magazines, educational suppliers, and user groups. --Supporting software written or planned for academic users, including changes to system software, program filters written to convert foreign software, an electronic mailing system, a program directory system, and a car system. --Working with faculty in obtaining hardware, using as example the terminals and microcomputers purchased at Triton and its plan for a computer lab. --Problems with computer education specific to a community college.

2')')4, to. Invited Sessions 221

TEACHING .COMPUTER ETHICS

Walter Maner Philosophy.and Computer Science Old Dominion University Norfolk, VA 23508

ABSTRACT --Wriinagemont, staff, and users of information systems of all kinds will benefit greatly from training in applied professional ethics. These professionals work in an environment where they must deal responsibly and knowledgeably with critical moral problems (such as breau4 of privacy, computer crime, and dehuman- ization) which are aggravated, transformed, or created by the advance of computer technology. A general rational for a course in computer ethics will, therefore, be developed along with course design criteria, a full course description, a set of proposed course objectives, associated bibliographies, and a taxonomy of subject matter areas within the field of information ethics. Attendees will break into small discussion groups to study cases illustrating moral dilemmas posed by the use of computers in education.

23 .f Tutorial

PASCAL TUTORIAL

Harry P. Haiduk Amarillo College P.O. Box 447 Amarillo, Texas 79178

ABSTRACT --YETrEutorial is concerned with (1) a brief historical view of PASCAL in terms of its philosphy and stated design goals (2) a brief review of its current relevance in a diverse set of applications, particularly as it may relate to the new Department of Defense Common High Order Language, ADA (3) a comparison of PASCAL's logic and data structures with those of BASIC, COBOL, FORTRAN, and PL/I (4) actual running program examples con- trasting PASCAL with BASIC and FORTRAN.

2 3 5

222 IMM'y A11111111===1.11 Pre-College Instructional Materials

COMPUTER -BASED INSTRUCTION FOR THE PUBLIC SCHOOLS: A SUITABLE TASK FOR MICROPROCESSORS? Dr. Timothy D. Taylor Computer Based Education Center 308 Carroll Hall The University of Akron Akron , Ohio 44325 (216) 375 7848

OVERVIEW required for much of the current instruc- The microprocessor is proving itself tion in the public schools is the display as an attractive, low-cost device which of upper- and lower-case alphabetic char- has the capability to serve its owner as a acters. None of the low-cost micropro- record-keeper, accountant, entertainer, cessors is sold with upper- and lower-case and tutor. The attractiveness of the de- displays, although most of them can be vice has led teachers and other educators modified, at extra cost, to add this capa- to explore its potential for providing city. The use of color and sound can add computer-based instruction. While no de- significantly to the appeal of micropro- vice, whether it be a microprocessor, a cessor-based courseware, but the lack of large computer, or something in between, dual-case alphabetic characters has to be should be viewed as simply good or bad considered a serious limitation for the for education, all devices have charac- teaching of such subjects as English and teristics which must be considered before reading. an investment is made. One might expect that a computer terminal connected to a SIMPLICITY OF OPERATION A public school student using a termi- large computer (resident or non-resident) nal tied to a large computer typically be- might have capabilities that microcomputers lack, and this is certainly the case. gins his session by being seated and typ- Whether or not a school should purchase a ing one short message. At that point, he micro, however, depends entirely upon the has entered the computer-based education course and is working where he left off function which the machine will have. during his last session or where his pre- Following is a discussion of the educational services that can and cannot be vious performance records indicate that he provided by off-the-shelf, low-cost, should be. One CBE course may contain microprocessors such as the Apple IX or enough material to teach the student and track his progress for a year or more. A Radio Shack's TRS-80 (Model I). student using a microprocessor (presumably DISPLAY CHARACTERISTICS with at least one disk drive) typically The variety of characters, graphics, has a more complicated procedure for begirt- and colors that microprocessors are ca- ming his session. He will obtain his disk, pable of displaying has been a major fac- be seated, insert his disk into the disk tor in the success of its sales during drive, power-on the machine so that the recent years.The Apple II and Ohio initial_ program will be loaded into memory, Scientific micros, for exawle, can pro- and theft he is ready to begin,He may or duce graphics, colors, and sounds, in may not have to type a message telling the addition to displaying standard keyboard computer to start the lesson. He probably characters. Graphics are often helpful, cannot start where he left off last time; particularly for the ruching of such sub- however, a carefully programmed lesson may jects as physics and hisher level mathema- enable him to choose his own starting tics, but the most important characteristic point. One disk may contain enough meter!?

223 236 224 NECC 1980

al to tutor the student for a number of degree of individualization is not avail- hours. able today. The educational software cur- In the public schools, students using rently available for microprocessors con- terminals connected to a large host compu- sists of short lessons, each addressing a ter can be given the freedom to use the specific topic.Seldom does any continu- terminals without teacher supervision be- ity exist among these short lessons. To cause the sign-on procedure is simple and employ these microprocessor lessons for because no peripheral materials, such as genuine computer-based education would floppy disks, are required. Use of a mean that decisions as to which students microprocessor may require the monitoring need which lessons at what time would be of a disk storage area and a check-out, made by teachers, aides, or by the stu- check-in systeu for students to borrow dents themselves. Whether these persons disks. Since a standard 5-inch floppy would make the correct decisions, and disk contains only a limited amount of ma- whether continuity could be provided among terial, the student will frequently need the lessons collected, is questionable. to exchange his disk in order to switch to Assuming that continuity could be achieved another topic. students would have to be given paper-and- In terms of simplicity of operation, pencil tests constantly in order to deter- the advantage appears to rest with the mine which lessons were needed by each terminal connected to a large host compu- student and in what sequence. Such tests , ter. are unnecessary on a large computer sys- ADDRESSING A VARIETY OF LEARNING NEEDS tem since branching occurs automatically A typical, well-planned CRE course con- as a result of present and past student tains a sufficient quantity of instruction performance. In a well-designed CRE lesson, students are generally unaware that to meet the needs of a variety of differ- such intricate branching is taking place. ent students: the slow learner, the rapid learner, the student who knows nothing They know only that the computer is moving in s logical direction from the beginning about the topic being presented, the student who already has familiarity with the to the end of topics that they need to learn. subject matter. As the student works through the instruction, his performance TRACKING STUDENT PROGRESS will cause him to be branched forward for Microprocessors and large CIE systems advanced material, backward for additional both have the capacity to provide on-line practice, or laterally for s discussion of assessments, to give the student some in- topics tangentially related to the primary formation about his progrese, and to topic being presented. The number of branch him to various locations based on paths that a student could take through his performance. (As mentioned before, the material is infinite because the the physical size limitations of the structure provides almost limitless bran- microprocessor's disk would frequently ching opportunities. Since no two stu- necessitate disk-swapping in order to dents possess identical needs, it is pro- switch topics.) Out in order to build a bable that no two students will progress genuine performance history on each stu- along the same path of instruction. A dent, a large system is required. Large- computer-based course should be extremely system CRE can relieve the teacher of the flexible, broad in scope, and capable of burden of administering periodic tests and handling students' individual performance plotting each student's growth. By typing styles. Failure to attain these charac- appropriate keywords at the computer ter- teristics constitutes a failure to place minal, the instructor can view immediately CIE into a category eeparate from other any student's current and past performance media such as lectures, textbooks, and records. Typical information includes the audio-visual tapes, all of which possess following: a predominantly linear flow. :- 1) the student's first and last date The above characteristics make present on-line. computer-based instruction impractical, if 2) the total number of hours and not impossible, on today's most popular minutes spent. microprocessors. Microprocessor tapes, 3) topics where mastery has or has disks, and memory have a size limitation not been demonstrated. which prohibits them from delivering a CRE 4) areas of particular difficulty. course with the breadth of scope described 5) areas not tutored due to excellent above. At best, a student using a micro pretest performance. for this type of instruction would find 6) average response time on drill-and- himself swapping disks constantly in order practice items. to branch ahead, backward, and laterally. 7) length of individual sessions. Rut microprocessor courseware with this 8) students' comments on parts or all

V. 4'. 23 7 Pre-College Instructional Materials 225

of the instruction. what the date is or how long it took him This information can be displayed for one to answer a question, and then the stu- student, a group of students, or all student's response can be recorded, but the dents in a teacher's class. Hardcopy ter- reliability of such data would have to be minals are often used so that the report questioned.The absence of an internal may be kept for future reference. Teachers clock precludes the recording of dates or can quickly identify any students having elapsed time by the microprocessor. Other difficulty, and students can become very types of data, such as topics mastered and highly motivated when they see that they scores earned during on-line assessment, have achieved progress from one week (or can be recorded by the microprocessor if month, or year) to the next. the lessons are programmed to do so and if The quantity of student performance a system of data-collecting disks, like data afforded by the microprocessor is the one described above, is established, trivial when compared to the data routinely In short, the collection and reporting kept on the large system.Many (perhaps of student performance data are essential most) of today's tutorial microprocessor in order to track progress, provide student lessons give no performance statistics motivation, and isolate learning problems. whatsoever. Those lessons that do give Large CBE.systems address this need very performance information do not store it well. Microprocessors could provide some permanently for the teacher's later inspec- (but not all) of the same student perfor- tion. Performance data are collected and mance records only after a fairly complica- reported to the student as he is working, ted system of specially written lessons, but when his session is over and he re- student data disks, and multiple disk-drive moves the disk from the disk drive, all hardware is created. Retrieval of informa- collected information is instantaneously tion from this system would be significant- lost. Although virtually no microproces- ly more cumbersome than it is presently on sor courseware exists that builds a his- larger CBE systems. tory of student performance, it is possible to do so. The most practical way to INSTRUCTIONAL LANGUAGES accomplish such a task is to use a multi- Like spoken languages, computer ple-disk system. In a two-disk system, languages are numerous and varied.Many for example, one disk can be devoted to computer languages perform similar tasks collection and storage of performance equally well, but some languages are data while the other contains lessons be- specialized to perform certain tasks better ing presented. In a public school, each than the .others. student would be assigned one or more Computer languages that have been de- disk for storage of his own performance signed to provide instruction are sometimes data, and these disks would be checked in referred to as instructional languages. and out along with the disks containing Such languages typically record some basic the instruction. Instruction would have information automatically for every student to be programmed carefully so that each on every CBE course, such as his first and lesson would store performance data on a last usage dates, his total time on each different part of the data-collecting lesson, and hia current location within disks. An instructor wishing to view the each lesson. In addition, storage performance histories of his 30 students ties are provided to enable the course would sit at the microprocessor with his author (programmer) to store an infinite variety of performance data permanently or collection of 30 or more disks and slip them in, one at a time. As each student's temporarily. A student progress report is progress is displayed, the information merely a systematic display of these stor- could be duplicated in hardcopy form if age facilities. An instructional language a printer is attached to the micropro- also permits the course author to accept a cessor. Obviously, this inspection of wide range of responses from the student. student progress would be a time-consum- For example, if the author expects "false" ing task, and the performance data pro- as the answer to a question but realizes bably would not be as complete as the that students may misspell the word, he can data collected on a large system. use a atat-ment which essentially says, "If It was mentioned above that typical the student types a five-letter word begin- performance data provided on large CBE ning with 'f' and ending with 'e', I recog- systems includes the student's first and nize this as the correct answer." Another last dates on-line, total time spent on characteristic some instructional languages instruction, and average response time have is the automatic re-starting of the on drill-and-practice exercises. Unfor- student at the point where he left off dur- tunately, many of today's popular micro- ing his list session. Characteristics such processors cannot record any of this infor- as these are not found in computer languages that have been designed for purposes mation. It is possible to ask the student

233 226 NECC 1980

other than instruction. tive course. Such duplication of effort BASIC is an example of a non-instruc- is a general characteristic of micropro- tional language, not because it is defi- cessors. Each user has his own computer cient in some way, but because instruction and his own copy of each program he uses. is not the purpose for which it was crea- Fifty users wishing to use a mathematics ted. It is a general-purpose computing lesson would need to purchase 50 copies of language which is standard on virtually that lesson. On a large system, only one all microprocessors. The language has no copy of the lesson is needed, and it can pre-defined storage area for recording be used on any number of terminals simul- student performance data, no provisions taneously. for accepting a wide range of responses, Another important feature of large and no automatic restarting of the student systems is that all users are accessing where he left off last tine. Naturally, the same disk storage area. It is this the BASIC language has capabilities that characteristic which permits teachers to some instructional languages lick, notably obtain class reports on groups of students. in the area of mathematical computation. Such reports often point out the best, Wu if BASIC is to be used on microproces- worst, and average oerformance record of sors for CBE, numerous sophisticated sub- the group (sometimes including statistics routines must be created in order to give such as mean, range, and standard devia- BASIC the characteristics of an instruction), thus giving the teacher an immedi- tional language. These subroutines will ate picture of the group's homo or hetero- occupy space on the disk and permit less geneity. With micro-based CBE, perfor- space for instruction and collection of mance data are likely to be scattered performance data. The subroutines must be across many disks, thus making such group duplicated and stored on every disk used reports impossible or impractical. for instruction.And, in spite of their Large systems often possess a message- sophistication, the subroutines will not handling capacity which is not possible on permit the recording of certain types of microprocessors. Messages can be sent data such as dates and student response between course authors and students, for time unless the microprocessor chosen con- example. A student who is puzzled or who tains an internal clock. wishes to communicate with the author of As mentioned above, most languages can his lesson for any reason can usually log accomplish most tasks. But if a choice is a comment which the author will later read made available, it is only logical to pick and answer. Such communication assists the language which is least cumt.ersome. the author in locating any areas of,his On large systems, the choices are numerous lesson which may cause confusion among and include many instructional languages. students and enables teachers or other With microprocessors, the choices are few supervising persons to post administrative or nonexistent and do not include instruc- announcements to users of the system. tional languages.. On the basis of lan- Some large systems also permit on-line di- guages, it appears that present-day micro- rect communication between terminals. processors are not yet prepared to handle Microprocessors necessitate a duplica- the level of aophistication found in large- tion of effort and lack message capabili- system CBE courseware. ties, but they do provide a greater degree of independence and self-sufficiency. As LARGE - SYSTEM AND SHALL-SYSTEM CHARACTERISTICS mentioned above, a computer malfunction on Regardless of the language employed or a large system renders all terminals use- the performance data collected, there are less simultaneously, but one microproces- certain. inherent characteristics of large sor's failure has no influence on other microprocessors. A micro is also more systems which must be evaluated prior to making a decision to employ or abandon such portable than a terminal connected to a a system. On the negative side, when the large computer since the latter requires large computer malfunctions, all terminals either a hardwired communications line or a telephone line. And, of course, there connected to it become useless. But that are no monthly rental charges after a all terminals are using one computer also microprocessor and peripheral equipment has advantages. A major advantage is the are purchased. ability to avoid duplication of effort. For example, when a courseware revision WHY BUY A MICRO? becomes necessary on a large system, one A microprocessor can serve as a valu- correction constitutes a system-wide able educational tool, particularly in change. But if 50 microprocessors were teaching students about computer hardware, used in place of a 50-terminal network, and can help them develop an understanding then 50 individual corrections would need of what happens when a program is written to be madeone for each copy of the defec- and executed. Mathematics students can 239 Pre-College instructional Materials 227

also profit from a study of the BASIC lan- the defective lesson or data-recording rou- guage, which can be used effectively to tine. solve math problems using algorithms. The above steps vary in complexity Limited amounts of instruction, of course, and would probably require years to accom- can be provided in tutorial or drill-and- plish. The characteristics of such a sys- practice form as well But nothing with tem are contrasted with the characteristics the scope, complexity, and record-keeping of present day large systems in the Summary abilities of large-system computer-based Chart which appears at the end of this education can be accomplished ih a reason- article. able way on today's microprocessors. NOTHING IS IMPOSSIBLE LAUNCHING A CBE PROJECT WITH It is not the author's intention to MICROPROCESSORS state that microprocessors are useless for Assuming that apublic school has con- computer-based education. In fact, most of sidered the pros and cons of the low-cost the undesirable characteristics mentioned microprocessor as well as those of the CBE can be rectified through hardware modifica- terminal connected to a large system, and tion, addition of peripheral devices, or by assuming that the micro was chosen, these simply purchasing a larger microprocessor. are some of the steps that would probably But it is the small, comparatively unsophis- have to be taken: ticated microprocessor that has"been at- 1) purchase a microprocessor. tracting the most attention in the market- 2) purchase at least two disk-drives place. This author becomes concerned when in order to provide the capability teachers and other consumers become en- for storing and retrieving student tranced by those very attractive devices performance data. and when salesmen with little or no experi- 3) purchase floppy disks for data ence in computer-based education make exag- storage and for storage of the gerated claims. Quality microprocessor instruction itself. courseware which has been tested objective- 4) purchase a printer in order to ly and shown to be of educational value is provide student progress reports NOT widely available.Microprocessor in hardcopy form. courseware which provides the individuali- 5) modify the microprocessor so that zation and record-keeping of large systems it will display upper- and lower- is not available at all. Although it IS case alphabetic characters needed possible to buy a microprocessor with upper- for the teaching of spelling, and lower-case alphabetic characters, and English, reading, and related although larger disk drives CAN be used so subjects. as to reduce the amount of disk-swapping, 6) design machine-language or BASIC and although an internal clock CAN be added, subroutines to handle tasks al- such modifications result in a home brew ready designed in instructional system which is costly and which makes the languages. in-house development of courseware manda- 7) copy these subroutines onto all tory.It is this message that the sales- disks to be used for, lesson con- men are failing to convey. tent and/or all disks to be used for data collection. THE FUTURE 8) initialize each disk to be used for Microprocessors will probably play an lesson content or data collection. important role in future CBE systems for at least two reasons: 9) build courseware which branches forward, backward, and laterally 1) Their capabilities will increase. to accommodate individual learning It is hoped that a greater choice of compu- styles and which keeps detailed ting languages will be available along with greater storage capabilities and, conse- student performance records on the data-collecting disks. quently, less required disk-swapping. An 10) establish and supervise a system internal clock will enable the micro to for storing lesson and data- record dates and keep track of elapsed time. collecting disks, keeping track of Upper- and lower-case characters will be which disks are borrowed (and re- commonplace. turned) by whom, and assuring that 2) Downloading will become more each student has access only to his practical. Downloading essentially means own data disk. that the micro can communicate with a larger machine for the purpose of copying a 11) make copies of lesson disks if the project involves more than one program into its own memory or onto its own microprocessor. disk. Then the program can be used locally, 12) when errors are detected, make the independent of the large machine.After corrections on all disks containing the session is completed, the updated

240 228 NECC 1980

records can be transmitted back to the larger computer. It is rather impractical to attempt to emulate large-system CBE with today's smaller microprocessors. A tremendous amount of time and effort would be required, and the final product would contain built-in disadvantages. By the time such a system is built, better micro-based systems will probably be available. At that time a re- evaluation of the enhancements made on microprocessors as well as on large systems will be necessary. DESIRABLE CHARACTERISTICS

Typical Low-Cost Large Popular System Micro

X Involves a one-time cost without monthly rental charges. X Displays graphics and colors, produces sounds.

X Displays both upper- and lower-case alphabetic characters. X Has complete portability, not dependent on a telephone line. X Requires one step (one short message) to start instruction. X Affects only one user when the computer malfunctions. X X Provides performance feedback while the student is on-line. Provides detailed performance history for teacher's later X inspection.

Enables students to continue where they left off during their X last session.

Records dates and students' response_ times.

Records session time and permits instructor-defined session X length for individuals or for all students. X Generates class reports showing histories of groups or students. X Has well-constructed courseware immediately available. Is a system well-adapted for extensive branching to accommodate X individual learning styles.

X Supports instructional computer languages. X Permits nessages among students, authors, teachers, and others.

2 II Pre-College Instructional Materials 229

UNDESIRABLECHARACTERISTICS

Typical Low-Cost Large Popular System Micro

X Involves a monthly rental charge.

X Lacks graphics, colors, and sound.

---Requires a dedicated line or telephone line.

X Has upper-case display only, unless modified.

Requires executing several tasks in order to start instruction.

Requires constant disk-swapping if individual learning styles are to be accommodated.

Requires a disk storage and check-in facility which most be supervised.

Necessitates adding more disks proportional to the quantity of performance data desired.

Requires creating new courseware since virtually none exists presently. Necessitates considerable duplication of effort.

at? 230 NECC 1980

MICROCOMPUTER /VIDEODISC CAI DEVELOPMENT CONSIDERATIONS Ron Thorkildsen Him Allatd Exceptional Child Center UMC 68 Utah State University Logan, Utah (801) 730-3534

Providing an appropriate education fot all specifically by mentally retarded children and children in the least possible restrictive envi- adults. Even though the system is designed to be ronment will require drastic changes in our educa- used by e handicapped, it will have all the tional processes, ftom assessment and prescription components necessary for a general purpose CAI through instruction and monitoring.Recently system.With the appropriate courseware the Benjasin Bloom has stated, "If we are convinced system would be relevant to learners at any level that a good education is necessary fot all who of intellectual ability and expetience. live in modern society, then we 151st search fot The system, teferred to as the MCVD (Micro- the alterable variables that can make s difference Computet/VideoDisc) System, was otiginally devel- in the learning of children and adults in or out oped ftom funds of a small grant received from the of school" (Bloom, 1980). The alterable variables Univetsity Research Office. Subsequently a stoat Sloom emphasizes are time-on-task, cognitive ftom the Media and Captioned Films Division of euttY, and formative t4Stin. These variables the Bureau of Education for the Handicapped of help explain the learning interactions between DREW was awarded. The grant is for two years and teacher and student. Using these variables implies began October 1, 1979. individualizing the entire educational process, The major goal of the project is to develop, ftom assessment to the monitoring of learning. evaluate, and demonstrate a CAI system for use Mainstreaming is currently viewed as a with mentally handicapped learnets.The system is necessity in providing the least restrictive unique since it can commmajcate with nonreaders, environment.Mainstreaming handicapped students a capability nonexistent or very limited in past crestes a special burden for the classroom teacher, CAI applications. Computer-assisted instruction however, introducing handicapped students into systems have been devised and developed to commun- the regulat classroom greatly inctsases the range icate with nonreaders, but in most cases they have of intellectual capacity and experience of the used serial devices such as audio tope recorders students.This broadened range makes the individ- and slide projectors.The difficulty with these ualization of instruction a necessity, although systems is a relatively long access time when individualization will be extremely difficult branching to different segments of an instruction- considering the limited resources available to the al program. A short access time (less than 3 regular classroom teacher, seconds) is ctitical in most CAI applications, Educators have long felt that the computer but especially critical when wotking with mentally holds a special promise in providing individual- handicapped learners since, at best, their atten- ized instruction, but this promise has not been tion is difficult to maintain. fulfilled because of the limitations of audio The hardware fot the MCVD system consists of visual hardware ..id the high cost of computers. the MCA videodisc player, an APPLE II microcom- Recent changes in hardware should alleviate many puter with a digital disk, a Sony 12-inch tele- of these limitations.The microcomputer has vision monitor, and a Carroll Manufacturing light - greatly reduced the cost of computing and the lututruPt touch panel. The MCA 7802 videodisc videodisc has the capacity to provide the audio player was selected becauae of its random access and visual components necessary for effective capabilities, and the APPLE II microcomputer individualized instruction via computer-assisted system was selected because of its pottability, instruction (CAI). reliability, and color graphics capabilities. A CAI system utilizing a microcomputer and A color monitor was chosen because of the addi- videodisc is currently being developed as pert tional flexibility provided by color. The tout.% of a research project conducted by Utah State screen is a light -intetrupt system which allows Univeruity's Exceptional Child Center, The the learner to interact with the system by touch- system is designed to be used by nonreaders and ing the scream. The hardware components are

2.1:3 Pre-College Instructional Materials 231

enclosed in a cabinet that disassembles for trans.- signaled for help. While the learner interacts partation purposes. The monitor will be mounted with the system, data are collected by the micro- on an adjustable pedestal that swivels, tilts, computer and stored on the APPLE II magnetic disk. and can be adjusted for elevation. This flexi- Various data are maintained which track a bility and adjustment is necessary to accommodate leatner's progress through a program of instruc- the wide range of ages and varying degrees of tion. The data are then available for analytic motor ability. The cabinet is mounted on retract- purposes and are also used by the system to allow able wheels to facilitate mobility. the learner to start at a point where the previous The computer programs that control the video- session was terminated. This automatic restart disc and receive input from the touch panel are option can be overridden by the teacher, who can written in Applesoft BASIC. These programs specify the point at which the learner is to were originally written in PILOT, which we aban- continue. The data is readily available in hard - doned because of problems with the PILOT trans- copy form via a Centronics Printer. lator. PILOT was originally selected because of Pour instructional programs are currently in its ease of use. lie are currently exploring various stages of development.These instruction- various PILOT translators and compilers. We have al programs are developed by the Outreach and determined that a PILOT interpreter does not pro- Development Division of the Exceptional Child vide for sufficient execution speed to interact Center and have been field tested and validated. with the three devices.We are also currently They were chosen primarily because the instruc- investigating the potential of PASCAL as a lang- tional sequences had been validated and because uage for the system. they are adaptable to the MCVD format.The Communication between the three hardware following is a brief description of the programs: deilces is accomplished through an interface board 1. Matching Program (Hofmeister, 77) - This that was designed and built by USU's microcomputer program teaches the child.to match objects lab. The interface board also can programmatic- that are alike in_size, colpr, or shape such ally switch the video source to the monitor from as idencifying squares with other squares. either the APPLE /I or the videodisc. This capa- This skill is necessary before the child can bility allows us to transmit APPLE II graphics and learn the more advanced skills such as names audio to the monitor while the videodisc is search- of colors and shapes and reading.This ing. It also allows us to present video informa- program was not designed to teach a child to tion from the APPLE II and audio from the video- to name colors, sizes, or shapes. disc simultaneoua3y, which provides for a great 2. Timetellina Program (Rofmeister, 75) - deal of flexibility in altering the video images This program is to be used with any child from the videodisc. or adult who cannot look at a clock and At present the system interacts with the recognize what time it is to the minute. It learner by presenting an audio instruction and has been especially useful in tutoring the associated visual image on the monitor via programs for the slow learner, although it is the videodisc. The learner responds by point- not restricted to such use. Upon completion ing to an object on the monitor screen. When of this program, the learner should be able the learner touches the screen, two light beams to look at any clock, showing any time, and transmitted from each axis of the touch panel recognize the correct time to the minute. are interrupted, and the point of interruption 3. Recognition of Functional Words is detected by the touch panel.The X and T (Rofmeister, 76) - This program provides coordinates from the point of interrupt are instructions for teaching the learner to transmitted to the microcomputer. The computer recognize functional words. This program program in the microcomputer contains the correct teaches the recognition of twenty different coordinates for each segment of instruction, and words important to everyday living (e.g. stop, the coordinates transmitted by the touch panel go, pull, push, etc.). are compared to these correct coordinates. If 4. Identification of Coins (Rofmeleter, 77) the response is correct, the microcomputer This program teaches the learner to recognize responds by referencing the segment on the video- coins upon sight. disc which contains audio and visual positive Preliminary development of the Matching feedback. Other possible response conditions Program was completed in May 1979. A preliminary are a wrong response, a close response, and a field test was conducted and changes were made to non-response. (A non-response is detected when the system as a result of this field test. A the learner does not respond in a specified second preliminary field test was conducted and period of time. Recorded segments are contained changes were made to the system as a result of on the videodisc for these response conditions as this field test. A second preliminary field test well as a variety of feedbatk, including animation of the Matching Program is currently being con- and motion picture sequences. Each segment of ducte4. It was anticipated that this field test instruction has associated parameters that specify would be tompleted by the submission time of this the number of times a learner must respond cor- paper, however, the field test will be continued rectly to advance to the next segment and the through February 8, 1980, in order to collect number of trials allowed before the teacher is additional data. This second field test was

24.4t 232 NECC 1980

necessary tom (1) evaluate the changes in the and recognize :gushers 1-12. system that were precipitated by the first field 2. Learners must be able to identify basic test; and (2) further define the population. geometric shapes (e.g. circles, squares, The following problem areas were determined etc.). as a result of the first field test: 3. Learners must have sufficient motor 1. Even though the search times were rela- skills to hold and use the pointer. tively short (less than 2.5 seconds), the At the completion of the first half of the attention of the child was lost when the program, learners will be able tom monitor went blank during a search time. 1. Sequentially count and place the numbers 2. The system was giving negative feedback 1-12 around the clockface. to a correct response when the child was 2. Recognize the digital format (e.g. 2: ) required to touch multiple objects on the for the little hand. screen. 3. Determine the value of the little hand 3. The child seemed to lose interest in with digital cues on the clock face and the specific positive feedback segments if they big hand as a distractor. were repeated several times. As a result of information gained during the la an attempt to rectify these problems a second field test of the Matching Program, it was number of changes were made to the system: determined that physical prompting and practice L. A programmable switch was added to the; in_using the. pointer are required before inter- interface board that allowed the computer acting with the abstract objects on the tele- program to switch the source of video from vision screen. The Timetelling Program is being either the microcomputer or the videodisc. developed to involve the teacher (or another This switch enabled the system to present a student) in the beginning phase of the program. computer-generated graphic while the video- When appropriate interaction with the CAI system disc was searching. is achieved, the system takes over, allowing the 2. The algorithm in the computer program learner to work independently with the CA1 used to detect Nultiple responses on the system. screen was changed. We determined that the Based on the unique capabilities of the computer program was not responding rapidly MicroComputer /VideoDisc System, the following enough to detect rapid responses from the courseware design considerations were formulated learner.The algorithm was changed, and a and incorporated into the development of the rapid increase in response time was realised. Timetelling Programs 3. The algorithm used to select positive 1. Although the availability of a second feedback segments was changed so that the audio track affords many instructional segments could be presented randomly. options, such as instruction with different The six children involved in the second voices or in a second language, as course- field test are from the special education class - ware development proceeded both positive roomest the Exceptional Child Center. These and negative, it was determined the second children are functioning at a lower mental devel- audio track should be used to store specific opment level than those involved in the first feedback, to the learner on his performance field test. The purpose of this change was to (see Figure 1).This feedback is specific determine if children at the lower level of to a particular segment of instruction as development could effectively interact with the opposed to the standard feedback blocks system. The only criteria established for which are used by numerous instructional selection was that the children have sufficient segments.Generally, it is the instructional motor skills to hold and point .ith a small sequences at the beginning of the program pointer and sufficient receptive language to that require specific feedback.An example understand simple commands such as "touch the would be "good touching the one." one like this," "that was good," "that was not 2. Strategic placement of the standard correct," and "try again." feedback blocks can substantially reduce The field test has been in operation for the access time required by the videodisc three weeks at the time of writing.During player to search for a feedback segment. this time numerous problems in working with the Because of the nonlinear nature of the MCVD test population have been identified. Since system, standard feedback is used often the field test is continuing, the data c.onot throughout the program. A regression at this point be analyzed and specific con» equation developed by Woolley and DeBloois elusions cannot be made. allows the prediction of access time based The first half of the Timetelling Program on the number of frames (Woolley, 1980). A is planned for production in March 1980.The criterion-of 2.5 seconds was established for following entry-level skills have been estab- the maximum search time for providing either lished for learners using timetelling courseware positive or negative feedback. Based on the (these entry level skills will define the popula- regression equation, the feedback segments tion): would have to be located within 6900 frames. 1. Learners must be able to count to sixty of any instructional segments that access it.

241.- Pre-College Instructional Materials 233

Meeting our 2.5 second search criteria re- Parent Teaching Package - Word Recognition." quired multiple placement of the standard Available at the Outreach 6 Development feedback blocks. Division, Exceptional Child Center, Utah State 3. Periodic testing is facilitated by the University, Logan, Utah, 1976. nonlinear capabilities of the MCVD system. Woolley, R.D. and DeBloois, M.L. "Preliminary The timetelling courseware and microcomputer Benchmark Data for the PR 17820 Discovision software are designed to branch to appro- Associates Videodisc Player." Center for priate practice frames for criterion test- Instructional Product Development Technical ing. This branching insures that the learner Report 11. Utah State University, Logan, will be tested only on the material the learner Utah, 1980. has already covered. Criteria have been established at 00X accuracy.Testing, consisting of II frames, begins and continues only as long as the learner meets the SOX criterion. In other words, as soon as two frames have inaccurate responses, the software branches the learner back to the initial lesson for continued practice. Conversely, should the learner meet the BOX criterion on his or her seventh frame, the last test frame will be omitted. 4. The amount of repetition can be varied for individual learners based upon their previous performance. The learner who achieves a predetermined level of accuracy (determined during field testing) may be branched ahead to the criterion test and given the opportunity to demonstrate mastery of that particular concept. Learners having difficulty with Coe same concepts can be given additional repetition and subsequent prectia:-z,r The two preliminary field tegiti and production experience have yielded valuable information on the development of videodisc-based CAI. Pre- sently one of the most limiting disadvantages of CAI involving the videodisc is the high cost of disc pressing.This presents instructional courseware developers with an entirely different set of instructional development problems. After a disc is pressed, you have to live with it.Therefore, any experience gained from this project or other projects should provide valuable information in developing videodisc programs. Essentially, a developer needs all of the up -front knowledge available before a disc is pressed.

REFERENCES Bloom, B.S. "The New Direction in Educational Research."Phi Delta Keegan. 1980, Vol. 16, No. 6. Nofmeister, A.M. 6 Gallery, M.AProgram for Teaching the Identification of Coins. Niles, Illinois: Developmental Learning Materials, 1977. Nofmeister, A.M., Gallery, M. 6 Landon, J.J. "Matching Sires, Shapes and Colors." An Instructional program prepared by the Excep- tional Child Center, Utah State University, Logan, Utah for the Utah Division of Family Services, 1977. Nofmeister, A.M., Atkinson, C.M. 6 Nofmaister, J.B. Programmed Time Telling. Eugene Oregon:E.B. Press, 1975. Eamaister, A.M., Patten, M. 6 Rosen, A. "A 234 NECC 1980

Audio Track 1 Audio Track 2

Video

Control Track Sync Track Audio Mad( 3 One Segment of Instruction- 4 Seconds on Vk isotope 120 Frames on Videodisc

Directio of Tape Travel One Inch Type C Helical Videotape Used For Initial Production

Figure 1

2A. Linear Framework of Timetelling Package

;i1A4. 1. AY. , 4, ....f.:::: instruction Instruction Instruction 1,417`c Instruction _1 v.v.42 \

Standard Feedback Block ISFBI

Figure 2

2-1s 236 NECC 1080

THE NON-TECHNICAL FACTORS IN Tat DEVELOPMENT OF CAI

Michael Mocciola Computing Center/Academic Computing Pace University New York, New York 10038

ABSTRACT The most challenging opportunities and the most serious problems that inhibit the development of computer-aided learning in education are non-technical. In most educational institutions decision making, communications, and bargaining about the funding of computing, the purchase of computer equipment, or the application of computer methods is divided among the major populations of the schools; school board members, administrators, teachers, department heads, and students. To understand the socio-political factors related to computer-aided learning, the job-related responsibilities of each of these populations must be carefully examined. Computer educators then need to use the socio-political forces to the best advantage of students and teachers to provide enhanced classroom instruction. Educators must recognize and foresee the needs of the schu'l community, have early access to pertinent information, and communicate effectively.

21o; Invited Sessions

DATA SETS AVAILABLE FROM THE FEDERAL GOVERNMENT

chaired by Thomas E. Brown General Services Administration National Archives and Records Service Washington, DC 20108 (202) 724-10,d

ABSTRACT SPEAKERS: Many agencies of the O. S. Government EFrationcn, Coordinator, College Cur- routinely make computerized data available riculum Support Project, Bureau of the to educators and researchers. To do this, Census: a discussion of the services some agencies provide reference service for which the Bureau of the Census provides the data which the agency created, other to enable researchers to gain access to agencies distribute data originally created data collected by the Census Bureau. by other agencies, and still ether agencies Ross J. Cameron, Archivist, National serve as a clearinghouse for a given sub- Archives and Records Service: an outline ject and refer researchers to agencies of the functions of the Machine-Readable having information on that subject. This Archives Division in its mission to inven7__ ""session will have representatives from toffy, obtain, and provide reference ser- three agencies performing these three vice :or data created by other Federal functions. At the end of the session, one agencies. should have a clear indication of the pro- Edward D. Mooney, Program Specialist, cedures involved in locating and obtaining National Center for Educational Statistics: data sets from the Federal government. a presentation on the Federal Interagency Consortium of Users of Educational Statistics whose purpose is to facilitate access to education data in the Federal government.

237 250 Minority InstitutionsECMI

ACADEMIC COMPUTING: A SAMPLER OP APPROACHES IN MINORITY INSTITUTIONS

Sister Patricia Marshall Xavier University of Louisiana 7325 Palmetto Street New Orleans, LA70125 504/486-7411

A variety of post-secondary minority year, three - fourths black institution es- institutions using a variety of approaches tablished after World War II. It enrolls to academic computing were interviewed over 9,000 students in day and evening campus-wide and in-depth late in Fall 1979. dictator Two campuses, one emphasising These interviews were part of a larger as- technical studies and one evolving toward sessment of needs in educational computing a focus on business studies, are united at 239 minority institutions.(1) under one administrative structure. Large numbers of minority students began to at- Faculty, students, and administrators tend this school early in the 7Os as a ..were_interviewei_on computing development, result of outreach by the institution. usage, problems, and successes.Diverse The stable faculty consists of only 28 approaches were discovered, corresponding percent minorities.Although this insti- to philosophic, demographic, geographic, tution may appear small by national stan- Historical, political, and cultural fac- dards, it is one of the largest of the tors. minority schools and one of the few two- year colleges with any large degree of The institutions were selected in order experience and planning in academic com- to obtain as broad a cross section as pos- puting. sible within time and financial limits and parameters such as ethnic composition, type Unencumbered by the weight of tradi- of control, date of establishment, highest tion and spurred by local employment level of offering, academic orientation, needs, the Community College of Baltimore enrollment, type of access to hardware, and had established a separate academic de- experience. While not all of the most suc- partment for data processing by the early cessful institutions were chosen, we did date of 1965. Thia department, now called try to include schools which would not feel Computer and Information Systems (CIS), threatened by the interviews and which had attracts 300 majors and graduates about 25 had enough experience to identify factors each year with an associate degree in com- inhibiting and promoting progress. Con- puter studies. Four full-time and numer- sciously excluded were such institutions ous part-time faculty staff the department. as the Universities of Hawaii and Puerto Rico; partly for financial reasons and The school has weat'ered tnree hardware partly because they are so much larger than phases and is entering its fourth. Ini- minority institutions in general. The tially an IBM 1620, acquired for the data table on the next page lists the institu» processing department in 1965, was the tions interviewed and some key parameters only computer. A committee from electron- used Li their selection. For confidential- ics, data processing, and administration ity we identified them only by number.We hired a manager of computer services and will report here on four of them; these are developed specifications for a UNIVAC 9300, named below with permission. which arrived in 1968. The college allo- cated its funds to lease it and pay support staff for administrative services. Depart- Institution #3 ments and individual users, however, have never been charged. The Community College of Baltimore is au Eastern seaboard, public, urban, two- 1972 brought a UNIVAC 9480 (131K) with

238 Zvi Minority institutionsECMI 239

TABLE 1

MINORITY POST-SECONDARY INSTITUTIONS INTERVIEWED ON ACADEMIC COMPUTING, FALL 1979

HIGHEST LEVEL FALL ETHNICTYPEOFOFOF- ORIENTATION 1978 TYPE OF HARDWARE /NWT-DATE TYPE CONTROLRERINGLIBTECHENGICARENROLL- ACCESSIBLE BY STUD - TUT/ONESTAB. BSI PUBPRI'2'44MART/VOCNEERERRHENT ENTS AND FACULTY

1 1947 x x x x i'I 9152 UN/VAC 9480 ,

. 1 IBM 1620 APPLE II (5, used / as terminals and stand -alones) DEC 10 (remote)

2 1873 600 HP 2000 /BM 1130

3 1969 x x x x 876 DEC 10 (remote) PDP 11/70 (remote) PDP 11/34 IBM 360 (remote) IMSAI 8080 TRS-80

--, 1891 x x x x x 5,395 DEC 10 HP 1000 Micros

5 1884 BIA x x 1,013 DEC 11/34 IBM 1401

6 1867 x x x x 1,526 DEC PDP 11/34 IBM 1130

7 1968 x TRIx x 839 DEC PDP 11/45 BAL

8 1881 .. x x x x 3,296 HP M2000/158370/158 (remote) SOL (3, micros) SWTP (micro)

9 1966 xx x x 4,315 IBM 3031 (remote) IBM3033 (remote) AMDAHL 470 (remote)

NOTE: Campus-wide interviews were conducted at the institutions listed above as part of a needs assessment of educational computing in minority post-secondary institutions.

ETHNIC TYPES: B - Black S - Spanish speaking 1 I - American Indian

1 HIGHEST LEVEL OF OFFERING: 2 - Two years .- 4 - Four years 4M - Master degree

252 240 NECC 1980

four terminals for administrative data en- moral support. The evaluator who conducted try and inquiry. COBOL, RPG, ernd FORTRAP the interviews at this community college were supported for students, as well as commented: ASSEMBLER (on the 9480 and the 1620). This UN/VAC is still used for student These individuals represent a batch runs. In addition, ten ports were scarce resource at any institution rented in 1975 on a companion community and seem to represent a necessary college's HP 2000, an arrangement that was if not sufficient condition for terminated in 1979 in favor of access to a progress in academic computing. DEC 10 at a nearby private university. (A broad base of faculty awareness Students use ten DecWriter terminals which is probably not a necessary condi- are dedicated during labs for a course in tion for groiEg in an institution's BASIC and used for other courses during computing capabilities.) open times. Periodic ad hoc committees on computer The CIS faculty purchased five micro- utilization WEISii exemplify the planning computers in Fall 1979 through a Title VI mechanisms for computing established within grant. These have replaced the 1620 in the college. Plans also exist to establish teaching assembler languages and computing a line position for a director of informa- principles. A fourth system is under con- tion systems reporting to the president sideration now because of the two campuses, through a dean of planning, development, growth of institutional and faculty re- and communications. This new position search, and growing concern for academic would assume responsibility for academic, computing. The institution had done its as well as administrative computi:3. own local needs assessment before considering such a step. Tensions have existed between administrative and academic computing, as at many Thus, the Community College of Balti- Other institutions. The manager of comput- more hoe steadily built its computing cap- er services reports to the vice-president abilities, if at a slower pace than some for administrative services. The CIS De- majority institutions certainly also at a partment, on the other hand, is within the faster pace than most minority and/or two- Division of Business, Secretarial, and Com- year institutions. This stable growth has puter Sciences under the vice-president for resulted from support by the college of academic and student affairs. Support increases in computing capabilities through staff have grown through several stages to its own funds, supplemented by aggressively the current manager of computer services, sought federal grants. Because of its three programmers, two keypunchers, and two careful planning and budgetary practices, computer operators. No students are em- the school will be able to assume ancillary ployed since the manager claims "it doesn't costs when grants expire. work."

Factors influencing the expansion of Students submit batch jobs through a computing capabilities at this institution slot in a. door.Output pickups are sche- have been the local employment market, size duled twice daily for fifteen minutes each. of the student body (necessitating automa- However, during the two lab hours assigned tion in record handling and giving early each course in a batch-mode language, exposure to administrators and faculty), turnaround time is closer to immediate. support by administration, key faculty mem- Students, who probably know of 40 other bers with interest and dedication to sup- alternatives, accept the pickup arrange- plement the budget for computing resourcea, ment, complaining only about the number of 7 and explicit planning mechanisms within the keypunches. Terminals are available for college charged with studying computing learning BASIC, and the APPLE microcomput- needs. ers were made available in Spring 1980 for learning ASSEMBLER. The commercial option Administrative support is evidenced by in CIS will be offered at the campus near- hiring practices, release-time practices est the business community and the scien- (for proposal writing and planning), and tific option at the other campus.Academ- activities involved in acquiring computing ic applications exist also in the business equipment. A few key faculty nembershave department (managing data on patient care visited other successful sites, attended and student performance). CMI(3) is ex- conferences (ECMI(2) and others) and con- pected to be developed in remedial reading ventions, served on committees in profes- and CAI(4) in science courses after the new sional associations, sought external funds, configuration is installed. Faculty with a and lent one another intellectual and good track record in attracting federal Minority InstitutionsECM1 241

grants will be responsible for developing for both administrative and academic pur- CAI at the science learning center. The poses. It was also used cooperatively by project director for the science learning two other small, church-related, primarily center had attended an ECMI conference and white institutions in the same city. All attributes his current grant, in part, to three schools used it for registration that experience. processing. Five years ago a Title III grant made possible the purchase of the HP 2000 with fifteen terminals for CAI de- Institution #2 velopment. Basic skills disciplines ac- counted for early CAI use; CAI is built Bennett College is a private, women's into course requirements in these disci- four-year, historically black, liberal plines. arts college with a "ILy, though stable, enrollment of about by°.Located in The success here with CAI has given Greensboro, a mid-Atlantic city, the in- rise to new problems.Students need twice stitution houses its computing hardware in as many terminals, and the 1130 is too one attractive location on campus. This small and is no longer supported by IBM. hardware includes an INK IBM 1130 batch sys- Additional personnel trained in computing tem used for administrative purposes and are needed. Release time for faculty is student programming coursea and an inter- needed to develop additional CAI course- actsve HP 2000 with fifteen terminals used ware. Despite the success, the support of extensively for CAI tutorial and drill and the president, end-the small size of the practice in mathematics, English, and biol- institution, some faculty are still una- ogy. Two of four keypunches are available ware of the potential in their disciplines. to students. In such a small school, this is seen as a problem by faculty who do use the computer. Computer center hours are 8 a.m. to 10 p.m. (Saturdays on request), and consulta- Attendance at ECMI conferences helped tion is available during these times. to develop a strong team of faculty, how- Freshmen receive computer ID numbers at the ever, who have used instructional computing beginning of the school year, but upper- heavily.Througn a federal grant the in- class student* have open-shop, direct ac- stitution began to share its expertise in cess to both computers. They run their own Spring 1980 by conducting a regional con- decks and retrieve their own output immedi- ference similar to the ECMI conferences. ately; however, lines do develop when the Bennett collaborated with North Carolina business office is running at the same time A & T University (Institution #4) in con- on the 1130.Tables are provided students ducting that conference. Thirty small in the computer center for such use as colleges participated. examining output and correcting programs. One faculty member cited CAI as ex- Reports on the success of the CAI pro - tremely helpful to entering freshmen, so j ect, in which faculty prepare courseware many of whom are in need of remedial work (or modify existing courseware) using IDF, due to inadequate preparation in secondary have been given at CCUC(5) and ECMI(4) schools. Students said they appreciated conferences; a great deal of consulting and the CAI but not the downtime. The presi- sharing with other colleges has also taken dent of the institution would like to see place.In addition, this college has been separation of administrative and academic cited as an academic computing exemplar by computing.Plans for upgrading the exis- HumRRO(6). ting hardware are on the drawing board.

Computing began at Bennett College ele- Despite its small size and its depen- ven years ago. A rented teletype connected dence on hardware and software that are the college to TUCC(7) and provided for less than stets-of-the-art by today's some instructional work, as well as for im- rapidly changing standards, Bennett Col- proving computer literacy of the faculty. lege has taken a position of leadership Influential in establishing the initial in the development of transportable course- capability were the president of the col- ware for use in basic skills courses. Suc- lege (who later attended an ECMI confer- cess has brought with it new problems, but ence) and the mathematics department chair- it has also nurtured confidence, plans, and person (who later became the computer cen- determination to solve the problems. ter director). A year later, the college purchased the 1130 through a federal grant in connection with another university. This computer was used, from the beginning, 242 NECC 1980

Institution #3 Two instructors are employed in the IDP program, one the chairman. An additional The Rio Grande Campus of Tens State slot is open but not filled. The two in- Technical Institute is a technical-voca- structors bear heavy teaching and lab tional, two-year, Hispanic institution lo- loads, with most emphasis on tab work. cated in Harlingen, close to the Mexican They share the burden of supervising the border in the southernmost tip of the facility. Second-year students are trained state. One of four such institutions com- not only to program (and maintain programs) prising a system in Texas, this school was for some administrative and academic appli- established on a World War II air force cations, but also to assist as consultants base in 1967. Its fast-growing student to first-year students when instructors are body numbered well over 1,100 at the time unavailable. of the interviews, and some of the growth is in the data processing program. The /DP chairman has spent much of his own time on outreach to other departments, Curriculum design is a high-priority providing demonstration projects, seminars, and on-going activity at the school, as and classes. The degree of interest from evidenced by its special stuffing anJ the other departments likely to use the comput- existence of school-industry cooperative er for instructional purposes ranges all committees and advisory committees. An- the way from "take it away" to "you can't nual evaluation ensures that curricula are begin to meet MY needs."Those least in- up to date and graduates are well prepared terested are traditional inatructors. In for employment. (Some students command programs T.% individualized instruction entry-level salaries as high as $18,500 (or in u instructors have had previous without graduating.)Just minutes away experien-e), faculty are eager to attempt from this institution are vast farmlands computer-managed instruction or computer- on which thousands of Mexican Americans aaaisted test generation.Attitudes appear barely survive. Thus the existence of an to stem from educational philosophies industrial corridor and this institution rather than from familiarity with areas of to serve it is pivotal to economic change expertise, such as nuclear technology or in the area. (On the first day of the mathematics. interviews here the Wall Street Journal stated that this region was one of the Two weeks after the move to the new four fastest - 'rowing industrial areas in building, when the interviews were conduc- the country.) ted, a terminal room contained ten CRT's, two teletypes, and a line printer in fairly A new building houses the computing constant use. A Radio Shack TRS-80 and an and electronics programs and computing /NSA' 8080 microcomputer were also avail- facilities, as well as some other mathe- able. Six additional CRT's, a printing matics and science or technical programs. terminal, a digital plotter, and a tape Nearly one-third of ita 17,000 square drive were also being readied for use. feet is occupied by computing facilities Three keypunches with acoustical shells and classrooms used in the industrial were available. Bulletin board displays data processing program (IDP). Prior to included charts, lists, schedules, comput- the interviews (just two weeks after the er-generated Mona ',isms, and a sign pro- move to the new building) less than a claiming "The Dirty Dozen." ("The Dirty fourth as much apace had been available. Dozen," it turned out, were the secondyear students who had survived out of a IDP students dominate the facility, much larger original field of beginners which, like the entire campus, is outin the IDP program.) atandingly clean and well organized. Terminals are dedicated to students, or The terminals were on-line to a DEC to entire classes, from 8 a.m. to 5 p.m. FDP 11/70 located in a department store ften the facility is open till 6 be- 45 miles away. COBOL& not the latest use the IDP chairman stays late. Since (1969), and RPG card (Mks are sent to a

. t students are Mexican-American under- DEC 10 at a university 50 miles away, with gr.Iuates who live with their families in a turnaround time of weeks. But students the area and are not accustomed to being are obviously learning and being hired. away in the evenings, the facility is not The IDP department, having grown from four kept to at night.A demand does exist, students in 1974 to more than seventy at however, from working adults in the area various levels of advancement at the time for an evening program. Finances and of the interview, has gone through several staffing are the obstacles to be over- upgrades of remote connections. Current come. plane call for an on-aite computer with 32

2 S6 Minority InstitutionsECMI 243

ports for student use. However, instruction #2), this school was the last (and tors who want to support individualized in- only minority) institution to procure a struction, record keeping, and test genera- mainframe computer through the National tion in open-entry/open-exit courses feel Science Foundation's original Office of they may still not have sufficient capacity Computing Activities. That computer was since they would be in contention with IDP a CDC 3300, and it followed the first com- for use of the resources. Meanwhile, ad- puter acquired in 1964, an IBM 1620. It ministrators are working on additional ca- was replaced by a DEC 10 in 1977. Federal pacity for administrative work (remote ac- and state funding was combined in each of cess to a large state institution's network). the latter two cases, with the CDC being sold to add to the funding for the DEC. One instructor has done some work of his own, some oz which was destroyed in one of Although some impetus came from the the upgrades.Most recently he has been computer center (especially recently), a using a TRS -80 microcomputer on his own key role was played by former dean of arts time. He envisions a cluster of micros in and sciences Arthur Jackson, after whom a classroom, which he sees as a cost-effec- the computer center is now named.More tive solution to his problems in a tradi- than 80 terminals are on-line to the DEC tion-oriented department.The electronics 10, 26 of them in the computer center.A program, which trains almost a hundred variety of languages are available: BASIC, students for customer engineering on DEC PASCAL, APL, COBOL, ALGOL, LISP, and oth- equipment dedicated to the program, can ers. The computer is available around the use twice as much hardware. The chairman clock all week. The staff includes nine of this program, a former ECMI(2) parti- programmer/analysts, and experts are cipant, keeps up with hardware period- available to faculty and students as icals and literature but is far too busy needed. Students access the computer in- to move into academic computing generally. teractively through class accounts or individually (both authorized by departments) Students who are beginners tend to see as well as by batch jobs.A monthly report no problems with existing hardware and provides usage information but is not cur- software, but the advanced "Dirty Dozen" rently used for charging.Most use is by talk like data processing managers. Com- students in coursework, but faculty do re- pletely at ease in the jargon and what's search, and administrative applications behind it, they speak knowledgeably about abound. Turnaround time is good except at the shoreconings of the available soft- peak times.Jobs are limited to fifty at ware, the need for this version of a lan- a time, which can cause a connection delay guage and that many ports, and even the of ten-to-fifteen minutes.Work space in additional justification needed in the the computer center for students is also current proposal for new equipment. limited.

Administrators give moral support, but During the past two years North Caro- their budget requests have to move through lina A & T University has experienced what several layers of state bureaucracy and the computer center director refers to as compete with other technical institutions. "a quiet revolution" in computing. From a Even when funds exist, it is difficult to single-job, batch-mode machine with 49K find qualified staff in the area who sre main memory and 25 megabytes of disk, the not already working for burgeoning indus- university took a quantum leap to 256K, tries at high salaries. Nevertheless, the 600 megabytes of high-speed disk and the industrial data processing program grows, ability to process up to fifty jobs at a and other departments are beginning to time in many languages interactively, in voice their needs. Among two-year tech- batch mode, or in combination. nical institutions, TSTI-Rio Grande may be on its way to becoming an academic comput- However, increased demand for computer ing exemplar. time has lowered response from excellent to poor. Engineering, which six months before the interviews could run circuit Institution #4 analysis, finite element, operations research, and other sophisticated jobs at North Carolina A 61T University is a any tine,'must now normally execute many state-controlled, four-year, master degree- of these between 5 p.m. and midnight, and granting, black institution, with liberal several between midnight and 8 a,m. Engi- arts and some engineering emphasis.The neering school jobs also include CAI pac- enrollment is 5,400 students, Locatea 4C kages needed for educationally disadvan- Z1188 town from Bennett College (Institu- taged students, and it is difficult to

256 244 NECC 1980

find time during the day to run them. Conclusion Plana are afoot for a memory upgrade at 8100,000, half of which must be raised In the cases of the institutions des:- from outside the university budget. Fur- cribed above, though a variety of approach- ther upgrades will probably include a re- es have been made toward academic comput- mote-site laboratory and a processor up- ing, certain factors surface as common grade. Staff and space needs were cited ingredients of success. These include in addition to hardware and stations for campus-wide planning (or at least plan- remote access. ning beyond the walls of a single department or class), dedication on the part of An active computing advisory committee, key faculty or administrators, careful chaired by a chemistry professor, func- budgeting practices, the ability to put tions as an agent of change and supple- together funding from various sources, the ments positive pressurf.s from all types of ability to learn by experience (as well as users and administrators.Proposals have by capitalizing on the experience of oth- been written, and a computer science deg- ers), and the will to get maximum mileage ree program is to be launched in Fall 1980. from the resources at hand. Interesting- An academic computing director and an ad- ly, historical factors that could have de- ministrative computing director will also feated some actually seem to have caused be hired. these institutions to try harder.

Some usage came about through participation of faculty at ECMI conferences(2), where they learned about MINITAB, test as- References sembly and scoring, test item banking, and CAI, all of which are now used.Although (1) Needs/Strategy Evaluation of Minority most usage is by the engineering school, Institutions in Educational Computing other departments (especially mathematics, (NSF Grant No. SPI-7821515, in prog- chemistry, and business) are becoming ac- ress) tive users. The list has expanded each month in the last two years. Electrical (2) ECM (Educational Computing in Minor- engineering has also developed several ity Institutions): three working microcomputers for instructional use and conferences end one workshop for an HP 1000 that is used for instructional minority institution faculty in 1975- purposes and research in upper-level cour- 1977 to assist faculty with little or ses. no knowledge of educational computing to learn about it in their own disci- Instrumental in the growth of computing plines. Courseware was developed by in engineering at this university has been a small group at one summer workshop. an engineering accrediting agency, according to some faculty.Employment of faculty (3) CMI: computer-managed instruction by local industry was a factor in initially creating awareness of computer potential (4) CAI: computer-assisted instruction among engineering faculty. ECMI conference attendance is credited with increasing (5) CCUC: Conference on Computing in the awareness among mathematicft, chemistry, and Undergraduate Curricula (annual, business faculty. 1970-1978)

Political problems associated with fed - (6) HumRRO: Human Resources Research Or- eral. efforts at encouraging integration ganization have caused difficulties in maintaining orderly development of academic computing, (7) TUCC: Triangle Universities Comput- according to some, Also cited as a nega- ing Center tive factor was the pattern of federal funding which historically favored main- stream institution and left out minority institutions. This is one of the few minority institutions, however, that managed to begin to beat the system as early as the '60s. Minority InstitutionsECMI 245

COMPUTER USE IN CHEMISTRY AT A MINORITY INSTITUTION

James D. Beck Department of Chemistry Virginia State University Petersburg, Virginia 23803 (804) 520-5481

INTRODUCTION titative part for incoming freshmen. Many students in minority institutions Many incoming students have had limited encounter severe difficulties in introduc- exposure to science in high school. As tory science courses. These difficulties an example, 16 percent of the students in often prevent minority students from ob- a general chemistry class this past year taining degrees in scientific fields and had not had any chemistry in high school. thus effectively shut them out of careers The problems of inadequate prepara- in the sciences, engineering, and the tion are compounded by the student diver- health sciences (1). These students are sity which is encountered in many courses. often classified as "underprepared," a Incoming freshman chemistry majors, for classification which implies that their example, had reading scores ranging from backgrounds, in science and mathematics 9.0 to 15+, SAT verbal scores ranging especially, are not strong enough to en- from 240 to 480, and SAT mathematics able them to succeed in rigorous, quan- scores ranging from 280 to 580. This titatively oriented college courses. range of abilities, backgrounds, and in- A number of characteristics of these terests presents a tremendous challenge underprepared students have been identi- to an instructor. fied (1,2). These range from poor mathematics background and lack of exposure INSTRUCTIONAL APPROACH to science in high school to poor study For several years we have been exper- habits and a lackyof self-confidence. imenting with various approaches to the Poor reading ability and a general weak- teaching-learning process in trying to ness in communication skills appear to meet the varying needs of our students be significant factors which inhibit who enroll in chemistry courses. Most of success fcr these students. Many of our efforts have been centered on our these students have low personal stan- general chemistry course. This is the dards for academic achievement, expect chemistry course with the largest enroll- failure, and fear science courses. Al- ment and the one in which the problems of though numerous approaches aave been diversity and weak backgrounds are most taken to improve the performance of pronounced. The approach we are currently underprepared students in Acience using employe the computer in several courses, limited use has bo4n made of ways (3), and we are exploring new and ex- computer-based learning modes. panded types of computer use. In general, Virginia State University is a we try to make available to our students four-year stated supported institution a wide spectrum of different ways to located in Petersburg, Virginia. About learn chemistry in the hope that some of 3400 full -time students are enrolled, these will meet the needs of all of our with over 95 percent of the undergrad- students. The variety of materials and uate population being black. Reading methods allows students to select learn- testa given to incoming freshmen have ing modes which suit them while permit- shown that most of them have serious ting students who are not well-prepared reading problems, with nearly 70 percent to engage in extra activities to help of them reading below the ninth grade them master the material. level. Median scores on the Scholastic Our approach has been to retain the Aptitude Test were ;bout 300 on the traditional components of our general verbal part and about 350 on the quan- Chemistry course--textbook, lectures,

25s 246 NECC 1980

recitation, and laboratorysessions. To considered to be less useful than the com- these we have added a selection of supple- puter programs, problem sets, voluntary mentary materials and activities, includ- practice sessions, and media programs. ing a study guide, sets of performance ob- The copies of old examinations were ranked jectives, lecture otes, slide/tape and nearlyeven with the practice sessions. filmstrip/tape programs, voluntary practice Students nearly all desired to continue sessions, copies of old examinations, and having a variety of alternative methods suggested plans of study. The computer and materials available. has also played an important role in this smorgasbord of learning activities. About COMPUTER USE fifty interactive computer programs are The instructional approach has been available for student use; most of these described in some detail to emphasize the are drill-and-practice or simple tutorial integral part played by the computer. Al- programs. A few are simulations. The though the computer is not an essential computer has also been used to generate part of the instructional process,it is individualized problem sets for student a valuable component.The computer can use. These are used off-line and the an- do some things better than others, but it swers may be submitted for grading by the cannot do all things well.It should be instructor. looked on as one of many instructional Students are free to select which of modes available to the teacher, appropri- these alternative learning strategies they ate for some students studying some types wish to employ. No specific activities of material in some situations. When the are required, but students are expected computer is made a part of the instruc- to complete a certain number of activities tional package, its use should be care- during the semester. Students are encour- fully planned with consideration given to aged to do more than the minimum number, the learning objectives that are involved. especially if they exhibit weaknesses in The use of the computer in this pro- identifiable areas. ject is not new or unique. However, these We have evaluated the effectiveness types of use can be duplicated at nearly of these learning activities in two ways. any institution, even a small college, a One involved a simple frequency of use minority institution, or a high school compilation in which we counted the num- where extensive computer facilities may ber of times that students selected a not be available. The methods also offer particular type of activity.The other the potential of expansion to more elab- evaluation method used student surveys to orate and sophisticated computer applica- ascertain the students' perceptions of the tions. relative usefulness of the activities. Most of the interactive BASIC programs Both studies produced similar results. that we have used were written at Virginia The interactive computer programs were State University and tailored to the needs used heavily and were perceived by the of our students. While simple in concept students as being very useful tothem and in construction, the programs have in their efforts to learn chemistry. proven to be quite useful in this setting. The computer-generated problem sets were They have been designed to be short and used by nearly all students and were also single-concept in nature, each one re- perceived as being very useful.The quiring that the student spend only a few voluntary practice sessions ranked minutes at the terminal. Most are open- slightly below the computer programs and _ended, allowing students to obtain addi- problem sets in extent of use and in tional practice as theyfeel that they perceived usefulness. The slide/tape need it. In general, the programs cover and filmstrip/tape programs were rated basic topics which must be mastered in still lower. order to solve more complicated problems It is interesting that every activ- and understand more complex relationships. ity was used by some students and that A few programs are simulations of chemical every activity was rated by some to be situations. "very useful." It is also interesting These interactive programs do not use that while all of the non-traditional graphics or special effects and generally items that were described above were use the most common BASIC instructions. considered, on the average, to be Thus, they will run on nearly every com- "quite useful" to "very useful," the puter system and can be used with either highest student ratings were given to CRT or hard-copy terminal. Some of them the lecture notes and to the lectures. have been adapted for use on several of However, the textbook, study guide, the popular microcomputers (4). Although recitation classes, suggested plans of the programs may not be suitable for use study, and sets of objectives were all in all environments, they are simple

227 Minority InstitutionsECM 247

enough that they can easily be modified. Although we have been using computer- The generation of individualized generated individualized problem sets for problem sets has also been done in a very several years, we have just begun to ex- simple manner. Two different approaches plore the possibilities involved with have been tried here. In one we have gen- using them. For example, the program erated the problem sets in batch-mode, that is used to generate the problem using a FORTRAN program which merely se- sets can also be used to generate indi- lects questions at random from a question vidualized repeatable exams or quizzes. bank. Any system could use this method Computer - grading, either on-line or off- since a similar program could be written line, is'a possibility. in any language that has a random number Problem sets offer some significant function. In the other method of genera- advantages over interactive programs. tion, individual BASIC programs have been They do not require extensive connect written for each problem set. In these time, since the actual work is done by the random number function is used to se- the students off-line. They permit the lect individual questions and to generate inclusion of topics and problems which individual data within questions.With are not appropriate for interactive pro- this method, we can run the problem sets grams. For example, multi-step problems in batch mode or have each student ob- that require extensive calculations can tain a copy from a terminal. Hard-copy be included on the problem sets. Ques- capability is obviously required for this tions that require the use of the text- mode of generation. book or of other sources of information are suitable. In fact, there are almost EXTENSIONS no limitations on the types of questions While the two modes of computer use which can be included on a computer- that I have described are relatively sim- generated problem set. If grading is ple, they can be adapted to fit a variety not a problem, even essay questions are of instructional settings, they do not possible. require special hardware, and they have One exciting potential use of com- been proven effective. There is, however, puter-generated problem sets is for diag- a great potential for developing more so- nosis and prescription which we have phisticated, and perhaps more useful modes tried on a limited basis. After the stu- of computer use, building upon these sim- dent has completed the problem set, an ple applications. We have explored some interactive program is accessed on a ter- interesting extensions at Virginia State minal. This program checks some or all University and plan to try out more new of the student responses. At this point, things in the future. some of the advantages of an interactive Prior to September 1979, our inter- mode can be realized. Immediate feedback active programs were limited somewhat by can tell the student which answers are our terminals--slow and noisy IBM 2741 correct, and the correct answers or meth- hard-copy units. Installing IBM 3278 CPT ods of calculation can be presented. A units and 3287 printers has increased our course of action can be prescribed to capabilities tremendously. By reducing help the student overcome the identified the time required to run each program, we weaknesses. Prescribed activities may have been able to increase the number of include reading the textbook, doing ad- programs available and allow students to ditional problems, running interactive spend more time in remedial or advanced computer programs, viewing slide/tape work.Although printers are available, programs, or seeing the instructor. students do not generally need to get These prescriptions are individualized hard-copy for their runs, since their re- and are based on the student's perfor- sults are put into a file and can be re- mance on the problem set. In addition to trieved by the instructor for grading. providing useful information to the stu- Although our new terminals do not dent, the results can be made available have true graphics capability, their rap- to the instructor, enabling areas of id print speed has enabled us to program weakness to be identified, for individual material which requires more extensive students and for the class as a whole. layouts than we were able to use pre- This capability is similar to the TIPS viously. This equipment has expanded the program (7). realm of programming possibilities to The problem sets might also serve as some areas that we could not consider be- a core for a system of computer-managed fore the change in hardware.We would instruction. We have been exploring like to have the capability to use real some different ways to use them and have graphics; this would undoubtedly enhance been encouraged by the results.Our stu- learning even more (5,6). dents have responded positively to our

26o 248 NECC 1980

primitive attempts at diagnosing problems Harold Weinstock, editors, NATO Ad- and prescribing remedies, and we hope to vanced Study Instatute, Leyden, Neth- enlarge this effort in the near future. erlands, p. 21 (1977). Several years of working with com- 9. Lower, Stephen, Gerhold, George, Smith, puters in chemistry at a minority insti- Stanley G., Johnson, K. Jeffrey, and tution have convinced me that many stu- Moore, .7_ W., "Computer-Assisted In- dents, underprepared ones in particulur, struction .n Chemistry," Journal of can benefit from computer-based instruc- Chemical Education, 56, 219 (1979). tion. I believe that this is possible 10. Cauchon, Pauf,-Wimistry with a Com- without elaborLte computer hardware, ex- puter, Programs for Learning, Inc., tensive programming experience, or a huge P.O. Box 954, New Milford, CT (1976). investment of time and m^ney. Alfred 11. Moore, John, Gerhold, George, Brene Bork has stated that there is no one man, G.L., Owen, G. Scott, Butler, right way to use- computers xneducation, William, Smith, Stanley G., and Lyn - or even a most profitable way (8). Cer- drup, Mark L., "Computer-Aided Instruc- tainly there is no one best way to use tion with Microcomputers," Journal of computers in science instruction or with Chemical Education, 56, 776 (1979). underprepared students or at a minority institution. But there are many meaningful and wortl-while ways in which computers can be used to enhance learning for all students (9,10,11). Instructors at minority institutions should not consider computer-based instruction to be beyond their c. Abilities. The computer is an important instructional tool that needs to be included in planning present and future educational delivery systems.

REFERENCES

1. McDermott, Lillian C., Iiternick, Leonie K., and Roseng.tist, Mark L., "Helping Minority Students Succeed in Science," Journal of College Science Teachin , 9, 135 (1980). 2. Ketn k, Louis J., "Teething Science to the Disadvantaged Student in an Urban Commbnity College," Journal of Chemical Education, 50, 46,71-973177. 3. Beck, James D., "Using i.L4 Computer in the Teaching of Science," Pro- ceedings of the Minority InstTEUrions Curriculum Exchange Conference, Wash- ington, D.C., p. 37 (1979). 4. Several programs are available from Programs for Learning, Inc., P. 0. Box 954, New Milford, CT 06776. 5. Bork, Alfred, "Computer Graphics in Learning," Journal of College Science Teaching, 9, 141 (1980). 6. Soitzberg, Leonard J., "Computer Graphics for Chemical Education," Journal of Chemical Education, 56, 644 (1979). 7. Shakhashiri, Bosom Z., "CHEM TIPS - Individualized Instruction in Under graduate Chemistry Courses," Journal of Chemical Education, S2, 581------11371.477- 8. Bo; *4-ee "Computers and the FuAA .ucation," Computer -Based S. In.w Istruction, Andre Jones and Minority institutionoECM1 249

EDUCATIONAL USE OP COMPUTERS IN PUERTO RICO

Prank D. Anger Department of Mathematics University of Puerto Rico Rio Piedras, Puerto Rico 00931 (OM 764-0000

ABSTRACT Puerto Rico presents a unique blend of problems and promise in educational computing. Both the lack of an adequate technological base in the community and restricted capital hold back the development and implementation of computerized instruction in all forms.Public schools, facing many of the same problems of main- land inner-city schools, are totally unable even to start in this area. Nonetheless, since about 1974 there has been a rapid growth of instruction with and about computers in the universities, colleges, and private and parochial secondary schools. This cowth promises to increase radically during the eighties. The educational problems to which this development responds, the kinds of computer strategies attempted, some interesting results, and projections for the eighties will be discussed.

262 Computer Laboratories in Education

MICROCOMPUTERS IN THE TEACHING LAB*

Dr. Robert F. Tinker, Director, Technical Education Research Centers, 0 Eliot St., Cambridge, MA 02139 :617) 547-3090

AN OVERLIOXED AREA While computers have many uses in sci- meets are often too stringent for simple ence education, their use in instrument- time-sharing systems. It is possible to ation has been largely overlooked.This use tine-sharing for laboratory measure- paper will show the many powerful things ments, but it is considerably more complex that can be done with a properly inter- and expensive than using terminals for faced computer to gather, analyze, and time-sharing. present real laboratory data. The arrival of microcomputers on the The major reason that this educational scene has completely changed the hardware tool has not been developed in the same situation. Many laboratory applications way as other educational applications of of computers require very little computa- computers relates to hardware.Twenty tion and place only minimal requirements years of experiments with educational op the computer. As a result, very simple applications of computers have almost and inexpensive microcomputers can be entirely involved large time-shared main- successfully used in the laboratory. Now frame computers, which are not well that the hardware question is being solved, adapted to the timing requirements im- it is time to begin a vigorous effort to posed by real laboratory measurements. make up for lost time, to begin using com- A time-shared computer determines when puters in the lab, to develop related data is going to be read from the term- hardware and software, and to research the inal, and so communications with it are educational implications of this tool. determined by the computer's timing requirements. While that condition may SAMPLE APPLICATIONS usually be adequate for human inter- Our group at TERC began experimenting actions with a terminal, it is not 'ade- three years ago with applicatioac of quats for most laboratory interactions. microcomputers in the laboratory because In lab applications, the timing require- we were writing course material on instrumentation as part of our modular electron- * Some of the materials incorporated.in ics project, which is designed to prepare this work were developed with the fin- .tmdents to use instrumentation. It was ancial support of the National Science clear from the beginning that no such Foundation Grant Nos.7711116 and SED course would be complete without a fairly thdrough coverage of microcomputer appli- 79-06101.

250 Computer Laboratories in Education 251

cations in instrumentatim.research labs which can be observed as it evolves.The that are not already heavily dependent on temperature detector is an inexpensive computers for gathering and analyzing signal diode. Because the transducer is data will certainly become so by the time small, a small sample can be used, and the current generation of students are on the experiment can be speeded up so that the job for any length of time. it only takes a few minutes. As a result, it is feasible fat students to obtain data The Coolina Curve Experiment on mmy different samples and focus their Historically, the first application we attention on the phenomena, rather than developed used the KIM computer to record on the boring details of gathering and temperature of a liquid sample of naphtha- displaying the results. lene as it cooled through the crystalize- tiop temperature. This is the cooling A Pot ourri of Applications curve experiment familiar to those who SStce this initial venture into micro- have taught introductory physical science. computer applications in the laboratory, There is some nice physics in it because, we ham developed over a dozen other lab as the sample loses heat to the surround- applications for simple microcomputers. ing water or air, the temperature does not These ire listed below with some of the drop uniformly. At the transition temper- more general applications described first, ature, heat flows out but the temperature followed by some programs tailored to the does not drop, giving rise to a plateau- specific needs of certain laboratory like graph of temperature as a function of exper4ments: time. This shows very clearly somethiug Counter/Timer. In this program, the that many beginning students do'not appre- computer displays counts or elapsed ciate! namely, that there is a distinc- time. It has all the functions of a tion between heat and temperature; that normal counter/time but with a lower at the solidification temperature, heat frequency response. In addition, it is flowing out without a temperature can record for later recovery the num- change. ber of times that multiple events hap- There is some good science in this pened after an initial trigger period. experiment. First of all, of course, the IC Testin .,This convenient program plateau temperature is a good way of de- earns the logic expected !too an in- termining the melting point of a sample. tegrated circuit (IC) by running In addition, mixtures of napthalene and through the permutations of a func- paradichlorobenzene (moth flakes and moth- tioning IC and than comparing this to balls) display some interesting melting questionable ones. phenomena. In some ratios, the plateau Function Generator. Using an analog disappears, as it will for most mixtures, output, this program generates a var- but at one particular ratio -- called the iety of functions that one would only eutectic -- a plateau reappears at a temp- expect from a very sophisticated func- erature that is below the melting tempera- tion generator. Ten different func- ture of either of the pure substances. tions, incliding white, pink, and blue All these phenomena are accessible to noise, are available with selectable someone with as little instrumentation as amplitude, offset, and frequency up to a mercury thermometer and a stopwatch. a bandwidth of 20 kHz. However, gathering and plotting the data Transient Recorder.This program can be a tedious and boring task, parti- triggers when the transient starts and cularly when it is necessary to repeat displays the result on a standard the experiment several times for different oscilloscope, thus effectively con- mixtures. A typical cooling curve run verting it to a storage-type oscillo- takes 20 to 30 minutes. This run can be scope. Up to 256 samples of the input speeded up by using a Koller sample, but signal can be recorded at intervals then many fewer points are obtained and as fast as 20 microseconds. In effect, many of the important features of the the cooling curve experiment is a cooling curve are lost. special adaptation of this facility. The microcomputer solves this problem Fourier Synthesis. This program can entirely. We developed # very simple require: the amplitude and phase of interface that can be ad to record the up to 30 terms in a Fourier synthesis. temperature of a sample at any rate The resulting output waveform is dis- desired and to display on a simple oscill- played On a common oscilloscope and oscope a graph of the resulting tempera- can be beard with the help of a simple ture versus time. The microcomputer can power amplifier. log in at a rate of one per second to gen- Fourier Analysis. This program uses erate an apparently continuous graph, the idea behind the transient recorder

26,1 252 NECC 1960

to capture an input waveform, which is mysterious, thus weakening the connection then frequency analyzed. Using a Fast between reality and result; and third, it Fourier Transform algorith, the com- isn't really necessary at all. These ar- puter displays the power content of the guments are taken up in order below. first 128 frequencies in the captured signal. Complexity Radioactive Half-Life Experiment. Microcomputers introduce a whole new Here, pu &Mt from a Geiger Tube are level of complexity that requires master- counted over time, and the resulting ing sophisticated hardware, software' half-life decay function is displayed and operating concepts. In order to use on an oscilloscope. computers effectively, the students will Pulse Height Analysis. This program have to investigate all these new ideas uses a special interface to capture and will get sidetracked, so that any the height of pulses from a standard possible advantages that the computer can photo-multiplier particle detector. bring will be offset by the enormous in- The resulting pulse height spectrum is vestment of time and intellectual effort then displayed on an oscilloscope. With required. less than a dozen integrated circuits This would certainly be a compelling in the interface, a standard microcom- argument if it were true, if students were puter can be used to replace special- required essentially to develop all the purpose PHA instruments costing ten hardware and software needed in the labor- times as mach. atory.. However, this is an inappropriate The Computer of Average Transients. We way to use a computer in the laboratory. have developed a geological sounding The desirable approach is to use previous- experiment that requries careful pro- ly developed canned.programs that are cessing of signals from a geophone that carefully designed to require the minimum detects seismic waves generated by hit- amount of additional knowledge.Lab pro- ting the ground. The common technique grams should operate just the way elec- is to use a compliter of average trans- tronic pinball machines do; anybody with a ients, which the KIM simulates, using a quarter can walk off the street and begin simple program. using those microprocesmor-based machines. Solar Collector Analysis.Temp4rature There is no reason that laboratory instru- from a number of sensors distributed ments using microprocessors should be any around a solar collector, as well as a more difficult to use. Admittedly, some light detector measuring the input of our first efforts in this area do not light levels, is simultaneously logged meet this requirement. But there is no and selectively displayed through use basic reason that prevents microcomputers of this program. from being easy to use in an application Rotational Dynamics. This program, area. It is simply a question of good written by Ken Flowers, measures the programming; providing menus for selec- angular acceleration of a disk commonly tion and help files to clarify any diffi- called for in lab experiments involving cult points. rotational dynamics. Linear Dynamics. An electronic version of spark tape uses clear tape with Mystification black lines. When attached to an object The skeptic would hold that an instru- and pulled through a special detector, ment that instantly gives result is un- the computer can display position, desirable when it is not clear to students speed, and acceleration instantly. how those results are obtained. If there is no alternative way to cross-check the EDUCATIONAL IMPLICATIONS -,,cults, if the experiment must stop when These examples illustrate that it is the apparatus stops.fuectioning, and if technically and economically possible to the computations are too difficult to dis- have students use the computer asa gen- cuss, then the results may as well be eral-purpose laboratory instrument: the magic. big question in many people's minds is Since the results appear to be magi- whether such use is desirable. I will cal, the student loses any chance of hav- begin addressing this point by first ing any intuitive understanding of the developing the skeptics' argument more meaning of the results and connection be- fully. tween those results and the physical phen- There are essentially three main For instance, arguments against the use of this kind of omena under investigation. sophisticated instrumentation. First i* you pull a strip of marked paper through a that it is too complex and difficult to little gadget, and the computer prints out The connection use: secondly, it makes the laboratory some acceleration data. Computer Laboratodes in Education 253

between the way you pull the tape and the of inputs and outputs I call "intuition acceleration is incomprehensible and un- calibration" because it is directly anal- verifiable, and therefore, the students' ogous to the calibration one performs on understanding of the term acceleration is instruments. not enhanced by this experiment. With a The thesis, then,is that if a micro- spark tape, the student can see the marks computer is used to measure acceleration, on the tape, understand the relationship this will appear to be mysterious only between those and the motion, and trace until the students' intuition is calibra- the calculation of the acceleration ted. Now is this done? The direct ap- through a series of simple calculations roach is to give students this tool and based on the marks. Furthermore, the cal- let them play around with it in a directed culations are based directly on the defini- manner in order to get a good intuitive tions of velocity and acceleration, and feel for what it is producing. In tests therefore, the student gets practice in with students, we have found that it applying those definitions by performing takes very little time using this tool to the calculations in the laboratory. all appreciate, on an intuitive level, the awareness of these relationships are lost relationship between the acceleration and when the computer automatically calculates the change of speed, namely, that rapid the accelerations and graphs them. changes in speed results in large accel- ,An important part of laboratory exper- eration and that slowing down is just iences for students is their enabling another form of acceleration with a neg- function. That isby learning how to ative sign. These intuitive explorations frame and answer questions, students are demystify the tool, but more important, then better able to function as scientists they foster an intuitive feeling that no and as citizens. Understanding is the key other approach can create. Students come to this enabling function, and the inclu- away with more than understanding of an sion of a major thing, the microcomputer, isolated term, such as acceleration. They which is inherently incomprehensible, is talk about acceleration in intuitive terms disabling rather than enabling. that are directly related to its defini- These are important arguments and point tion. They see that it is position-indep- up some of the pitfalls of using the com- endent, that it only depends on how quick- puter in the laboratory.Like any new ly speed is changinl. Thus, the measure- tool, the microcomputer can be misused. ment and the definition come together in However, the statement of the problem the laboratory without the use of compu- points to the solution. Most people can tation. learn to use mysterious or incomprehen- The one argument that can't be count- sible devices -- so long as they have a ered is that students do not gat as much clear idea of cause and effect.Many practice performing computations in a aspects of diving a car are essentially laboratory where the computer does the mysterious to drivers: the details of the computation. The obvious cure for this connection between the wheel and the dir- is to assign some additional computation ection of motion, between turning the key or to do the experiment once without the and the operation of the motor, between computer. the position of the accelerator and speed. The same arguments can be applied to most 1 tt of the technologies in our environments ---TEirikepticts argument hers is that the television, radio, calculators, tele- computer, while attractive, is not really phones, and aspirin.People use them all necessary in the laboratory. However, more or less effectively without any de- the microcomputer makes a number of things tailed understanding of their mechanisms. possible that could not otherwise be done Through practice people have learned the convenientli or economically, and there- connection between inputs and outputs. fore, unless cne takes the fan.estic pos- They have learned what to expect when ition that there is no Seed to improve the accelerator is pressed, when the science instruction, one cannot ignore record button on a tape recorder is these rather substantial improvements. punched, whena telephone is dialed. The The use of a general-purpose laboratory reproducible and predictable nature of computer broadens the kinds of experiments these technological aids, which can be that can be contemplated in a teaching understood at an intuitive level through laboratory and also increases the rate at practice, completely removes the mystery which data can be gathered and analyzed. surrounding their operation and obviate* For example, the dynamics measurements the need for a detailed understanding of permit a detailed analysis of the Connec- how they work. This process of gaining tic..1 between force and acceleration that an intuitive appreciation of a connection in simply tot possible through any other

2 c

.1MM. 254 NECC 1980

mechanism. This application alone opens things that can be done, and we also eval- up new andas yet, uncharted possibili- uate and distribute hardware and software ties in the introductory laboratory. Sim- so that schools and teachers can find most ilarly, the Fourier Analyzer, the Pulse what they need at one place. We have de- Height Analyzer, and even the Computer/ veloped three laboratory interfaces and Timer open up possibilities that would are currently working to Interest some otherwise not be considered in elementary manufacturers in these or other related laboratories because of cost considera- products. tions. For a number of years, we have been The speed with which experiments can giving workshops for teachers on the use of be analyzed is often an important educa- microcomputers in teaching. These work- tional factor. Instead of a simple fall- shops have been very well received and, ing ball experiment occupying an entire in fact, we face the problem that there laboratory, many accelerated motions can are more people interested in taking these be studied together in one laboratory; workshops that we can possibly enroll. To instead of one cooling curve, a set of meet this need, we are in the process of cooling curves of various mixtures can be developing an exportable workshop -- a one- studied in the same length of time. The day workshop that former workshop partici- result has to be that students will have pants can themselves offer to others. We more raw data available to them on which are currently looking for groups of to fashion their theoretical model of the teachers who will participate in the first natural world. Because less time is de- round of these workshops and be willing, voted to the calculation and presentation then, to host these workshops themselves of data, more time can be devoted to un- in their immediate locale. derstanding the scientifice phenomena. In many cases, the hand calculations are not Long-range Efforts only time-consuming but distracting; We see a time in the near future when because they occupy most of the lab time, microcomputer-based instrumentation will students tend to focus on them and totally be as widely used in introductory science forget the scientific phenomena being in- teaching as microscopes, clocks, and ther- vestigated. mometers. The microcomputer will not be a A secondary but important argument for frill added on top of other instrumentation; the inclusion of microcomputers in the it will be the primary instrument. Because teaching laboratory is that by the time of the ease of programming and flexibility our students are employed, those that go of input and transducers, grade school into the laboratory will find microcom- students could use microcomputers in much puters there. the way they currently use microscopes, to extend their senses, to help them perceive THE FUTURE physical phenomena that are outside the I feel that the microcomputer viewed as range of their immediate senses, and to a laboratory instrument is an emerging provide the questions and motivations for resource which is cost effective today in scientific studies. Most of the elements certain teaching situations and will be necessary to fulfill this vision are cur- used in the immediate future in an rently available, and it is certain that extremely broad range of science teaching prices will drop over time to the point at environments. which few schools could justify not having microcomputers in the laboratory. Current Efforts Meanwhile, however, we see four areas The pestary impediment to the wide in which important work needs to be done scale use of microcomputers in the labora- before microcomputers will be widely uged: tory now is that software and hardware are education research and development, improve- not readily available from any one source. ments in the laboratory interface hardware, This, in turn, is partly because commer- network software for sharing limited cial vendors do not perceive the teaching resources,.and development of curricula laboratory as a commercially viable area that use laboratory-interfaced computers. because there hus been no broad expression EdttionResearchdDeveloent. We of interest by laboratory teachers. suggest t at t of As a non-profit organization dedicated microcomputers in the laboratories, while to improving science instruction, our having many advantages, also creates cer- group at TUC have undertaken a number of tain pitfalls. Both the advantages and the projects to speed the introduction of the pitfalls need careful researching. In what microcomputer into the laboratories.We sense can students learn to master such a have, of course, generated some sample powerful tool? What intellectual pre- programs that illustrate the kinds of requisites must they have?The questions

2e'V Computer Laboratories in Education 255

raised about mystification and enabling of puters to be used for extensive data students must be studied in greater detail. analysis, combining data from many students There is a possibility that the introduc- and generating complex displays. This tion of laboratory experiences that use the local network would be tied into a central microcomputer will be of benefit not only computing center, which could be used for to science instruction but also to mathe- long -term, storage, high-speed printing, matics learning. When combined with the and the maintenance of large data bases. power of the computer to analyze and model, We are cooperating with a number of the introduction into mathematics ,:tstruc- schools to work towards an initial imple- tion of computers that can deal with real mentation of this ideal. phenomena will have an important role in Curriculum. The most pressing current motivating students in building intuition; neorii-Wargeneral area is to develop this area needs to be researched and de- curriculum material that uses laboratory veloped. microcomputers.One can view the programs Interface Hardware. We see a continuing that we have developed so far as samples, need to keep abreast of hardware develop- a shotgun blast that illustrates what can ments and continually incorporate relevant be done.There is an urgent need syste- innovations into interface designs. It matically to apply laboratory-based micro- is only by producing and distributing in- computers to courses in various disci- creasingly sophisticated prototypes that plines. An important aspect of this we will learn how these can be used in edu- effort is to try to define the electrical cation, how costs can be minimized, and characteristics of the interface that is how dissemination problems can be reduced. required, so that the curriculum material One of the benefits of our work in build- does not haye to be tied to any particular ing and distributing prototypes is that it hardware.We naturally feel that our lab helps build a market we hope commercial interface defines this standard, but we companies will want to exploit. would be del.ghted to discuss any other We see a clear need to develop an in- possible standards with any educators. terface that contains its own 16-bit CPU The only way extensive curricula will and communicates to a host computer over be developed is to use teacher-generated the new general interface bus or over material.The major problek with this other standard communication lines. Such approach is the evaluation and distri- a device would have greatly increased bution of this material.We have estab- performance and would be compatible with lished an experimental Microcomputer almost all commercially available compu- Teacher Resource Center that uses a com- ters. While it may not be affordable by mercial data base to store reviewed files a large number of institutions this tine, and programs. This data base is public it is important to learn, through limited and contributions are welcomed. If distribution, what educational' advantages teachers will evaluate what they use accrue from the increased performance this from this data base and contribute some unit would have. new material, great strides will be made Networking. The expensive part of in filling the missing curriculum gaps. computers now is not the central processing unit or the memory but rather the SUMMARY peripherals: printers, disks, specialized A microcomputer equipped with a labora- interfaces, graphic displays, and the tory interface offers a very exciting new like. The appropriate way to maximize resource for science instruction.With computational performance is to have a the proper hardware, the computer can be means of sharing the expensive resources turned into an altogether general-purpose among a variety of inexpensive computers. instrument. It can replace a large number Thls is called networking and is the of standard instruments at a fraction of microcomputer alternative to time-sharing the cost, and provide the major measure- which is really only appropriate for ment and analytic tools needed in a wide computers that have expensive CPUs and range of science courses. Equipment and memory'. We have some interesting net- software are available today that begin working software and hardware currently to do this, while future hardware develop- operating in our lab that need to be ments and declining costs will make lab- brought into the teaching environment based microcomputers a necessity in nary and expanded to laboratory applications. teaching situations. We envision that the teaching laboratory of the future will involve a network of inexpensive computers at each laboratory station, networked with a number of more sophisticated special-purpose microcom-

20w 0 256 .4E0C 1980

THE COMPUTER LAB OF THE 80S

Guy Larry Brown Head, Data Processing Piedmont Virginia Community College Et. 6, Box 1-A Charlottesville, Virginia 22901 (804) 977-3900

INTRODUCTION er for hardcopy output occupy only a small During the past 20 or so years that the part of a single disk. The remainder is computer has been used as an academic roof, set aside to store student programsand there has been considerable discussion for other usesas may be desired. For about computer-assisted instruction (CAI). example, a program can be,colled by a stu- CAI, in fact, has been a luxury affordable dent to provide information about a problem by only a few. In recent years, with pres- he does not fully understand. Or a series sure being applied to educational budgets, of tests can be stored for use by the stu- it appeared that CAI would remain a dream dent at the instructor's discretion. to be realized only if the good foivy god- Two of the micros receivedminor modi- 'other waved her magic, wand in a class- fications to permit I/O througb a standard room. However, the advent of the micro- cassette player. This capability permits computer witb its by cost and super cap- the units to be detached from the system abilities has made it possiblefor that and moved out of the lab for demonstrations marvelous wand to bedirected into even or other purposes.Zr also allows for the the most ispoverisbed academic niches. transfer of a program stored on a cassette This paperdescribes an operating system into the micro's memory and henceonto a of microcomputers that is unique in capa- disk for storage and subsequent availabil- bilities and costs. ity to all micros. All computers were manufactured by Ohio THE SYSTEM Scientific, the printer is aCentronics

The system in operationtonsists of 14 . 779, the video monitors are Sanyo VM 4209, microcomputerseacbwith 8,192 (8K) of and the CRT is a TEC Series 500. The sys- randomaccess memory (RAM) and 8K of read tem is capable of operating with 16 of the only memory (ROM). The RAM allows storage micros; however, to keep total costs below of instructions written in Microsoft BASIC $19,000, only 14 were purchased initially. and data in the form of letters of the al- This figure included all equipment,instal- phabet, digits, and special characters. lation, and softweref Equipment selection The ROM contain' an interpreter that tran- was influenced by the availability of a slates BASIC into the computer's language. reputable local dealer whoservices what Each micro has its own keyboard for input he sells. Also favoring Ohio Scientific into the computer and a video monitor for is the large inventory of software,' small visual output. business system capability with hard disks, Stored programsareavailable from and design featuresmaking expansion and shared dual floppy disks, each capable of upgrading simple snd relatively inexpen- holdingup to 275K characters. The disk sive. unit is part of the hostcomputer tbreugh which controlof the micros is centrally OPERATION exercised. This host computer has 48K RAM The system ins operated almost flawless- with its BASIC on disk instead of inROM. ly since installation in the fall of 1979. Control of the bost.is through a CRT. At- Typical micro systems use cassette I/O tached to the host is a 100 character per which is very slowcompered to disk. And second (CPS) matrix printer which is also a dedicated disk for stand -alone micros Shared by the micros. adds significantly to the costa the unit. The program (software) that provides The shared systemallows each micro user the capability for the micros to share the to (1) load a program stored on disk into disks for input/output(1/0)and theprint- the memory of hiscomputer, (2) save a 2 ComputerLaboratories in Education 257

program on disk, (3) print a listing of DISADVANTAGES the program, or (4) print the output from Disk and printer allow only one user at the execution of a program. Only one micro a time (this is only a minorincon- can use the disk or printerat a time. venience). The use of passwords for files, message Cost is so low that academic departments sending to the host computer, and line can afford procurement without in- advancing on the printer are recent, volvementof computer management locally added enhancements to theoper- (maybe this is an advantage). ating software. Documentation is not complete or error free. USES Operating software lacked capability Presently thesystem is used only for for printing the execution output teachingthe programming language BASIC. of a program (locally modified to One 20-student class was taught in the acquire this capability). fall of 1979, and two such classes are Service may not be locally or quickly being conducted during the winter quarter. available. During registration, several units were Cost is so low that one cannotexpect moved from the labfor demonstration pur- the same level of manufacturer sup- poses. Curiosity was aroused in many who port as previously experienced when stopped to engage in conversation with the systems were priced much higher. computer or play a game. Demand for usage night be so higb that the establishment of priorities for FUTURE PLANS use may be difficult. In addition to expansion to 16 micros, there are plans for the acquisitionof a SUMMARY . quality character printer for word proces- A computer lab for the &Os is available sing. Thiscapability will permit in- today. It consists of up to 16 microcom- structionof secretarial science students puters, each of which can baveup to 32K and members of the staff in word proces- RAM. They have rapid access to programs sing. and data through a shared dual floppy disk Expansion of courses to cover assembly drive which also provides for equallyznpid language is anticipated as well as a course storage. Hardcopy listings of programs and in the operationof small business computer output from the execution of thoserrograms systems. Longer rangeplans call for are also availableon a 'bared printer. teaching COBOL. The system, including the host computer and Effortsare being made to make the lab all other hardware, installation, and nec- available to any of our faculty for use in essary software isavailable foraround their courses. Programs will be developed $20,000. for math, chemistry. biology. and a variety of business courses.

ADVANTAGES Cost. Reliability. Graphics capabilitiesof 256 characters. Standardization of components. A single disk for program/data storage. Displayof 64 characters/32 lines on micro videos. Extensive software available from manufacturer. Expansion of capabilities. and updating inexpensive. Notdependent upon trained computer personnel. Rapid transfer of programs/data through a disk system. Hardcopy available to all micros. Control is facilitated thxdOsh the host computer. No reliance on telephone connections. Students learnon a micro frequently found in businesses. Micros still function independently.

270 25$ NECC 1900

THE EDUCATIONAL TECHNOLOGY CENTER

Alfred Bork, Stephen Franklin, and Barry Kurtz Educational Technology Center University of California Irvine, California 92717 (714) 833-6911

This paper reports on the formation the past, so that the Center has a of the Educational Technology Center at nationwide effect beyond its immediate the University of California, Irvine. The activities, materials, and publicity. primary focus of the Center is the use of The Center will publish a newsletter the computer as a learning aid. The reviewing the activities and results of Educational Technology Center was started on its projects. Although no set schedule January 1, 1980, with University funds is planned, we expect this newsletter to providing staff support. The Center con- be published three times a year. Anyone tinues the activities An computer-based interested in receiving the newsletter learning conducted by the Physics Computer should write to the Center. Development Project during the last eleven The following list gives the active years. projects at the Irvine Center. Further information about any activity is avail- NEED able on request. The Educational Technology Center was formed because we believe strongly that 1. A. Testing and Tutoring Environment the next decade will be a critical period for Large Science Courses. in American education.Such centers are Authoring for personal computers needed to guide us toward a future where Testing environments tte computer will play an extremely im- Physics - waves portant role An education. It is impor- Statistics tant to develop a number of continuing National Science Foundation--Compre- groups that are not fully dependent on hensive Assistance to Undergraduate grant funds but have an existence beyond Science Education (CAUSE) support for particular projects. 2. Scientific Literacy in the Public We have pursued for some years within Library. the University of California the possi- Public libraries, shopping centers, bility of one such Center. Estill pro- science museums vide guidance to others working in this Public understanding of science area. The Center will work on a wide Personal computers range of research and development activi- Fund for the Improvement of Post- ties leading to more effective use of the secondary Education (TIPS!) computer and associated technologies in 3. Mathematics Competency Tests for learning environments. Beginning Science Courses. University of California/California CURBINT ACTIVITIES State University and Colleges The Educational Technology Center 4. Translation of timesharing materials intends to engage in many activities con- to personal computers. cerning more effective and more efficient Universit of California/California use of information technology in learning, State University and Colleges emaphasising learning materials on the 5. Biology materials ecology. Wien.' computer. Some of the activities University of California, Irvine, will be pure research, while others will Committee on Instructtonal Develop- have& strong applied and developmental ment component. We shall work closely with individuals and groups elsewhere, as in

27j Computer Laboratories in Education 259

6. Development of Reasoning Skills in activities increases. We should aim for Early Adolescence. the best possible materials at the least Junior high students developmental cost. Transition to formal reasoning 5. Expanding technologies. Com- Personal computers puter and associated technologies are National Science Foundation - Devel- evolving rapidly. We must learn quickly opments in Science Education (DISE) to use an.expanding range of capability, developing materials which are not immed- PRODUCTION SYSTEM iately outmoded. In addition to specific products, 6. Thec,terinewinterac- such as those just mentioned, the Center tive mediumW-ew has developed a production system for earrrnigmedium, we must learn how it generating computer-based learning mater- differs from older media. For example, ial. The emphasis is on both efficiency reading from computer displays has many and effectiveness and on techniques which differences from reading print medium, but will allow natural extensions to large- the empirical details are not known. scale production of such models. The 7. Dissemination.New media also production system is based on a systems demand new males of dissemination. analysis of the problem and on our many S. 'sew course and institutional years of experience in producing a wide structurss. As computers are more widely range of learning material. Literature is used, they will have major effects on available describing the system and the course and institutional structures. supporting software. The Educational Technology Center intends to pursue these and other issues. ISSUES FOR THE FUTURE Currently we can distinguish a number of very important issues that will shape the future of computer-based learning: these issues indicate directions the Educational Technology Center will pursue. No order of priority is intended in this list. 1. Full-scale course development. At present, with a few notable exceptions, computer-based learning materials are supplementary to course structures. Very few full courses mike heavy use of computers to aid learninc. We need experience in developing such complete courses and in integrating computer and other learning aids. We need additional experience in computer-aided delivery of such courses. 2. Expanded acquaintance.Very few teachers, and even fewer members of the general public, have seen any effective computer-based learning material. Often the examples seen have been weak examples; so the learners have formed inaccurate opinions of the value of such material. We need more acquaintance with the full range of possibilities, more computer literacy with a learning emphasis. 3. Research in learning. Presently we have conflicting theories about learning. We need to know more about how students learn so that we can develop better learning aids. 4. Production techniques. Older strategies for developing materials often were not suited for the large -scale development needed in the years ahead. The types of systems approach followed at Irvine and elsewhere needs further exploration and refinement as the scale of 2 Invited Session

MIS EDUCATION: INDUSTRY NEEDS AND EDUCATIONAL SOLUTIONS Chaired By Eleanor W. Jordan Department of General Business Business-Economics Bldg. 600 university of Texas at Austin Austin, Texas 78712 (512) 471-3322

ABSTRACT PARTICIPANTS (Listed in order of Presen 7574FEhe past decade a considerable tatzon number of discussions in industry-oriented publications like Computerworld have Marguerite Summers, Chairperson focused on irrelevant education as a reason Western Illinois University for prevalent software problems and DP personnel shortages. In this session, David Naumann educators will discuss efforts made by University of Minnesota their institutions to resolve the supposed MBA and PhD PIS programs relevance problem in programs for business application's software designers at the Joyce J. Elam graduate and undergraduate level. Two Wharton School industry representatives will also partic- University of Pennsylvania ipate in the panel. MBA concentration in MIS Eleanor W. Jordan University of Texas at Austin Undergraduate DP program

James Cook Southwest Texas University Undergraduate business CS program

Ken Truitt ARCO Oil and Gas Company

Willis Ware Rand Corporation

27ki

200 Computer Games in Instruction

SHALL WE TEACH STRUC- TURED PROGRAMMING TO CHILDREN? Jacques E. LaPrancc Dcpt. of Mathematical Science Oral Roberts Univ. Tulsa, OK 74171

ABSTRACT able to code some substantial programs. --liZZT011ege programming experience has The task ofcomputer science education been observed to be detrimental in many will be made more difficult by this back- cases to college-level study of computer ground experience because the students science. The problem is the lack of un- will have to do more unlearning than is derstanding of the principles of struc- currently the case and sometimes unlearn- tured programming. The solution proposed ing is more difficult than learning. here is theintroduction of structured programming games at the elementary THE PROPOSAL school level which will prepare the A possible solution to the above prob- children to do structured programming lem would be to introduce structured pro- later on with languages such as BASIC and gramming concepts to children before they FORTRAN that are not designed to promote are able to begin to use conventional structuredprogramming. An example of programming language. Other oulutions doing this called ANTFARM-s presented might be to replace BASIC wich PASCAL as and the results of using it with one the most common microcomputer language or group of children are discussed. to have all manufacturers' BASIC manuals based on structured programming. Neither THE PROBLEM of these alternative solutions seems RUFF-7E7 morestudentsare entering feasible, The simplicity of BASIC will university computer science programs with cause itto continue to be poptIlr, es- some form of prior.experience with com- pecially with the smallest computer con- puters. This is typically a high school figurations. Thelarge number of pro- course in FORTRAN or BASIC or experience grams already coded in some dialect of with a personal computer. There seems to BASIC will also serve to keep support for be nocorresponding exposure to struc- BASIC strong. The manufacturers may be tured programming or design, however. more or less willing to include struc- The students believe they know a lot be- tured programming concepts in their man- cause they have written programs in BASIC uals, but the bottom line will always be or FORTRAN, but in reality they know onk what enables them to sell their product. coding; they understand nothing of struc- The general public will have to become ture, top down design, good style, or more knowledgeable, discriminating,and documentation. It is sometimes difficult particular before the manufacturers will toget them to break the resulting bad feel any serious pressure to include ma- programming habits. This problem is like- terial on structured programming. lytobecompounded by the increasing Although the first solution mentioned availability of small personal computers above is not a shoo-in, itis one that without proper accompanying instruction can be developed more easily than either in good top down programming. An in- of the others and would be worth inves- creasing numberof youngpeoplewill tigating. It is not realistic to have likely learn programming by reading their all elementaryschools begin teaching microcomputer BASIC manuals. The learn- structured programming, but a few well- ing of coding will thusbe encouraged published success stories plus the avail- rather than the learning of programming. ability of well-written materials for use Yet these students will be misled into in schools wouldhelp to promotethe thinking they know a lot because they are idea. A fewsuccessful projects at

261 274 282 NECC 1980 t,

schools in several areas across the coun- Itacceptsfive basic commands: MOVE,

try would encourage other educators . io TURN LEFT, TURN RIGHT, EAT, and PLANT. somethingsimilar and inspire comp...er MOVE means move forward onecharacter manufacturersto include structured de- position in the direction the ant is fac- sign and programming in their manuals. ing; TURN, LEFT or RIGHT, means a 45° ro- Structmed programming 'conceptscould tation about the first body segment; EAT easily be introduced to elementary school means to consume whatever isunder his thildren if they were presented in e gene head; and PLANT means to plant a new seed form at theirlevel of sophistication. with his tail. Seeds starts out as "."; Gamos could be developed for which the 100 time units later they germinate con 11 language is inherentlystruc- (","); then after50 more units they tue and the children would be moti- sprout ( " ; "). They grow into a stalk val.ea by the attraction of the game to ("1") after 50 more units and then into a master the control language as well as branching plant ("Y") after 50 more. At possible in order to do more with the the end of 75 more units,a flower: ("P") game. Because the language would be in- appears and then after 75 more or a total herently structured, they would automat- of 400 units, the plant natures into food ically begin to develop structured prob- ("S"). One time unit is the tire neces- lem solving skills and structured expres- stry to execute one operation. Such a sion of their ideas. Manuals could be growing rate seems to be a satisfactory provided that would describe structured choice though unrealistically fast. design and programming in a way that most Each time the screen is updated, the teachers could follow andwouldalso cursor is moved to the home position, and describe how to use the games with the the first line on the screen is erased to children to develop their understanding await the next input. The user may type and skills. any combination of commands on this line. The LOGO system at H.I.T. Artificial When the return key is pressed the com- Intelligence Laboratory (12) already con- mends sze performed. We began by simply tributes to these goals. This system was showing the children the effects of indi- not developed for introducing structured vidual commands. Then we introduced the programming concepts but aould be easily concept of sequence by showing that sev- adapted to that purpose. It is exciting eral com.ands can be listed on the same for the children to use (14) and there- line. fore has the necessary motivational char- Since the sequence of commands cannot acteristics. Several othertoolsfor belonger than one line,the limit on teaching structured programming could be program sizeis quickly reached. Other developed withdifferent appeals, dif- commends are then introduced to allow the ferent levels of sophistication, and dif- ant to do more things. From the author's ferent organizational structures. point ofview, however, additional commands are given to develop the additional THE EXPERIMENT structured programming concepts of itera- (71.:7taing of these ideas was made tion, selection, and refinement. in April 1979 by the author and Dr. Mary The first additionalcommand illus- Dee Fosberg of Central State University, trates iteration: "DO a cc.amand sequence Eemond,Oklahoma. To present the ideas n TIMES," where n is an integer. Instead ofprogramming to a group of gifted of "n TIMES," therecan be "TO ROW children, we designed a gaze called integer"or "TO COLUMNinteger." The ANTFARM. It was implemented on an IMSAI rows ate numbered down the left of the witha Z80 processor, Digital Micro- screen and the columns across the top on systems disk,-Aid Infoton 200 CRT using the second row. This command allows the the UCSD PASCAL system which we took to ant to go through a sequence many times t'.f presentation and set up there. before stopping for the entering of the The ANTFARM program consists of drawing next command. (The "DO" command can be in ant facing up in the center o2 the used in conjunction with the do nothing ..creen and two rows of food ("e's) in commands "WAIT" and "REST" to watch the upper left. The goal is to have the plants grow.) ant move over to the food, eat it, and The command to illustrate selection has plantnew food. The ant looks like: the form: \ *l *1 / PLANT [AHEAD) 0 or -0 or *00-. IF SEE (FOOD LEFT command. 0 / \ ,SPROU1 RIGHT 71\ 1\ LDIRT etc., depending on orientation Computer Game': in Instruction 263

If the character space in front (or to The final commands availab. are "FORGET front left or front right) contains non- name," which eliminates th- named module space for "PLANT," " @" for "FOOD," ";" or from the table, and "nu," which lists "1" for "SPROUT," or space for "DIRT," on the screen all the "REMEMBER"ed names the single command following will be exe- andtheir definitions. 14'...n the use" cuted; otherwise itwill be skipped. pushes the return key afterthe "TELL" This command allows for some interesting display,the screen iscleared and the conditional tasks such as having the ant ant's farm scene is redisplayed as it was look for a row of food to eat or an open prior to the "TELL" display. A future furrow in which to plant more food. Al- addition will be commands to save and re- though the chi'dret were able to use this call the'set of defined modules to and command it seems to need more development from a disk file. Thece commands will to be stronger. In its present form, it allow a child to build his program from seems too restrictive. dayto day rather than having to type To encourage the development of hier- everything overfrom thebeginning at archicalstructure and refinement, the each session. command sequences are limited to one line, and a command called "REMEMBER" is THE RESULTS given. This command has tho form "REMEM- We had about13/4 hours with a group of BER name command .sequence END" which al- gifted children between the ages of nine lowsthe child to give numesto command and twelve. Interest was universally sequences for subsequent use as commands high throughout the session. The format themselves. One popular sequence is of the session was that for about 20 to 30 minutes the children were shown some REMEMBER TURNAROUND DO TURN LEFT 4 TIMES of the hardware aspec..,, of computers; END then the ANTFARM was introduced. Since the Infoton has a detached keyboard, the after which "TURNAROUND" can be used as a author satto the side and did all the comman.l. This subprogram capability is typing while the children looked at the especially helpful when used in connec- screen and said what to type. We intro- tion with theselection command since duced the ANTFARM features roughly in the only one command maybe selectedor sameorderas in this paper, stopping skipped. Tie commands "QUIT" and "STOP" frequently to ask the children how to get allow the currant subprogram to be ended the antto do sou specific task. The prematurely (or the ANWARM program it- children picked up the ideas very quickly silf to end if it is used at the command and were soon telling us of th-ngs they :t.ae level). This capability allows for wanted to see the ant do. By the end of powerful searchfeatures when used in the VI hours, three or four of the most submodules with selection and iteration involved children were beginning to use commands, such as topdown design to achieve some specific task they wanted the ant to do, defining REMEMBER MOVETOFOOD DO MOVE IF SEE FOOD their own modules and giving them names. AHEAD STOP 70 TIMES END The only weaknesses observed were the selection commandsand ,:he bugs. The This will cause the an- to move forward selection commands were too difficult for 71) times or until there is food right in the -hildren to use unaided in their pre- front o' it, whichevercomas first. sent form,primarily because the condi- One of the tasks the children raced VAS tion was something happening out in front the planting of a field of four rows of of the ant instead of right under him, ten plants each. The following is an ex- and the seeingleft or right does not ample of a structured development of a correspond well to Where the ant is solution: facing when he turns. The bugs occasion- REMEMBER FIELD TWOFURROWS TWOFURROWS END REMEMBER TWOFURROWS FURROW NEXTRIGHT FURROW NEXTLEFT END REMEMBER FURROW DO MOVE PLANT 10 TIMES END REMEMBER NEXTRIGHT DO TURN RIGHT 3 TIMES MOVE TURN RIGHT MOVE MOVE END REMEMBER NE.;TLEFT DO TURN LEFT 3 TIMES MOVE TURN LEFT MOVE MOVE END

A field could be planted in the upper right part of the screen by

DO MOVE TO ROW ? TURN FIGHT TURN RIGHT DO KOVE TO COLUMN 65 FIELD

2 76 264 NECC 1980

ally caused the program to abort, losing REFERENCES all the defined modules and all the farm development to that point. Actually the 1. Benet, Bernard, "Computers and Early program worked quite well considering it Learning," Creative Computing, was developed and implemented in three or #4-5 (Sept.-Oct. 197107 pp. four man-days and consists of over 400 90-95. lines of PASCAL code. (This accomplishment in itself is a credit to the value 2. Beyer, Kathleen, and Stuart Milner, of top-down design,structured program- "Elementary School Curriculum ming, and PASCAL.) Task Group," ESSS Re wort, Sep- tember 1979, PP. 2 THE CONCLUSION TheANTFARM program has demonstrated 3. Brady, J. M., and R. B. Emanuel, "An Chat it is possible to create tools with Experiment in Teaching Strateg- high motivational value for children that ic Thinking," Creative Comput- contain ellthe concepts of structured 111I, 14-6 (Nov.=T78), pp. programming. Furthermore, the way this 106-109. program isdesigned. forces the children to use hierarchical or structured devel- 4. Cohen, Harvey A., "Oznaki and Be- opment to achieve their goals. Even in yond," Proceedings of the one session we were'beginning to see some National Educationaraiqutine. of the children using structured develop- daraWice, 1179, pp. 170-178. ment, on their own. It seems reasonable suggestuggest that continued use of tools 5. Dahl, O. J., E. 4. Dijkstra, and C. such as =FARM over an extended period A. R. Hoare, Structured Pro - of time could develop these concepts so ramming, New York: AcaaZiic well in children that they would continue rens, T972. tousetop-down design and structured programming even with languages such as 6. Hakansson, Joyce, and Leslie Roach. BASIC and FORTRAN which are not naturally "A Dozen Apples for the Class- oriented toward them. We hope that we room," Creative Coruting, 5-9 will have the opportunity to continue to (Sept. 1979), pp. 2-54' explore these ideas and that others will have similar opportunities. We need 7. Larsen, Sally 'Greenwood, "Kids and long -term data onchildren using these Computers: The Future Is To- tools to determine the effects on their day," Creative Computing, 05-9 future success in programming or computer (Sept. 1979), pp. 58-60. science. 8. Lieberman, Henry, "The TV Turtle: A LOGO Graphics System for Raster Displays," MIT, A. I. Memo 361 (June 1976).

9. MECC, "Computer Literacy Objectives from MECC," ACM SIGCUE Bdlle- tin, 013-4 ON T. 1979). 10. McGowan, Clement L., and John R. Kelly, Top-Down Structured Ptogramming TechaqUIT7--- New York: Petrocelli/Charter, 1975.

11. Milner, Stuart D., "An Analysis of Computer Education Needs for K-12 Teachers," Proceedings of the National Educational com- utiriaMience, 1979, pp.

12. Papert, Seymour, "Teaching Children Thinking," MIT, Arttti 'al Ir- telligence Memo 247 "%I L:. 1971). Computer Games in Instruction 265

13. Papert, Seymour and Cynthia Solomon, "Twenty Things to Co with a Computer," MIT, a. I. Memo 248 (June 1971).

14. Papery, Seymour and Harold Abelson, Jeanr0 Bamberger, Andrea diSess, Sylvia Weir, "Znterim Report of the LOGO Project in the Brookline Public Schools: An Assessment and Documentation of a Children's Computer Labor- atory," MIT, A. I. Memo 484, (June 1978).

15. Perlman, Radia, "Using Computer Technology to Provide a Creative Learning Environment for Preschool Children," MIT, A. I. Memo 360 (May 1976).

16. Ragsdale, Ronald G., "A Program

. Package for Introducing the Top-Down Approach to Computer Programming," SIGCSE Bulletin, #11-1 (Feb. 1979174. 113-117.

17. Solomon, Cynthia J. and Seymour Papert, "A Case Study of a Young Child Doing Turtle Graphics in LOGO," MIT, Artificial Intelligence Memo 375 (July 1976).

18. Watt, Daniel R., 'A Comparison of the Problem - Solving Styles of Two Students Learning LOGO: A Computer Language for Child- ren," Proceedings of the National EducaticTarini'euting U5EfiFiFice7-11707-pp. 255-260.

19. Weinberg, Gerald H., The Psychology of Computer Promammjag, RFw York: Van Nostrand Reinhold, 1971.

2:7S 266 NECC 1980

STRUCTURED GAMING: PLAY AND WORK IN HIGH SCHOOL COMPUTER SCIENCE

J. M. Moshill, G. W. Amann (The University of Tennessee)

W. E. Baird (West High School) Knoxville, Tennessee

PROLOGUE We propose that a similar choice ofamin Question 1: When, is a computer,gare not a features can foster the development o game? When is it an ok class-activity? logical problem-solving skills, while Question 2: So what's wrong with games, retaining the kinetic/esthetic motivational anyway? structure of video games. (We have all Exasperated answer to 2: known programming hacks who have made the Students won't work on programs when they game/program connection.) We want to use have access to games. Games are fin and midrecomutercolorgmapidestomakeonoputing programs are work. more Mtn color crayons and less like Reflective answer to 2: arithrettc. The usual computer games are either Having said that much, we will answer hand-eye (with occasionally some small , Question 1 and then flesh out our answer amount of brain-) coordination contests, with a description of the curriculum we such as "Lunar Lander," or fantasy-land, are developing. interactive do-it-yourself storybooks such Question 1: When is a computer game an ok as "Dungeons and Dragons.* These activi- class activity? ties take part in the allure of broadcast Answers 1) The game must be designed with television: namely, they involve the a set of concepts and skills in student kinetically and emotionally, but mind and a plan for how the they do not have a cumulative component. game teaches theses You can walk in on television (or computer 2) The things learned in the acti- games like "PONG" or "SPACE WAR") vity must contribute toward a anytime; no prerequisities or logical- cumulative body of knowledge, deductive skills are required. (For an a toolkit that the student can excellent exploration of this theme, see perceive and make use of, as Postman, 1979.) toys and toy-making tools, and We cannot call this kind of attention 3) The game must be su erseded by passive -- observe any kid watching an a more interesting, more nter- action TV show or playing a video game. active game, chosen with extreme Nevertheless the interaction is non-ana- care to be cnplayabln unless the lytical. It has more in common with student has mastered the skills baseball than with reading, more of re-. taught in the previous lesson/ cess than of curriculum.No wonder game. teachers of computing have game trouble CHALLENGE whenever interactive terminals or micros Our mission, in the University of become available. Tennessee/NSF High School Computer Science The problem this paper explores is Curriculum Project (HSCS), is to make com- the development of an introductory cosputer skills available to average students. puting curriculum built around a kind of Computers may indeed become as ubiquitous struc*ured gaming. The computing commu- as telephones and televisions, although we nity has begun to understand that care- believe that the introduction of another fully chosen programming language features technology as soporific, captivating, and can guide our thought in ways that make anti-thinking as television could be a code work across time, that. is, remain major social disaster. M. hope, rather, adaptable, comprehensible, repairable. that computers will become convivial tools like the telephones "convivial" means that lualr7ThruatiallyeupEortedbytSF Grant their use is determined by the user, not SED-79,-914.4 Computer Games in Instruction267

by some central leastcommon denominator declarations become of concern.A PASCAL such as broadcaster. We don't have utopian interpreter is used which scroUs the ideas as to what future generations will source psgram being executed on the bottom do with computers. (Who could have pre- of the screen (at a controllable rate) dicted in 1915 what problems we'd have with while the program produces its output on automobiles?) We do, however, have a the top part of the screen. A working strong feeling that the question of system will be on exhibit at NECC2. whether individuals will be able to program We will conclude with fairly detailed their computers, or merely ha Programs,s explanatiqn of the first two lessons, an open and important questiori. The chal- quilting and TURTLEGRAPHICS. lenge, then, is to give every citizen who TWO EXAMPLE LESSONS can dial the telephone some ability to pro- Lesson ls WM-W.-- gram a computer. Behavioral Objectives: Students should METHOD learn how to insert a disk and start the This section will be brief; we have system. They will run the Quilt program, published elsewhere (Aiken, Hughes, MOshell explore its features and limits, and use 1980) the nuts and bolts description of it to generate :static and moving color HSCS. We are using a cartoon-animation patterns. software system called RASCAL which runs Conce ts: Students will learn the follow- as part of UCSD PASCAL on the APPLE ng: microcomputer. The basic installation 1) how to enter an initial pattern of costs about $3200, including a single information into the system which floppy disk, color television, and 100 then controls the repetitive be- character-per-second printer. Each lesson havior of the computer in a one - semester (18 week) course consists (Introduction Activities). of approximately a week of work, divided 2)116:7; to construct simple experiments into these parts: to determine the parameters of the Introductory activities computer's operation, such as how Exploration project many colors it has, how big the Skill-building project screen is (Exploration Projects) . Buttoning-up activities. 3) how to hand simulate a formal pro- A class consists of about 15 students per cess of discrete steps to predict computer (our collaborating schools have the computer's behavior and under- only one APPLE eachwe hope to try the stand it(Skill-building Projects). curriculum in multi-computer classes later), Resources: Five groups of three students alternate IT-The Quilt program. Students insert computer use with planning work using a disk and on the computer. graph paper and marker pens. The off-line When the greeting message appears, students ere planning their strategies, thuv type X (execute). The doing hand simulations and observing the computer asks EXECUTE WHICH FILE? on-line students, fur to graduate to the The student types QUILT and a re- next activity, a group must successfully turn. QUILT then prints the fol- predict the outcome of an assigned "seed" lowing message: (e.g., geometric pattern, algorithm. pro- THIS PROGRAM LETS YOU DRAW A PIC- gram). The activities develop during 18 TURE.AND THEN MAKE A 'QUILT' OF IT weeks from a non-linguistic, color-pattern BY REPEATING THE PATTERN. TO DRAW: process called "quilting" (next section), SET PADDLE 0 FOR A COLOR; through immediate-mode and straight-line- TYPE KEYS ARCUND K TO MOVE THE DOT. code entry of TURTLEGRAPHICS (Papert,1970) FOR INSTANCE, 'I' MOVES THE DOT commands and the introduction of PASCAL UPWARD. control structures such as REPEAT... TYPE 'R' TO SEE THE PATTERN UNTIL and IF...THEN, to the creation of REPEATED. cartoon characters and their animation TYPE '0' TO HALT THE REPEATING. with complex programs using the RASCAL TYPE 'C' TO CLEAR THE SCREEN. animation system. The output is always TYPE 'Q' TO END THE PROGRAM. color graphics and music; the curriculum NOW HIT THE RETURN KEY TO BEGIN. steadily increases its interaction as When the student hits the return students learn how to use the joystick key, the following reminder message to control motion. There is always an remains on the bottom of the underlying lesson about how programs screens work. Al. code is in a completely REMINDERS: structured language (PASCAL) and is KEYS AROUND,MOVEsIstUP,J=LEFT,U=DIAG, taught from the inside out. Only at later ETC stages do environmental details such as C) LEAR REPEAT MALT DE REPEAT= QMSITTETWOUSPROMM

2S9 268 NECC 1980

which groups have caplets) their tasks and gives them the next challenge sheet:the skill-building The game paddle supplied with the projects.Since the first groups got easy explor- APPLE computer consists of a knob ation projects (and finisher] earlier), they are controlling a computer input; the student given tougher skill-building projects. Groups turns it and observes that the dot in the that haven't figured out their ecplorations by the centerofthe screen changes color.She middle of the third day are helped to fine the types an experimental sequence of answer and roved on. keystrokes such as LLLI and observes that The object of skill-bsilding exercises is to the dot moves and leaves a colored trail be able, to accurately predict what a given seed shaped like loigure 1. She than triestyping R. will do. Two parts are given.In pert I, each The coseuter immediately reproduces the pattern, group is given a different list of three or four starting where the cursor ins left and finis; the seeds and asked to hand-sisolate the result and screen with a periodic pattern. (Since the screen then to try it.when they can do this assignment is waived, when the cursordot fallsoffscreen correctly, they move to part II.Here, the work- left, it returns onthe right, and similarly for sheet gives them the finished pattern, and they are top-betben.)If youRP:levet the. above seed, you to find the seed that creates it.In this exer- get Figure 2. cise, the gang really Incases like a game. Moving 2)Miscellaneous resources:cusbzweede patterns are tensible, for instance, if the seed 40x40 graph roper, five or sax sets of goes basic over itself in the background color before color crayons. continuing.Again, because the students are having Sequence of Events: to hand-simulate durino their off-machine time, The first day of this lesson consists of skills are being built ladle students are planning introducing the program and allowing students to =ewes. The teacher nqa-only give very short play with itfora few minutes each.Theyare sent machine wows times during this phase, requiring any with copies of a sheet of exploration pro that a filled-out crayon simulation be presented as jeststo beundertaken the awn day.These pro- a ticket to allow its being tried on the ccmputer. jects are carefully ordered by increasing difficulty On the fifth day the class is given over to a so that the firstgroup onthe machine will have show-and-tell, in which each group explains km it sane chanceof success, whilethe second group 411 found the mow to its exploration question and have todo ,scaleplanning before getting on the on:- shows off the most interesting seed.Polaroid puter Mile the first group is working) in order photos or slides of each group's best pattern are bDsucceed. The exploration projects are: taken for use in the et-of-course Parents' Day Group 1: 93Wmany colors are there? ed-Abit. Are they distinct, or are somerepeated? Lesson 2: ACS How are you going to be mare? Behavioral Objectives:Students will learn how to Group 2:Can you paintwitieblack,or isit like control the screen tte-tle via individual amends the paintbrueh not touching the painting? no the TWERP progran. They will explore the fea- WMhigh and wide isthe screen? How tures of =ME:GRAPHICS and of teSTE, a neusic-output many colored blocks are there? function. Geoup 3:How cony blocks (pixels)per second can Concepts:Students will learn the folleArsg: this =cuter draw? 1)the idea of using cateends that are weird* Group 4:Can you fire apattern(seed) belch, when to get desired output. They should under- run, 411 fill theentire screen with one stand that connaais consist of operators color? Whet is the shortest or* you can (verbs) and (nouns or adjectives). f ird? 2)the conceptscperirent construction Group 5:Let's say you were allowed only seven and hand-simulation (which are introduced keystrokes to make your seed.If I stuck by Lesson I and reinforced here). a piece of tape on the screen somewhere, Resources: can you make a seed which will zap a line nINTERP, the PASCAL interpreter program. through my =irked text?(It might go MEM will be used throoehout the Munn. around several times; that's ok.) ,In this lesson, the insediate :rode will be The second and part of the third days are used; as each command is entered it is spent giving each work group (two or three students) executed.The canaries used this week tea minutes access time to the caqouter, while the (listed below in Day 1 Evens) control the other groups catch or plan. The teacher :roves position of an imaginary screen turtle about observing, providing hints, and showing which leaves a colored trail as it moves. students low to pretend to be the ocmputersusing A musicetaking canard is also explored. graph paper and crayons.We turn the scarcity of 2)Six small protractors. assurer access to our advantage here by requiring 3)Six pre-made, but unmarked, cardboard the students to be careful and precise in their rulers long enough to reach diagonally hart-ainulation of the repeat-process in order to across the Te screen. predict :bat their prcgrams will do. Sequence of Events: During the third day the teacher deteridnes bo the zirst day, the teacher shows students

2S3 4. Computer Games in Instruction269

how to start INIERP and enter the immediate mode, will he given IURITEGRAFH/CS-oarmands to produce a then types the following six lines, pausing after design at abet* level S above andasked to draw the each while the Class observes the effect: figure on graph paper as it would appear when drawn PILISCFEEN(HIES) by IVICETAWAPHICe. FILISCSEEN(BLACK) THE Ox IT& STYLE OF THE HSCS CURRICUILlet PBCOLORHereret At idusTeint-the reada may ask, lefty do you mow(20) call these activitiesgames at all? What have your Itge (50) lessons in cannonwith Blackjack or lunarLander? HOVE (20) The fact that unifies Quilting, TURTLEGRAPHICE, These contaands are written on the blacklaard. and more traditionalinteractive gazes is that they one sore emend is added, AVVE'IO(20,20), to be sell thenselves. No one has tocanpel studentsto used if the line being drawn goes off the screen. do their assignments. Students are allowed two minutes each on thecon - The point at which our curriculum diverges pater to try these cowards. The concept of can - fran closed genes isthat the only realopponentin mends (operator, operands) is briefly explained. traditional games is a pseudo - random amber gener- Fbur discovery questions are then given to the ator or perhaps another Inman.In a cognitive game class: the onaenent is the rich structure of our own I) *at is the size of the screen? Where is ignorance. The excitenent of being able to create the center? pattern and order is as old asthe wallpaintings 2)Had Bony angles are there? What do nega- in thecaves of France. It is an essentially tive angles do? hewn activity, one at which all players can win. 3)How many colors are there? What are they? It is also a meta-gene; in which an infinite number What happens when various eeneolors and ofspecific games like shoot-the-dot can be tee/wound colors are used together? expressed.The relationship between cognitive 4)Try the camend BYTE(nuater, timber) .What genesand traditionalgenes is analogous to that does it do? Oat is the effect changing between a set of blocksand a preassembled toy. A the first number? The second number? What different order of learning becomes possible. are the legal ranges of_each Tarter? THE DEEM STILE C*? THE HSCS CURRICUILI4 Each group is given responsibility for one question. The rundenetwardesign princWre have fol- Tie groups get question 4. lowed is to attewpt bomake.each lesson augmentthe Cat the second day, work groups attempt to students skills in three areas:discovery, con- ewer their questions.They are given a descrip- trol, and design.We elide,/ studentsto play with tion of the concept of sheeting a dot, which they the* system as each new feature is introduced, but may try when they have answered their questions. the? have discovery questions Onee answers they (They need those answers to sheet the dot.) seek as they mass around in more or less structured 'lb shoot a dot, a piece of colored tape is ways. They need to find the answers to be allowed stuck to the television.The students try to pass access to the next levelofthe system.Students a turtle track through the cot.Doing it with one develop discovery skills by experimentally answer- TURN and one WIVE is a successful shot. The teacher ing questions nee "what does this cameed do?" will provide protractors and (blank) rulers and Ile ask students to undertakespecificchal- will suggest that students make television oilers, lenge, such as the shoot -tine -dot game, to develop but without telling them lame their ability to control the receputer by selecting After having deronstated the ability to shoot the correct cartrand and pzovieing correct values a fixed det on the screen (beginning of Day 3), the for its operands.Their understandingof the system team will graduate to production geometric figures is built by simulation exercises, which alive thee of the Collaring types increasing in difficultyt to predict the behavior of a catmand and tteteto 1)square choose the right command. 2) rectangle Later in the sweeter, students will begin 3)isosceles right triangle writing prcgrantsr but even at early stages there is 4)equilateral triangle the impetus to desi input sequences to produce 5)rectangle with diagonals the desired pate. Students must be able to pro- 6)trapezoid luoe a sequence of cayman& which produces the pre- 7) 5-point star natal output on first submission in order to B) Star of David graduate to the next level of the system. 9)"naked dandelion" (radiallines from a Another principle we have followed can be sum- cam= point) med up in the phrase "design from the first experi- 10)a short Mlle, in block letters ence." Mt believe that =touter science (or any- The first fivedesignsabove will be attempted thing else) should be taught firm the inside out. by the teens on Day 3, with the first teem to the That is, first experiences mist incorporate the tezminal getting design 1, the second tarn getting heart of the trebter at hand, with as little ettrai. design 2 andworking first at their desks with graph nexus matter as rossielc. For instance, paper, protrictors and rulers, and so on through all teaches the frrlamental coreofthe computing the teens.On Day 4 or after they have mastered all experiences in zepetitiotrif a =trolled process of the first dive designs, the teens will tackle there is greet weer, The Quilting lesson is taught designer 6e10.on Day 5, the class as individuals without introducing a word of jargon, pi vices

2S:3 270 NECC 1980

assigements, or complex camend sequences.alin- REFERENCES ing, and its fundamental message, can be taught to illiterates.The second lesson similarly teaches Aiken, R. M., Hughes, C. E., and Mbsbell, J. M., the relationship between operands, operators, and "Computer Science Curriculum for High School results.Only after students have firm operational Students," Proceedings Prbi/SIOCSE Conference, skill with a given tool, do we introduce terminol- Kansas City, Montana, February 25, 1980. ogy, written reference materials, and the ultimately Papert, S., "Teaching Children Thinking," Proceed- necessary environmental details such as data ings WW World Congress on Computers and Education, declarations and control statements. Preswerdam, 1970. This paper has addressed the important question of "Om Vans" microcomputer education? We Postmen, Neil, "The First CUrriculonsComparing are excited by the prospect of transforming gaming, School and Television," Phi Delta Kappan, 61:3, a traditional problem area for computing teachers, November 1979. into one of their primary tools.

Ihe authors aclarmaedge and appreciate the assistance of their collaborators:R. M. Aiken, C. E. Hughes, C. R. Gregory and J. A. Ross (University of Tennessee); L. Demarotta (H. C. Maynard High School); E. Miner (Alcoa High School).

Figure 1:The Seed Pattern

..;',1-. -i- .1 - ., .--:-.;-.. -, .r , I 3-42:-.-- , ; --r-T ---.--111-1. o 1 ,l.e. 13:1-irtI-1- , I'l ""."11-1-1-1- r r --1--1 -; +-Ealmi-4 .---i-n-r. ' :- :- I l I .1--i-417'1- --i-7 17:::-.-t-rrrrizi-r ,.,..!_rt.". 'frri-r 14" +1:1-1 : 111 11:0:1T 1- 4-i-

II if H

Figure 2:The Repeated Pattern

2-(% Computer Games in Instruction271

TAPPING THE APPEAL OF GAMES IN INSTRUCTION

Peter 0. McVay Educational Services Digital Equipment Corporation 12 Crosby Drive Bedford, MA 01730 Tel. 617-27S-3000 Ext. 2217

Games and the Computer effective computer game. Human languages and thought processes Around the time when computers began simply arenoteasily trans- appearing in classrooms and ceased being ferred to a computer, except on a novelty, thetheory arose thatthe the simplest level. (Note the ideal way of teaching with a computer word "easily " - -there are bril- was to put instruction in the form of a liant exceptions to the above game. Actually,this teaching method statement.) has been around for years; pioneers in the Dewey teaching method used many The pendulum recently has appeared games in their curriculum, and a number to swingin the opposite direction: of Montessoritechniques also involve games are now anathema and viewed as a games. frivolous pastime at best. But this Probably one reason the game idea so approach ignores that games are immense- caught and held the attention of educa- ly popular and have a tremendous at- tors with computers was the mania for traction for game players and designers computer games in general. Really bril- both. This paper proposes to extract liant students(and instructors)spent some of the good points about games and an inordinate amount of time designing then apply them to real computer-aided and programming new games. Other in- instruction. structors and students spent a huge amount of time playing these games. Why Salvaging Valuable Parts fight it? Design instruction to take advantage ofthe built-in motivation; What points are worth saving? What there is a poolofcourse developers can games do that other methods of in- anxious to developinstruction (i.e., struction do not do as well?The obser- games) and a large audience of enthusi- vable characteristics of computer games astic students (players). (and computer game players) are: A high level of motivation. What Happened: Expectations Denied Game players and designers spend hours working at the With a few exceptions,the results terminal. were notably disappointing. Problems Clear and consistent goals. arose on two fronts when games were made All true games have a clear an integral part of computer-aided in- ending in mind. How much instruction: struction becomes bogged down because ofmuddy goalsand The game quickly took over objectives? the instruction and frequeAtly A high amount of player in- became more important than the teraction. Game players are content of the lesson. doing something: they are Language-oriented, subjects manipulating the terminal, the required a huge amount of pro- computer, and the game struc- gramming effort to bend to an ture itself.

284 272 NECC 1980

Maximum choices for the Play- Questions for the Developer er. Studies of children's toys 'Rive shown that the flashiest Is the method of presentation con- toy is not theone that is sistent with the material? The best played with most often. The computer games require playeractions toy that gets the most atten- that fit the total concept of the game. tion is the simplest one that Players move tokens by giving directions allows the child to use imagi- to the computer in a manner similar to nation. The most popular games picking up the piece by hand (for board (on or off the computer)are games). If the game is a thought-type those with thesimplest cC1.- game,the player makes decisions that structs and widest range of are consistent with real-world decision choices. methods. The computer in both instances Simplicity. How much in- is a referee,informing the player of structionfails because the the consequences of his actions. starting sequence for the stu- Poorly designed instructional games dent is too complex? Complex- or simulations use methods of advancing ity in itself is not the hor- players (students) that are not related ror--the problem is that no one to real actions. In a number of compu- bothers to explain all the de- terboard games, studentsrace cars, tails to the user. Most good horses, or other items around a track. games come with very detailed But instructional racing games make the instructions. (Note: why does pieces move around the track by answer- the programmer who balks at ing questions. These games are not no- providingdocumentation churn tably successful because the action is out reams of instructions for too slow and because races are not nor- the game he just designed?) mally run by answering questions. This Creativity. Many of the as- design error occurs because the devel- pects of computer games-- wheth- oper has inferred that actions that game er they are traditional or in- players enjoy in ale setting must be vented by the user--have highly good in all settings. This mistake re- creative parts. An axiom among sults in a product that is neither good game designers that is fre- instruction nor a goo'' game. Lesson de- quently used to justify their velopers can avoid this trap by chosing interest in games is that some a presentation method without regard to of the most highly creative the observed popular appeal of a method ideasand programs aretirst in a different setting. A natural, or developed in a gene. consistent, presentation enhances the appeal of a lesson considerably. If the Applied Gaming Principles subject is math, the game should use mathematics naturally, not as a means of How can the principles from games be moving a token. The Minnesota Educa- applied to instruction? It turns out tional Computer Consortium has developed that the items that make games soap- economic models that allow children to pealing are intertwined, and by incor- run simulatedbusinesses (lemonade porating severalfacets of gamesinto stands, bicycle factories); both mathe- genuine instruction, several objectives matics and economic fundamentals are can be achieved at once (e.g., high mc- taught through thesegames. English tivation, strong interest,. good inter- teachers can use word processors avail- action). able on most computers in their clansen. The remainder of this paperis a Students take naturally to word proces- checklist of items to look forin any sors--they enjoy seeing their work computer-aided instruction. These are turned outin a professionalformat. items which have been found to most re- There are many other examples--the key liably, increase the quality of computer is to ensure thatthe presentation instruction and, it is hoped,' the skills matches the material. of students taking the course. An important point about this list is that it Is there a large amount of player Is not meant to be exhaustive--it is a interaction? f'requently the charge start. Persons wanting to use the ques- rated againsttelevision can also be tions as a checklist may also have their raised againsteducation: the student own principles and procedures to add to sits and absorbs the material passively. the list. The lectureformat certainly has its place, and brilliant and stimulating Computer Games in Instruction 273

lecturers are rare individuals that into both projects. This problem is a should be sought after. But computer- management, not a computer, problem. aided education as a lecture or elec- There can be several reasons for this tronic page-turning format usually does dichotomy when it appears: not rise to these heights--and is also a bad use of the medium. 1. Freedom to experiment and An examination of the most popular make errors. If you insist on games shows a high amount of player in- -)r-fand perfect instruc- teraction. This interaction is not sim- tion from your developers, that plykey-pushing; some of the popular is exactly what you'll get- - gamesare alsolimited to single-key with all the creativity and or- responses after lengthy actions by the iginality of the Saturday-morn- computer(football, adventure, road- ing cartoon shows. A perfect race). Butthe player is constantly lesson is not necessarily a thinking -- decisions must be made, paths good lesson. In fact, lessons chosen, strategies worked out. that haveincomplete sections Again, an objection will be raised may even be more useful, since that the material in education cannot be the student must fillinthe adapted to such a lengthy decision- missing portions. making format. But an examination of 2. Appropriate comments at the actions that students take when appropriate times. Kibitzers studying shows several possibilities for seem to abound in educational increasing the student's participation: development areas.The discussion of approaches and critical 1. 'ath choice. When a stu- analysis of lessons under dev- dent Tritairii away from the elopment is an important part classroom, he at least has the ofthe creative process, but option of choosing which page there are many cases wherea or chapter to start on. Why not promising bit of instruction include this choice in the com- hasbeen discarded because puter session? constructive criticism was 2. Notes. It is relatively heapeduponthe material--be- simple to provide an electron- fore it was developed and be- ic scratch pad--each student fore advice was requested. can be provided with a comment While a developer is writing a or note option during the les- game asa hobby, no one is son. At the and of the lesson, leaning over hisshoulder. the student can either store Leave the developers alone the notes or (if disk space is until thereis something to limited) take a printout of the criticizeother thana first notes away. draft. 3. More inquiry. This does 3. Responsibility. Managers not mean more questions. In- frequently lick confidence in quiry can bea form of path- the persons working under them, taking, or an invitation to andhope to avoiderrors by speculation. The student giving very detailed instruc- chooses which subject to take tions to the developers. If a first. Incidentally, this also developeris presented with a answers thefrequent question cut-and-dried package, there is of the correct order to study little room for a new and crea- subtopics. The student can tive approach. For someone to choose the topic and also skip becometrulyinvolved witha from topic to topic if neces- project, she must feel that the sary. Thesame material is projectis atleast partly eventually covered, butalong hers, and must have some in- paths that the student has cho- volvement with planning. And sen. responsibility extends down to some very low levels. Ina Is the material creative? One language arts project in Nor- puzzling problem in computer-aided in- folk, Virginia, the computer struction is that the same individuals operatorswere employed to that produce ho -hum instruction also are enter some of the questions turning out brilliantly designed games. students would be given. They Obviously the same effort is not going were asked to use their imagin-

2S 274 NECC 1980

ations when entering the praise 2. Method of presentation. Is response students would receive theformat(question and ans- on correct answers. The re- wer, screen display, essay and sults were a series of ques- choice) appropriate forthe tions with highly original re- subject? sponses that the students could 3. Effectiveness. Does the relate to. The operators also content of the lesson stay with took a personal interest in the the student? development of the course, and 4. Correctness. Is the con- frequently monitored student tent free of grammar, syntax, progress andaskedto change spelling and content errors? Lessons that were not popular. (Here is an area where younger students delight in showing Is everyone using the material en- off. If you choose to release thusiastic?Bete is another factor that them on this one, prepare for is a function of the people involved and an avalanche.) not the computer. 'You WILL be enthusi- 5. Additional lessons. What astic:" is obviously an unworkable ap- other material shodIrle added proach (butone whichis sometimes or changed? tried:). But if everyone involved with 6. Design of instruction it- the development of a particular piece of self. Developers can work with instruction finds it tedious and diffi- Mir students, or students who cult to work with, then perhaps the en- have just completed the course, tire instruction set should be looked to redesign and evaluate the at. If persons who are working in their instruction. subject area find the topic boring, then what about the students? What is needed Is the course simple and direct? is either a fresh and creative approach One fault of some of the indiVre (see above) or a different topic. If a courses is that they tend to be immense- subject is completely boring,then its ly complex. While at some time in the value is open to question. Tepid topics future humans may learn to think recur- usually resultif theindividuals can sively and in complex algorithmic pat- see no value in them whatsoever. ternsanalogousto the machinesthey Enthusiasm,once engendered,tends use, presently people still think in the to be catching. Enthusiasm most fre- same old way. Presenting highly complex quently stems from freedom and responsi- study structures, maps, objective bility--two subjects discussed under fields, and learning paths can set up a creativity and maximum choices, above. forest that quickly discourages even the Freedom and responsibility extend to all most determined students. levels: supervisors, developers, tea- If the structure istoo complex, chers, and students. Teachers and ad- there are three solutions: ministrators tend to be wary Of allowing students totake part inthe teaching 1. Check the presentation. In process, andfor good reason:student complex structures, a simple, freedom is difficult to manage and con- direct path can frequently be trol. Real participation in their own found in the material. education is also a novel idea for most 2. Turn routing over to the students; they have been passive consum- student. If given a list of ers of instruction for so long that they tWand objectives, the stu- may not be able to handle such a respondents can make their own sibility without careful pre-training. choices of what to study first, But the results can be spectacular. and when. Here are some areas in the instruc- 3. Ride some of the complex- tional process where personnel at all ilx. taiTrUistructure may level: can directly participate: be-due to the enthusiasm of the developers for marvelous struc- 1. Level of instruction. Too ture. A lot of the design can hard?Too easy? Should it be be safely hidden by having the presented to a different grade computer do the routing work. level or in a different subject area? Great advances have been made in the use of computers in classrooms in the past few years, and the pace of develop- sent and discovery increases with each

2 =, Computer Games In Instruction 275

new application. The procedures discussed in this paperare a starting point for developing better computer- aided instruction. The key to successful computer-aidedinstruction, as to any other form of instruction,is the people involved in the project. There is no substitute for a creative, dynamic, and highly qualified individual in the right position at the right time. Fortunately, these persons appear to be available because computer-aided instruction has made great gains in recent years. The continuing analysis of what constitutes good technique in the use cl computers will continue this trend. Computing Curricula

AN EDUCATIONAL PROGRAM IN MEDICAL COMPUTING FOR CLINICIANS AND HEALTH SCIENTISTS

Albert Hybl James A. Reggie Department of Biophysics Department of Neurology & (301) 528-7940 Department of Computer Science (301) 528.6484

UNIVERSITY OF MARYLAND School of Medicine Baltimore, Maryland 21201

ABSTRACT

The growing importance of computers in medi- medlOal computing has recently been singled out as cine implies that health professionals lacking ru- an area of great importance that deserves more at - dimentery knowledge of their potential use will be tentiOn than it hes received (7, 12, 18). Very handicapped in the future. In this paper we dis. little literature has addressed this gojl and there cues the content and implementation of a curriculum has been no clear consensus about which aspects of in medical bomputing that we have introduced at our computer science should be taught to medical stu- medical school to remedy this problem. In addition dents and physicians.How this material should be to formal classroom instruction our proves in taught and how it should be introduced into an chides such innovative features as a demonstration already information.dense medical education process laboratory for exhibiting computer applications in have been given little if any attention. medicine, the use of a family of knowledge mange. sent languages that are designed for computer - It is our belief that the growing influence of inexperienced clinicians, and the involvement of computers on medical care makes some familiarity interested individuals in ongoing researoh in nor- with their uses and potential important to health puter applications in medicine. Our description of professionals. We have thus designed and imple- this curriculum and its implementation should be of mented a program tc supplement conventional medical interest to anyone involved in tesohing computer. education with instruction on medical computing. inexnerienced individuals about the potential uses This program has two major OnjnotiVen: of computers. (1) to provide health professionals with a broad INTRODUCTION understending of current and potential uses of computers in medical researoh and clinical In recent years the role of computers in medi- practice; and cal education has been receiving a great deal of attention.The vast majority of this interest hen (2) tc give health professionals the ability to centered on oomputer.aided inatruotion for a wide directly use available computer reeouroea for range of olinioel and pro.olinical topics (3, 4, 8, solving clinical and researoh problems. 9, 14). At the present time many medical schools use such autcsated instructional material to In this report we begin by out ..ing the con- supplement conventional teaching methods. tent of a curriculum on computer applications in blomedisine that we believe achieves these objeo. In this article, however, we will discuss a Elves. he then disouse how we have approached the complementary aspect of computers in medical educa- implementation of this program at our institution tion: introduoing medical students, prastioing and the response it has generated aeons medical physicians, end other health-related personnel to students and fatuity members. Finally, our expert. the applications of computers in medicine. Iduoat. *noes are reviewed in the context of other resent . ing individuals in the health professions about eadOevors in this,area.

2Q9 278 Computing Curricula 277

A MEDICAL CONFUTING PROGRAM O Social and Legal Issues: A discussion or the present and future effects of computers on The content of our medical computing program the practice of medicine is given ano reflects our feelings about ghat topics are cur- eludes material on ethical issues, standard- rently important or will become so in the future. ization, and the economics of medical comput- 03r 00100 of topics is based largely on modioal ing. The Privaoy Act of 1974 (PL 93-579), (imputing systems that are presently used and a re» the Federal Brooks Bill (FL 89-306), and fed- view of recent literature on biomedical computing eral information processing standards are research.Our emphasis is on the use of the com. examined. voter as a tool to accomplish various tasks rather than the fundamental conoepts of computer science. The section on scientific and laboratory ap- The materiel in our program can conveniently be di- plications contains the following material: vided into five sections: I Real-Time Instrumentation:Configurations of O General Concepts microprocessors and minicomputers for data O Soientifio and Laboratory Applications acquisition ranging from stand-alone dedi- O Clinical Applioation cated facilities to hierarchical computer O Educational Applications networks are discussed.Techniques for col- O Administrative Applications lecting and communicating data to the oosput- er are stressed. The section on general oonoepts is tailored to increase student awareness of the oapabillties of O Graphics:The use of terminals with graphics computers and to provide then with the fundamental capabilities and plotting systems is explored.Their ability to extend the effec- knowledge needed to use a computer. It includes: tive use of computer resources by improving the presentation of certain types of informa- O Essentials of interactive Computing* In order to entourage students to become computer tion is covered.The plotting of results users, we begin by explaining the mysteries from the digital simulation of a biophysical of account numbers and password*, the tech. system is used as an example of the suppor- niques for signing on and off, the idiosyn- tive role of graphics. crasies of using certain terminal*, the mar- vels of the executive system, and how to save O image Prooessing:The role of computers in data and/or progress in retrievable files. various clinical scanning systems (e.g., CAT, (Batch processing or programs is disoussed nuclear) is discussed. Attention is given to the tuture..pntential of image analysis in the but not used by students.) . laboratory (e.g., ohromoscse analysis, cell oountins). O Document Prooesaing and Text Editing: These systems are easy to learn and iamediately useful to the students. Their mastery faoil- Statistical Analysis of Dates The basic cm- itates the learning and use of other oomputer oepts of data analysis are introduoed and resources. available packages for statistical processing of data are desoribed (0.g., SPSS, MOP). Example applications :mob as epidesiologic O General Information Processing:Programming languages for intonation prooessin4 and for data analysis Sr. provided. general problem-solving are surveyed. Guide The seotion on clinical applications covers lines are presented for selecting an seem- priate language for various tasks. topics dealing with the use of computers in patient oarc and clinical research including* O An Overview of Applications:An initial overview of the range of scientific, labora- O The User-Computer interface:The use of com- tory, alinioal, and educational application puters for obtaining s patient's history and the problems of making computer facilities of computers in medicine is given.This in. eludes SISOSSSIOO of boil working systems direotly usable by physicians are explored. such as a whole.body CAT scanner; a cranial A brief introduction to natural language CAT seamier; two nuclear scanners; computer- processing by aaohine is provided. ised baste solemn research laboratories; the Q8.2 System for ICU patient data colleetion, O Medics' Information Systems:This inoludes storage, and mitering (at the Maryland the use of registries, data bases, and hospital Institute for Emergenoy Medioal Services); a information system for storing and retrieving manic:al records for use in patient partially implemented PROM'S system (17) (at care and clinical researoh. the Baltimore Cancer Research Center); and Aspects of se ourity and integrity ere examined in detail. the Stroke Database Center (located adjacent The oonoept of a query language is introduced to the Neurology Mari, this unit serves as an and ourrently available systems are reviewod entry point to the sultioenter Nations' Strike Database OM. (e.g., COSTAR, PROMS, ARAMIS).

290 278 NECC 1980

Computer-Assisted Medical Decision-Making: the medical community. Tne potential uses of computers for assisting physicians with diagnosis and patient manage» To maximize the student's experience, we nave sent are discussed. Concepts of knowledge adopted a teaching approach that uses several representation and alternative inference different instructional methods. Much of the ma- methods (s.g., statistical pattern classifi- terial is presented as formal lectures that are cation, production rule systems, cognitive complemented by related reading assignments from models) are covered. the current literature. The use of lectures and reading assignments allows us to cover the wide The section on educational applications covers range of concepts outlined above in the brief time material relating to any aspect of the health pro- available. However, to give mediCal-students a fessional's training. It includes: more intimate acquaintance with the impact of the computer on health care and medical research, we O Computer-Aided Instructions The available have supplemented the traditional teaching approach systems for computer -aided instruction (e.g., in three ways. PLATO, ASET, MEN) are reviewed and informa» tion on how to author suitable lessons is First, students are made cognizant of the di- presented. versity of applications through tours to on-campus facilities where dedtcated computers are in active O Education in Medical Computing:. The atti- use (the sites are listed above under "An Overview tudes of health professionals toward using of Applications").These visits effectively empha- the computer for medical applications are size the present role of computers in the health explored.The concept of a curriculum on professions and contribute to student motivation to medical computing and a survey of existing or learn more about biomedical computing. proposed programs are presented. Tne second way we are supplementing classroom Referent:. Systemss Computerized literature teaching is through the development of a laboratory searches (e.g., MEDL1NE) and more advanced for the demonstration of applications of computers medical knowledge bases are described. to specific research and to clinical and educational problems that are not currently in day -to-day The section on admiestrative applications use at our hospital.EKO analysis (currently being surveys the use of computers in hospitals end phy» implemented) is especially suitable for demonstrat- sician offices for scheduling, accounting, analysis ing the major components of laboratory computing: of polioy-related issues, and other related tasks. data acquisition (analog - digital signal process» ing), feature extraction (R -wave recognition, ONST CURRICULUM IMPLEMENTATION classification, power spectrum analysis), and decision- making (presence or absence of myocardial Currently we have implemented formal course- infarction).Several computer-aided decision- work that encompasses most of the material outlined making systems are complete (e.g., diagnosis of above. This coursework began with only a single thyroid disorders and stroke; prediction of outcome one»credit minimester course that covered essen- following cardiac arrest) and are used to illus- tially all of the material listed above at a fairly trate the current level of state-of-the-art medical euperfioial level. We have recently added a second decision support systems. Feedback from medical one-credit @animater course and divided the ma- students and physicians is useful for guiding the terial between the two courses permitting it to be development of the experimental systems that are presented in greater (loath. Although both courses being demonstrated. begin brfatiliarizing students with the essential elements of interactive computing, they are largely The third aspect of our instructional approac: complementary in what they cover. One course ad- involves hands-on experience using the computer fa- dresses general concepts, scientific and laboratory cilities at our university. In one of the mini - applications, educational applications, and ad- wester courses students are expected to use the ministrative applications.The other course Con- Document Processing System to prepare a critique of centrates on clinical applications.Both courses a current journal article of their choloe. The Ore offered twine a year and are taught in a coor- students are also given projects involving digital dinated fashion.Students are permitted and en- simulation, computer graphics, general problem couraged to take both courses. solving, and using statistical packages. In the other course that covers mainly clinical applica- Since biomedical computing has not been gener- tions, students are asked to develop two small ally recognized as pert of the medical curriculum, computer -aided deoision.making systems using a fan our approach has been to gradually introduce this; ily of knowledge management languages designed for material into the medical school program. lie feel use by computer - inexperienced individuals (15). that the development and evolution of our program The systems supporting these languages provide a has profited from an incremental implementation. oomplete computer -aided decision - making system when The gradually increasing amount of ooursework ap- given a description of the problem to be solved and propriately parallels the amount of computer use by tha relevant knowledge noeded to solve it. Dy Computing Curricula 279

Ins the computer for these projects students are hands -On computer experience, visits to relevant able to learn directly about the capabilities and facilities, and the use of a demonstration limitations of the computer for assisting them with laboratory. a wide range of tasks. Our present plan is to continue the evolution At the present time our elective minimestar of our minisester elective program through addi- courses are reaching about 5 10% of the freshman tional courses and by instituting a more formal and sophomore *lasses.Students taking these evaluation of their success.The demonstration courses have a wide range-of ccaputing backgrounds, laboratory needs further work; we need to develop varying from essentially no previous computer and acquire additional software and hardware. science to an occasional student with an undergrad- Long-range goals include the establishment of medi- uate degree in a related field (e.g. biomedical cal student externships in advanced topics in medi- engineering). The variety of ihstruotional ap- cal computing, integration of at least some form of proaches used has been fairly successful in inking required instruction on biomedical computer appli- the material presented understandable to cosputer cations into the standard medical ourriculum, and inexperienced students. At the sue time the broad the creation of a continuing medical education scope of the material has provided a challenge to course to reach practicing physicians. even the most advanced students.Student evaluations of the courses implemented so far have been Atanaldikilltlitat he thank the Departments of essentially positiv4 other faculty umbers and Diagnostic Radiology, Pharmacology and Experimental u niversity administrators have begun to take an Therapeutics, and Radiation Therapy, the Maryland interest in the program. institute for :urgency Medical Services, and the Stroke Databank for graciously accommodating our DISCUSSION class tours of their facilities.Computer time for this project was supported in full by the Computer Previously proposed or implemented programs Science Center of the University of Maryland. Dr. for education in medioal =outing fell primarily Reggie gratefully ackLooledges support for this into two ostegories. First, and most common, are work from an NIB Teacher-Investigator Development those representing complete curricula that consist Award (5 KO7 NS 00349). of interdisciplinary studies in medicine and computer science. These ourrioula are designed to produce individuals who are specialists in biomedical computing and thus typioally had to graduate degrees.Several examples of such ourricula have been described (1, 6, 13, 19) and a resent review of relevant progress in the Federal Republio of REFERENCES Oermsny has addressed this issue (10).

The seoond category of educational progress in 1. Ackerman, L. V. & Harris, D. K. (1975). medical oomputing include those consisting of only rohitecture for a Graduate Level a single *Ours* designed to supplement the educa- Educations/ Program in the Area of Computer tion of health professionals (e.g., nurses, plume- Systems in Nedlotne,wAuggproopentwolszi elate). For the soot part, this courses have coo- MagmaLansaw cont*ranatVP765468' *red a very limited range of topics (2, 5, 16). 1. Buokwell, L. J. (1979). *Education of Health The medical computing ourriculum that we have Care Students to Acoept and Use the described ih this report lies in between the two Cominutar, ItHOgullnii SetIna 3tst.,1280111111 different oetegorles outlined above.On the one Soimakoz 1881liaLlosAnManisaSam band, it is designed to supplement the traditional edited by Dunn, R. A., 206-209. Nov York* educational experience of medical students and oth- IBEB. er health professional' stellar to the single course programs.On the other hand, it consiats of 3. Dora, J. G, Ne;4an, A., Nasmond, V. E. 6 more than a single course and oovers a wider range Haynie, C. (1978)."Computer-sided of topic".The depth to whioh these topics are Instruotion in Clinical Neurology,* journal oovered is gradually increasing es the available sitilesilaalignaatia..,23(8):693. oouraework 'woes's. 4. Deland, E. C. (1978). IducgatiggIgghgalggg In this report we have attempted tc desoribe AnAIWA &WM llibioAtutog New !oft: Plenum n ot only what the content of a medical °eluting Press. currioulus should be, but also how it night be implemented within the constraints of the tradi. 5. Dumas, B. P. (1979). *A Computing Course for medioal school experience. In particular, Pharmacists,*lamunliana sitant .3rsL. we have presented our belief ih the need for an Zionntan an Smut= laalinattaa jii Salm/ incremental implementation and an argument for SAKI, edited by Dunn, R. A., 210.213. New supplementing oomventional lecture material with Yorks IEEE.

292 280 NECC 1980

6. Duncan, N. A., Austing, R. H, Katz, S., 17. Schultz, J. R. & Davis, L. (1979)."The Pengov, M. E. & Wasserman, A. I. (1978). Technology of PROM1S," ISSE/tooeedIngn, "Health Computing: Curriculum for an Emerging 61(9): 1237-1244. Profession, Report of the ACM Curriculum Committee on Health Computing Education," 18. Shannon, R. H. & Duncan,K.A. (1978). "Why Pr dinar of Ing eGli 1.921 &liana a Curriculum in Health Computing?" Conference, 277.285. New York: A.M. hsasslinta of Ihn Ardi 142 liatinna SPererena, 273-276.New York: ACM. 7. Duncan, K. A. (1979)."Educational Programs for Health Care Computing," Prooeedinen sit 19. Wasserman, A. I. (1979)."Educational the 3rd. =min an Daunt= lantiantinaIn Programs in Medical Computing Their Basis gattratfarg,edited by Dunn, R. A., 204-205. and Implementation,"hstnesunna,g,m3rg, New York: IEEE. Asonalui fagot= Andinatinn In Medical Slign, edited by Dunn, R. A., 217-222. New 8. Halverson, J. D. & Ballinger, W. F. (1978). York: IEEE. " Computer- Assisted Instruction in Surgery," Surgery, 81(6):633-638.

9. Kenny, G. N. & Schmulian, C. (1979). "Computer-Assisted Learning In the Teaching of Anaesthesia,"inalabgaii, 352):159-162.

10. Kocppe, P. (1977). "Education in Medical Informatics in the Federal Republic of Germany," of Infarnalina In Mining, 16(3): 160-167.

11. Kunitz, S. C., Havekost, C. L. & Gross, C. R. (1979). "Pilot Data Bank Networks for Neurological Disorders," prooeedluna slum IYERoiluman raw= Anginal= In Medical fare,editedby Dunn, H. A.,793-797. New York: IEEE.

12. Levy, A. R. (1977). "Is Informatics a Basic Medical Science?" Yroceedlnea.MC1101=2.7.. edited by Shires, D. B. & Wolf, H., 979-981. New York: North Holland.

13.Levy, A. H. & Chen, T. T. (1977). "Plana for aProgram in Medical Information Science," Erstantginant 1221 Intel CsinuLtr. Conference, 46: 321-325.

14. Madsen, B. W. & Bell, R. C. (1977). "The Delopment of a Computer - Assisted Instruction and Assessment System in Phareacology,"Mdicalsdueation, 11(1): 12 - 20.

15. Reggie, J. A. (1980). "A Domain-Independent System for Developing Knowledge Bases," Iracanslinna Ord.] Wang Sanatanne af Ma amain And= lac faunatntlana Malin of latsaLlinannt, victories B.O.' Canada. (in press).

16. Rowley, B. A. (1979)."Computer Medioine Educational Program for Medical Students and Others," Proceeding ithi 3Eg, $ ignaly. an Samna Anoliettian In Mina am edited by Dunn, M. A., 214.216. New York: IEEE.

2° r) Computing Curricula 281

A SECONDARY LEVEL CURRICULUM IN SYSTEM DYNAMICS

Nancy Roberts Lesley College Graduate School of Education 9 Mellen Street Cambridge NA02138 617-547-8844

Ralph N. Deal Chemistry Department Kalamazoo College Kalamazoo MI 49007 616-393-9452

SYSTEMS THINKING IN EDUCATION

For quite some time, the influence structure and selection of information). of science on society and education bus Mathematical models of varying degrees of been primarily reductionist, seeking to complexity are built to reflect the system understand phenomena by detailed study out of which a problem grows. Computer of smaller and smaller parts. Over- simulation is used because behavior implic- emphasis in this direction has led to a it in the structure of complex systems is separation of allied wciences like physics too involved to be solved by direct math- and chemistry. "Learning more and more ematical methods. Model building and com- about less and less" (Odum, p.9), as one puter simulation is therefore an integral critic described reductionism, has taken part of the system dynamics problem-solving place with almost the complete absence approach. of a movement in the opposite direction System dynamics starts from the pree to integrate knowledge. Education tends tical world of observation and experience. to follow the pattern of parts within It does not begin with abstract theory nor parts. Knowledge is divided into dis- is it restricted to the limited information ciplines, and disciplines are further available in numerical form. Instead it divided into subjects. The student is uses the descriptive knowledge of the oper- seldom exposed to materials which seek to_ ating arena about structure, along with integrate the various disciplines and available experience about decision. making. subjects. Such input* are augmented. where possible by Education ar well as the everyday wrLtten description, theory, and numerical world of social and economic analysis has data. Feedback principlee are used to a pressing need for synthesis and holism. select and filter the information that gives The systems approach seeks to understand the structure and numerical values. how parts fit together to form a whole System 1961 that functions for a common purpose. It when Professor Jay W. Forrester of UIT pub» Is in the world around us that most lishe Industrial:ta 6: students will be making decisions for the field of system dynamics broadened in its rest of their lives. Explicitly under applications, it also broadened in its standing the nature of these systems student audience. Initially taught as a from a structural and a behavioral point graduate level course at MIT, it has spread of view constitutes a most relevant and to the undergraduate curriculum in many useful education. universities. In 1975, ae a dissertation project, Roberts introduced the basic con- THE APPROACH: SYSTEM DYNAMICS cepts to fifth and sixth grade students (1978). The evidence that fifth and sixth System dynamics is a method for under- grade students could understand and apply standing and managing complex social systems. these basic concepts lad to the development It is built on traditional management (infermof the warrant secondary level curriculum motion, experience, end judgment) and feed- projectbeing funded by the Office of back theory or cybernetics (principles of Education. The project is based at Lesley

294 282 NECC 1980

College Graduate School of Education, realizing the central role models play in Cambridge, Massachusetts, under the our thinking (especially in the scientific direction of Nancy Roberts. method) allows asking the, proper question- not whether to develop a model, but what A SECONDARY LEVEL CURRICULUM ZN SYSTEM kind of model is most helpful for a given DYNAMICS purpose. System dynamics helps us to under.. The current curriculum development stand problems arising in dynamic systems, group is writing and pilot testing a set first by making our mental models explicit, of six self-teaching, introductory learning second by incorporating into them feedback packages. This material will make it possi- relationships (learning packages If, III, ble for a teacher in any discipline to and IV of the curriculum), and third by introduce model building and computer providing means for developing them into simulation as a ptJblem-solving method, unambiguous mathematical models (learning The titles of the six learning packages V and VI) when the complexities packages are: become too great for mental models to I. Basic Concepts: Dynamic handle. Problems, Systems and Models II. Structure of - Feedback Systems II.Stucture of Feedback Systems XII. Graphing and Analyzing the This part of the sequence introduces Behavior of Feedback Systems a way of thinking holistically abort IV. Understanding Dynamic Problems cause and effect. The principal tool V. Introduction to Simulation employed to analyze a system is the causal,. VI. Formulating and Analyzing loop d.agram. Cause-and-effect relation- Simulation Models ships are symbolized with arrows, forming An overview of the content of each chains of causal links. Loops result when learning package,follows. some or all of these chains return to their starting points. These loops of I. Basic Concepts: Dynamic Problems, causal influences are called feedback Systems, and Models loops; they are the central focus and Embarking on a course of study in foundation of system dynamics. system dynamics, students need to under, Causal»loOp diagrams are visual stand the basic concepts that underlie models of feedback systems. In part the field. Three concepts are central; of the curriculum they are used in a goner the focus is on dynamic problems; the ally descriptive way to summarize the com- intent is to consider the whole system plex interactions in a variety of stories, of interacting parts from which a problem problems, and systems with which the stu- arises; and models are explicitly employed dent is fel:tiller. The following is one of to carry out the analyses. Bach of these the storiea from this learning package concepts is explored briefly in learning followed by the possible csusalrloop die. packago I to provide a foundation for the grams that students might develop. study of feedback systems which follows. A problem is dynamic if it changes over time.Urban crime rates, for example, rise; the economy cycles, as do pendulums and people's feelings of depression and MS OIL CRIS2S elation; central city populations decline, and so do reserves of natural resources: A system is defined as a collection of parts operating together for a common purpose, but the concept is sometimes better left undefined to be inferred frog; examples. The notion should connote complexity, but it should also suggest a wholeness of perspective and the 'feeling that the whole is greater than the sum of its parts. A model is a representation»usually a simplification--of some slice of reality. Pictures, verbal descriptions, graphs, sets of equations, and laws are models. Think. Jug can be characterized as the manipula.- tion of mental models; the real system is never in one's head. Such a concept may seem too generalized to be useful, but

2():;

11 Computing Curricula 283

One aspect of the oil crisis, as the connection between feedback loops and explained by an economist, was the starting dynamic behavior. of a vicious circle. This vicious circle was begun by agreements made by the Arab M. Graphing and Analysing the Behavior oil-producing countries in 1971 called the of Feedback Systems Teheran and Tripoli Agreements (named for Dynamic behavior of variables in the cities in which the meetings were held). systems is the focus of this learning pack- Here these countries agreed to raise the age. The purpose is to understand how the price of oil. The rise in oil prices structure of feedback loops is responsible meant that these countries made more money. for the behavior of variables in the system They made so much more money that they over time. Graphs are introduced and used could not possibly spend it all. Real-, intuitively, often without specifying ising this, these countries decided not to nuaegical scales. The amte of the graph produce as such oil. They knew that even of a variable over time is the concern. tually their oil supplies would run out so Behavioral characteristics of positive they might as well make thelplast as long feedback loops and negative feedback loops as possible. are explored in simple situations and then Because there was less oil being exploited to analyse more complex multiple- produced in the world and more oil was loop structures. Positive loops are shown needed every day, a scarcity developed. to be responsible for uncontrolled behavior This scarcity of oil forced the oil prices such as exponential growth, while negative to go up even higher, continuing the loops are shown to produce goal?seeking vicious circle. behavior, though often with disturbing The following causal-loop diagram fluctuation* and cycles. S»shaped growth explain he economist's vicious circle in several apparently different systems is theory. shown to occur when a quantity is influenced first by a positive loop and then by a negative loop; a shift from dominance by the Oil Produ d of Oil positive loop to dominance by the negative ki) loop produces behavior which looks initially like unrestrained exponential growth but refits 4. then becomes goalvmeeking. An important principle will appear as different systems Figure 1. The Vicious Circle are explored: systems with the same feedback structure tend to behave the same over In causal-loop diagramming, as in time, in the sense that the shapes of their multiplicatin, two negative elements graphs over time are essentially the same. cancel each other out, creating a positive The role of delays in goalvseeking feedback effect. systems is explored intuitively. Students The diagram below expands on the will become familiar with the principle information given in this story and might that delays can cause a variable to oscil- result from a class discussion. late around itmgoal. The approach in this part of the + curriculum is essentially nonvquantitative, Amount of Amount except that some quantitative graphing will Oil Left Oilt Price of 40Used by be done to provide the tools required to In Ground Pr cord Oil Cons eers graph intuitively the behavior in causal- loop diagrams. ... Students completing learning package ProfitsJAPI. III will have a foundation for the principle that the bshaVior of a system is a cense Figure 2. The Oil Crisis quence of its feedback structure. They will be ready to try to apply their under- The principal skills the students will standings to more complex situation*. develop in learning package II are the ability to represent the essential Le» IV. Understanding Dynamic Problems finances in a problem or system as a 'A number of recurring themes in real., causal-loop model and the habit of search. world problem' are uncovered in this part Leg for feedback influences which close of the sequence. Each theme is explored causal loops. if students couplet* this in the oontext of several apparently dif- part of the curriculum but do not cony. ferent dynamic problems, making use of tinue on in the sequence, they will causalvlocp models and the properties of have ,gained a powerful tool for under, feedback systems developed in Sections II standing complicated interactions, but and III. The structure of feedback loops they will have only begun to understand responsible f.; the thematic behavior is

296' 284 NECC 1980

exposed in each system, provi'ding a common a structure of feedback loops are in focus for understanding the different doubt. Greater precision is required than problems sharing that theme. a causal-:oop diagram can provide. Section One of the themes explored here is V develops the skills needed to translate sometimes referred to as the counterintu- simple causal-loop models into quantitative itive behavior of complex systems= models which a computer can trace through intentioned policies often tend not to time, simulating the behavior of the actual produce the behavior expected, occasion. system. ally producing results opposite to those Two critical notions form the t 4481 intended. The phenomenon is traced the concepts of a level (or stock) ada the initially to the distinction between open concept of a rate (or flow). Level var- loop and closed-loop thinking. The former iables are pictured as rectangles; the overlooks feedback in contrast to the rate adjusts the flow of something into latter, which incorporates within the the level with which it is associated, boundary of the system all the essential just as a faucet adjusts the flow of water feedback influences. Counterintuitive into a tub, changing the water level. behavior appears again as the significance of feedback structure is explored; feed. back systems tend to resist certain kinds of change, unless the actual structure of Sires hpu- Birth Population the system is affected. In feedback Rate V Y systems the short -term effects of a policy may be different from, even oppo site to, its long-term effects. The tend.: Figure 3. Causal-Loop Diagram as It Relates ency of complex systems to become dependent to a Flow Diagram upon external controls is explored under the theme of "shifting the burden to the This part of the Curriculum returns intervener". The concept of tradeoffs is to the work of learning packages II and introduced. In complex systems, policies III, developing quantitative understanding rarely improve all aspects of the system of positive and negative feedback Loops at once; usually a policy improves some and the behavior of simple systems. Stu- areas and is deleterious to others, rem dents will expend their skills in under- quiring policy-makers to teke explicit standing and interpreting graphs of var- account of the tradeoffs. iables over time. In addition, they will The problems explored in this part develop abilities to write general level of the sequence range from peer pressure, equations and the elementary rate equations cramming for a test, and mowing lawns, to for exponential growth and decay and sig- criminal justice, pollution, urban growth.; moid (S-shaped) growth. Exercises include stagnationi.decay, drug-related crime, and first-hand calculatin7 and graphing the global population and food needs. model variables, then simulating the models Learning package IV is the canine with the aid of a computer. Familiarity ation of the essentially now.quantitative_ with computers and programming is not re- part of this introductory system dynamics quired; the introduction to the simulation curriculum. It shows the power of attempt language DYNAMO is self-contained. The ing to understand complex dynamic problems programming is presented as a means to an by focusing on feedback loops, to illustrate end: making assumptions sufficiently pre- some themes which recur so frequently in cise and suitably coded that a computer complex systems that they may be called can trace out their implications over time. "principles", and to leave the students Having hand-simulated the models first, the with a balanced view of the power and limitr students will know what the computer is ations of their understandings at this doing. point. as learning packages II, III and At the close of learning package V IV have progressed, the need for knowing the students are ready to understand more the relative strengths of feedback loops complex simulation models, they will have in a system will have arisen at various begun to see the power of simulation, and times, leading naturally (but not ;moos. they will have solidified their understand- eerily) to the next part of the sequence ings of the behavior of feedback loops in which methods are developed for making covered intuitively in learning package II the assumption. embodied in a causal- and III. loop diagram unambiguous by quantifying them. VI. Formulating and Analyzing Simulation Models V. Introduction to Simulation Several of the problems addressed in Computer simulation in system dynamics learning package IV are reconsidered, gen- becomes necessary when the implications of erally in greater detail. For each, in

2 Qt Computing Curricula 285

turn,a quantitative model is developed in ment Program. The secondary level design DYNAMO, and explorations of the system are should also be appropriate for many under- carried out by simulating different con- graduate college classes, as well as for ditions in the 'model. The central focus, other possible learning settings. The besides the significant problems them- learning packages have been planned to selves, is the understanding of complex meet the following criteria: system behavior. Where in a given system 1) Interdisciplinary and Unifying. The does intervention have the most effect? materials will provide examples from a Why does the system behave as it does? number of subject areas:physics, econ- What policies actually improve the be- omics, biology, ecology, sociology, anthro- havior of the system?Why does one pology, social studies, and literature. policy have a desirable effect while others The generic similarity of structure and which initially appear promising have little behavior in these fields will be illus- helpful effect or may even prove to be trated. harmful? 2) Relevant. The materials will deal with Explorations of the behavior of a significant world and national problems system are carried out, and alternative as well as problems encountered by students policies investigated by changing,numer- in their own lives. ical relationships in the model, alter- 3) Hands-on Approach. The set of learning ing, or adding equations. The computer packages will teach students the modeling is shown as an obedient servant tracing process through a series of exercises. out the implications of a modeler's Initially, the students will 4evelop on assumptions over time. Each simulation their own a set of feedback diagrams de- model and each simulation run appear rived from the verbal description of a not as ends in themselves, but as means problem. Then they will go on to quantify to understanding,the dynamics of a cer- the elements of the problem through graph- tain problematic system. The goal is ing, and finally write the equations re- understanding, and feedback models help presenting the structure of the model. us to understand certain kinds of problems. The students will then hand simulate their Learning package VI completes this model and finally simulate the model with introduction to system dynamics. Stu- the aid of a computer. dents continuing through all six sections 4) Supplementing Rather Than Replacing will have a new understanding of the causes Existing Subjects. The curricular of dynamic problems and the beginnings of materials are designed to supplement rather a set of tools for analysing and under- than replace current courses. For example, standing them. They will have learned the materials will be used to demonstrate to look at problems holistically and to integration in a mathematics course, popu- search for feedback loops responsible lation growth in a biology course, factors for the behavior of the system. They underlying social problems in a social will understand the role of models in science course. approaching problems and what is re- 5) Reiterative. Roth the problems selected quiredto develop such models. The stu- for study and the skills taught will re- dents will also,have an introduction to appear in several learning packages. The the meaningful role computer models and students therefore will study the same simulations can play in helping people problem areas with increasingly more details to cope with the complex dynamic problems of problem elaboration and increasingly they face. more sophisticated methods as they proceed from causal-loop diagramming through graph- THE CURRENT CURRICULUM PROJECT ing, to equation writing, hand-simulation, and then computer-simulation. Each skill The project is scheduled for comple- that is taught will be recalled for use in tion in August 1980. During the spring of later exercises that focus ostensibly on 1980 it will be pilot tested in six high other skills development. Further, the schools. Two teachers from each high student's sense of the similarity of under- school will integrate the curriculum into lying dynamic structures will also be in- their courses. These teachers are from creased as problems from a variety of dis- the disciplines of Computer science, ciplines are studied from a system's per- mathematics, chemistry, physics, English, spective. The students will begin to history, environmental studies. And biology. understand how identifying the underlying The targeted audience of secondary structure of a problem gives one a strong students need not be the only user, of the sense of comprehending the problem. materials. The materials that were created The curriculum development group. for fifth and sixth graders have been used hopes to do extensive field testing over in high schools, universities, and even the next few years. Should anyone be in in the MIT Sloan Fellows Executive Develop- terested in field testing the learning 25 286 NECC 1960

packages or wish more information on the projec-, please contact either of the authors. minim

1. Forrester, J.W. Industrial Dynamics. Cambridge, MAt MIT Press, 1956. 2. Odom, X.T. Energy, Power and Society. New Yorks Wiley-Interscience, 1971. 3. Roberts, N. "Teaching Dynamic Feedback Systems Thinking* An Elementary View." Management Science, Vol. 24, April 19794 pp. 836 -943.

2919 Computing Cunicula 287

THE COMPUTER SOFTWARE TECHNICIAN MANI

AT PORTLAND COMMUNITY COLLEGE

1980

,David M. Hata Portland Community College 12000 S.W. 49th Avenue Portland, Oregon 97219 (503) 244-6111

ABSTRACT evident that the software equivalent of the hard- -This paper is a status report on a new voca- ware technician did not exist. tional program being offered at Portland Community With the increase in the number of micro- College entitled "Computer Software Technician." computer-based systems being manufactured, shifts The curriculum was developed jointly by Portland in the skill sets required in the manufacturing Commohnity College and an advis'ry board composed environment were projected to move trward increas- of representatives from local electronics manu- ing emphasis on software skills.Translated into facturing companies such as Tektronix, Intel, and future manpower requirements, an addition of Electra -Scientific Industries and is targeted at several hundred software professionals was pro- .-- r- software/hardware technician positions in the jectedover the next five years, an impossible O.E.M. environment. task in the face of a critical shortage of technical people. Using past experience in developing INTRODUCTION electronic technician programs at the two-year Portland Community College is a comprehensive level, the two-year concept for training software community college serving all of Washington County technicians began to grow among managers. and parts of Multnomah,Columbia, Clackamas, and Development of the program began in March Yamhill counties, a region of some 679,000 people. 1978 when Tektronix formally approached Portland Within the geographic boundary of the community Community College with the idea of offering a two - college district are located several large manu- year associate degree program to train computer facturers of electronic equipment -- Tektronix, software technicians. An ad hoc committee was Intel, E.S.I., and others. formed and instructors from the Electronic Engin- Courses offered at Portland Community College eering Technology and Data Processing Departments are organized and integrated into a broad variety were assigned to course development tasks. of programs ranging from associate degree, certif- icate career, and college transfer programs to PROGRAM GOAL special interest and enrichment courses, appren------urrart programs in electronics and data ticeship training, and high school completion. processing/computer science emphasize either hard- The philosophy of Portland Community College is ware-related or software-related topics.The goal to offer learning opportunities to everyone re- of the Conputir- Software Technician Program is to gardless of prior educational experiences, a train a technician who has the skills to write and philosophy which has earned the school the name. develop microcomputer applications software under *the Educational Shopping Center.*It is in this the guidance of a software engineer and can bridge environment that the Computer Software Technician the hardware/software gap by posses:Mg-skills Program is growing. that will enable him/her to verify correct system operation. PROGRAM DEVELOPMENT eneedfor computer software THE COMPUTER SOFTWARE CURRICULUM technicians began in late 1977.A survey con- The proposed Computer Software Technician ducted within various groups at Tektronix showed Program consists of 101 credit hours for the ass- widely varying skills among those polled. For ociate of applied science degree. The first year example. the Information Display Group's skill consists of perallm eardware and software course set was much more computer oriented, greater than sequences combinpr ith mathematics and communica- 40 percent software, as opposed to less than 20 tions courses. C: wads develop basic problem- percent for the Test and Measurement Group. It solving skills whi:m Wilding a foundation in was also noted that groups with more software basic circuit theory and device operation. The engineers had fewer technicians. It became software and hardware areas merge in the fourth

300 288 NEcc

term as courses begin to integrate both software and hardware topics into a unified presentation. The program concludes with courses and on-the-job learning experiences, to create the ability to function as part of a development team and learn the art of project management.

TERM I Lecture Lab. Hrs. CST 2.211 Software Programming I $TN 125 Computer Oriented Mathematics I 5 5 2 WR 2.301 flatness Communications I 3 3 EL 6.117 Basic Electric Circuits 6 4 6

CST 2.221 Software Programming I 4 4 2 ITN 126 Computer Oriented Mathematics II 5 5 WR 2.302 Business Communications II 3 3 EL 6.127 Fundamentals of Semiconductors 6 6

TERM III CST 2.231 Software Programing III 4 4 2 CS 233 Intro. to Numerical Computation. 4 4 2 WR 227 Technical Report Writing 3 3 EL 6.137 Digital Logic Fundamentals I 6 4 6

TERM IV EL 6.248 Introduction to Microprocessors 4 3 3 CST 2.241 tow Level Languages 4 4 2 SP 100 Basic Communication 3 3 EL 6.247 Digital Logic Fundamentals II 4 3 3

TERM V CST 2.123 Language Processors 4 4 2 CST 2.136 I/0 & Data Cbsmunication Prog. 4 4 2 PSY1.546 Asychol &ammo 3 3 EL 6.257 Peripheral Circuits 4 3 3

TERM VI CST 2.126 Project Management 4 4 2 CST 2.132 Operating Systems 4 4 2 CST 2.141 Field ProJect 6 1 18 EL 6.267 Advanced Micro Systems 4 3 3

IMPLEMENTATION laboratory Facilities: Basic Circuits Laboratory Continued: Parallel efforts to upgrade existing electron- Triple power supply ics laboratory space used by the electronic engine- Function generator wing technology program and to acquire the specialized equipment necessary to implement the Com- Advanced Circuits Laboratory: puter Software Technician Program have produced a Dual-trace oscilloscope (100 MHz) well equipped set of laboratories to support both Digital voltmeter programs. Our initial goal to equip each labora- Triple power supply tory with industry-standard equipment has been Function generator realized. Specialized equipment for the Computer Digital counter Software Program purchased for laboratory use includes logic analyzers, microcomputer treinerc, and Digital System Laboratory: software development systems. Dual-trace oscilloscope Digital voltmeter Equipment needed by laboratories: Function generator Triple power supply Basic Circuits Laboratory: Logic analyzers Dual-trace oscilloscope (35 MHz) Microcomputer trainers Digital voltmeter -so

Computing Curricula 289

Computational Systems Laboratory: areas in which persons interested in computers, Microcomputer systems capable of running computer programing, and electronics can enter: BASIC and PASCAL Programs of this type have not been available in Microcomputer development systems the past. To industry, these technicians will provide the analogue of the hardware-oriented Cooperative Work Experience: hardware technician that has been missing. A cooperative education program has been established with local companies, and ten CST CONCLUSION students are currently participating in the pro- The need for computer software technician gram. Each cooperative education student is re- programs will grow, and the Computer Software quired to carry twelve credit hours of academic Technician Program at Portland Community College work while working twenty hours per week.Close can serve as a model for programs at other supervision is maintained between the cooperative colleges. It is the result of close industry - education coordinator and managers of the coopera- college interaction and is currently being eval- tive education students. uated through a cooperative work program with The response has been favorable among parti- local electronics companies. cipating managers; student response has been positive although many have expressed the feeling that the academic work load combined with work assignments has been demanding.

Faculty Upgrading: Program development has reinforced the need for a teaching faculty having a skill mdx that spans the hardware/software boundary.To date, faculty upgrading has been accomplished by technical seminars, independent study, and interaction with industry. Keeping pace with an ever changing technology is a continual struggle, and incorporating new teaching ideas into courses remains before our faculty.

Course Content Guides: Course content guides are being reviewed for content and updated.These course content guides are available upon request by writing to the author.

IMPACT ON COMPUTER EDUCATION -nate processing and computer science pro gramsto date have been very business applications oriented.Curricula are heavil), weighted toward programing languages and applications that do not fit the treditional product development cycle. The Computer Software Program was a totally new program rather than a modification of an existing program. The electronic engineering technology am data processing Programs were left intact since they are currently serving a well-defined industry need. in this manner, a curriculum was developed to meet a future need for technicians with both software and hardware skills. With the first gradmaiing class in June 1980,'software'technicians capable of fitting into systems development and manufacturing groups will be entering the Job market. It has been evident from the ftedbeck obtained through managers of our cooperative education students that the skill mix will be right for entry-level posngititheseons in software groups currently employ -

programs of this Ups are opening up new

302 290 NECC 1980

A COMPUTER SCIENCE MAJOR IN A SMALL LIBERAL ARTS COLLEGE

Joerg Mayer Department of Mathematical Sciences Lebanon Valley College Annville, PA17003 (717) C67-10SS

This presentation examines the diffi- separation between the two cultures shows culties, both practical and philosophical, no signs of dissolving, at least not in which stand in the way of offering a com- education. puter science degree at a small four-year Computer science as an undergraduate liberal arts college. AlsoI will present major brings this dichotomy in education the compromise which was struck at our into sharp focus, perhaps because this college. discipline is predominantly machine and Until recently, computer science as an process oriented. The aim of computer undergraduate degree program was almost science is not to seek truth or to under- entirely restricted to the larger uni- stand man and nature, and in that sense versities. Such programs were expensive, it is anti-humanistic. Computer scien- both in terms of teaching personnel and tists are as seen by the humanist, equipment, and required an environment essentially non-reflective, meaning that which included an engineering school. they do not concern themselves with the These conditions are easing somewhat, and ethical and moral implications of their it is becoming increasingly difficult for work. (Weizenbaum's book on that issue a small college to resist the incentives is not widely known; andin any case, to enter the cornuter science market. it does not seem to be very popular among Computer hardware is now within the means his colleagues.) Therefore, computer of most institutions, and for those with science is considered by many to be anti- inadequate funding there are ties into a thetical to the basic philosophy of commercia2 or educational time-sharing liberal education.*One only has to ob- network. Also with declining overall serve the nearest computer freak to find enrollments on the horizon and an at best proof for the dehumanizing, indeed deper- stable interest in most traditional sub- sonalizing influence of that machine.* jects, the continuing shortage of college- Computer science, as a relatively young trained computer personnel seems to offer discipline--exuberant, irreverent, and in the chance to shore up at least some of some ways irresponsible and antagonistic the drain in the student population by to deeply held cultural values- -does not offering a major in computer science. easily fit into the age-worn quilt of Encouraging though the emerging con- liberal education. ditions may be the arguments against Even if the misgivings were somehow to adding computer science to the program of be overcome, however, the practical prob- a liberal arts college are rather unset- lems of offering computer science in a tling. small liberal arts college are most for- Despite the efforts of many, there bidding. remains a sizeable gap between the defend- To begin withfinancing the needed ers of liberal education as the attempt to technical expertise is beyond the capa- capture the essence of the complete man, bilities of most such institution.. It and the proponents of an introduction to would be unwise, both politically and science and technology as the only means financially, for the college to hire a of preparing our youth for the technolog- beginning Ph.D. in computer science for ical future. The present mixture of major $25,000 when the average salary of the programs in most liberal arts institutions assistant professore in that college is reflects more an uneasy truce than the $14,000 and that of full professors is result of mutual understanding.The $22,000.Another manpower problem arises Computing Curricula 291

from the ACM standards set in Curriculum technical courses included programming '78. There it is recommended that *ap- (advanced), computer architecture and proximately six full-time equivalent assembly language, data structures, and faculty members are necessary. ...* With an independent study. Of the 36 mathe- a student :faculty ratio of 14:1 this matics graduates since 1977, six took translates into roughly 80 computer sci- computer-oriented jobs, seven became ence majors, a number not likely to be system analyst/programmers, and four reached in an institution of 1500 students, entered highly respected computer science which is the typical enrollment of a small graduate departments. four-year college. Encouraged by these results, and Another difficulty lies in the courses strongly persuaded by the Admissions which should be offered. According to Office, a major in computer science was Curriculum '78, a minimal program in com- added to the programs of the Department puter science consists of twelve of Mathematical Sciences. Its structure courses in computer science and seven is as follows. All students in the De- courses in mathematics, making a total of partment, regardless of their major, take 57 credit hours in the major. Rare is \ the same core curriculum which consists the liberal arts institution that would Of& allow such a concentration of a student's time in his or her major. Analysis 113 hours Finally, there is the need for hard- Foundations of Mathematics 3 hours ware and software. Again, Curriculum '78 Differential Equations 3 hours established certain benchmarks. "It is Introduction to Computer Science3 hours essential that appropriate laboratory Linear Algebra 3 hours facilities be made available that are comparable to those in the physical . . . During the difficult sophomore year, disciplines . . .. The initial budgetary the students assess both their capabili- support . . . may be substantial.* Also ties and their strongest interest. mentioned are extensive software systems. Accordingly, some will drop out of com- The whole package, even if conservatively puter science and others will shift into interpreted, presents a financial burden it. For that major the remaining re- that reaches well into six-digit numbers. quirements are: Realistically, then, it is impossible for all but a handful of the best endowed Abstract Algebra 3 hours small liberal arts colleges to find the Classical and Numerical Analysis 3 hours total budgetary support for the start of Computer Organisation & Assembler 3 hours a computer science major which approaches Data Structures 3 hours the standards set by Curriculum '78. So Programming Languages & Compilers 3 hours a student who wishes to major in computer Internship 3 hours science must limit his choice of a college to those with an enrollment of at The internship must be taken in some least 5000. It means that the smaller commercial or industrial computer opera- college cannot enter the market of com- tion, usually during the summer. The puter science education.Finally, it languages taught during the last three means that employers and graduate schools years are BASIC-PLUS, FORTRAN, and of computer science cannot reach into that Assembler (PDP11/40).We strongly enrich pool of above-average, motivated, courage the mastery of one additional and broadly educated graduates. language, such as RPG or COBOL. . At Lebanon Valley College we wore drawn There are other requirements.To give into the computer science field more by the students some bacuground in the basic accident than by design. In the early components of computer hardware, they seventies when it became clear that the must take a year of physics and work in a major in actuarial science was very Beall computer science laboratory. Also attractive to prospective Students and required are six hours in psychology, and could be so designed as not to violate three hours in an ethics course designed the precepts of liberal education too to deal with the ethical issues inherent severely, the Department of Mathematics in modern technology, and computers in decided to expand its offerings into the particular. Finally, having observed the applicable areas. The department grew often abominable technocratese of hand- steadily, so that in 1978 there were 70 books and manuals, we require that the majors in a school whose total enrollment computer science major take a three hour was 950. In 1975, enough computer science writing workshop. courses hadbeen added to offer an inform- Thus, the major consists of 31 hours al,concentretiOn in computer science. The in mathematics, including the Curriculum

3u4 292 NECC 1980

'78 courses MA 1, MA 2, MA 5, MA 3, CS 17, teaching the normal twelve hour load. CS 18, with MA 6 strongly encouraged; The Computer Center personnel must be -twelve hours in computer science, includ- sympathetic to the reality that the new ing CS 2, CS 3, parts of CS 4, also CS 7, major will tie up the system more often and CS 81 18 hours in what we consider re- than they are accustomed to. And the lated topics; and three hours internship. administration must be willing to accept The totals are 46 hours in technical the fact that one cannot start a tech- courses and 18 hours in supportive disci- nology-based major without improving the plines. existing hardware and hiring at least one These requirements are in contrast to specialist. It is well to remember that Curriculum '78, which sets a minimum tech- more often than not, the necessarysup- nical requirement of 51 hours, of which 30 port will come at least two years after are in computer science. it is needed. The disadvantages of our approach are For those who may be contemplating at least the following. It makes us Un- implementing n variation of our a2proach, comfortable to be in such disagreement let me close on a bright note. Within with Curriculum '78, and we would not be tilt first semester of the new major, four prepared to adjust if the suggested pro- students transferred into it from other gram of Curriculum '78 were to develop departments in the college, and two more from recommendations to quasi-accredita- students transferred into it from other tion requirements. Secondly, the rather schools because they wanted both computer narrow scope of the computer science science and a small liberal arts insti- courses must result in a limited appreci- tution. ation of the broad sweep of that field. And finally, the substantial asdifficult requirements in mathematics will most likely lead to a higher dropout rate than we are accustomed to because many students interested in computer science are mathematically not very gifted. The advantages are that we will be able to increase the enrollment in our department and that all the students in the department are being exposed to the new discipline.For them, the main and over- riding advantage is the flexibility which they enjoy during the first three years. To be able to switch, without loss of credit and required information, between mathematics, actuarial science, computer science (and probably operations research in the near future) is valuable to the mathematically gifted, who did not in their high sohoo1.4ears have any indication of the many job and graduate school opportunities for which they may be pre- destined. For me personally, the greatest advantage of our approach is that it shows our students, and not just in theory, the great breadth of mathematics and the mutual dependence and influence of the various disciplines we incorporate in our total program. Two conclusions appear to be clear. First, a small liberal arts college can offer a computer science major only if it does not adhere too closely to the recommendation of Curriculum '78. And second, such a Major has any chance of success only if it is incorporated in the operations of an already existing department. Its44Seeultyembers must be willing to work hard mastering most of the courses in computer science while they are still

3 06 Author Index

Abernethy, ,Kenneth 179 Johnson, Dale M.46 Allard, Me 230 Jordan, Eleanor U. 7, 260.44"' Alpert, Elizabeth96 Amenn, G.W. 266 Kneifel, David 37 Anger, Frank D. 171, 249 Kurtz, Barry258 Aronson, Michael 1 LaFrance, Jacques E. 261 Baird, W.B. 266-/"4"4". Little, Joyce Currie 220 Baldwin, R. Scott46 Lim, Tian S. 90 Bass, George H. Jr. 38 Lovis, Prank 75 Beck, James D. 245-it, Seidler, John A. 143 McVay, Peter O. 27101.46.0e40, Bishop, Judy H. 147 Hagnant, Peter214 Bork, Alfred 258 Manor, Walter 221 Bowman, William R. 16 Marshall, Sr. Patricia238 Boyle, Thomas 214 Mayer, Joerg 290 Brown, Bobby199 Heinke, John G. 143 Brown, Guy Larry 256. Mebane, Donna Davis 197 Brown, Thomas 237 Mebane, Rodney 197 Burrows, Robert L. 220 Hocciola, Michael 236 Bush, Steve 19 Morgan, Catherine E. 168 Hoshe1/4 J.H. Caldwell, Robert H. 31-A4-6'1- Hoursund, David 125 Collins, Ronald W. 74 Corbet, Antoinette Tracy 54 Oliver, Lawrence137, 169 Czejdo, Bogdan 112 Olivo, Richard F. 81 - 442749,01w.404

Davis, A. Douglas 86 Piegari, O. 179 Deal, Ralph H. 281 Poirot, James L. 130 DeKock, Arlan 138 Pollak, Richard A. 90 Dempsey, Richard F. 152 Powell, James D.1So Dershem, Herbert L. 65-Aw444. Dorn, William 75 Rahmlow, Harold 1 Reggie, James A.276--.44"4 Effarah, Jamil E. 25 Roberts, Nancy 281 Ellison, Robert J. 68 Rodriguez, Rita Virginia 171

Franklin, Stephen 258 Schimming, Bruce B. 58 Friedman, Frank L. 103 Shelly, Gary 3 Fdhs, F. Paul 205 Sobol, Thomas 155/14.0. Stokes, Gary 4 Garraway, Hugh 200 Stoutemyer, David R. 194 - 'e'~4' Carson, James W. 42 Sustik, Joan 199 Oats, A. John, Jr. 76 Gordon, Sheldon P. 169 Taylor, Robert P. 130,155 Oruener, William 198 Taylor, Timothy223 Groaa- Thornton, Joan C. 54 Thorkildsen, R'n 230 Thorsen, A.L. 179 Hagee, Michael W. 90 Tinker, Robert F. 250 Haiduk, H. P. 222 Turner, A.J. 6 Hannay, David O. 119 Hata, David H. 287 Varanasi, Murali R. 5 Hausmann, Kevin2, 170 Hazen, Paul 73 Ward, Darrell L. 19 Hopper, Judith A. 62 Wehrle, Howard F., III 12 Hunter, Beverly 168 Wet lore, David E. 139A4'4-'4V, Hybl, Albert 276 "A.e..4 Whittle, John T. 65/..r...et.. Wong, Pui-Kei184./4444

293