JCMST Volume 18 Number 3, 1999

CS-Ed research is an emerging sub-discipline, with all the energyof a young field, but also with attendant problems. Any emergentsubdiscipline must recognize its boundaries; a mature subdisciplinetypically defines itself in terms of exemplars, methodologies, andgreat practitioners. At this stage of development within CS-Edresearch, there are no standard texts, and there is no shared,systematically gathered corpus of exemplary research. This specialissue attempts to provide one collection of examples in order tocreate a visible identity for this emerging researchsubdiscipline—and to demonstrate its breadth and potential.

CS-Ed research is by its nature as an interdisciplinary specialty,drawing on Computer Science, Psychology, Sociology, and Education.Researchers trained in those subjects bring the researchmethodologies and knowledge verification systems of those disciplinesto bear in this de facto interdisciplinary area—giving it thepower of diverse methodologies and the potential to derive insightfrom “triangulating” different techniques and viewpoints.This special issue illustrates that diversity by assembling amulticultural body of examples that suggests the scope of thesubject. Our hope is that such collections will assist otherresearchers to integrate into the CS-Ed research community.

Computer Science is sufficiently distinctive as a discipline torequire that educational research come from “within” thediscipline. The accelerating convergence of technologies forcomputing, communications, and teaching poses a double challenge forComputer Science Education: developments affect both the subject andthe mechanisms of teaching. Educational methods race to keep pacewith the opportunities afforded by technology. Fortunately, theconvergence of technologies also affords an opportunity to integrateresearch and teaching objectives.

Hence, the theme of the special issue is “Research throughPractice,” to emphasize that our practice is our researchmaterial. CS-Ed research must be grounded in teaching and learningexperiences in order both to reflect on the nature and theory ofComputer Science Education, and to extend thinking about practice.This requires that research looks deeper than merely evaluatingimplementations, deep enough to examine what changes in teachingpractice reveal about underlying issues such as concept acquisition,development of skills and expertise, sources of misconception andsuperstition, learning processes, the roles of different types ofinteraction between teachers, students, and materials, and so on. Weneed to know not just the effect of introducing new technology ormethodology, but also the price. For this special issue, we sought inparticular examples of practice-based studies which gave insight intothe fundamental issues in teaching Computer Science, demonstrationsof methodology in Computer Science Education that included principledevaluation and appropriate generalization, and analytic evaluationsof practice that went beyond documentation to provide a strongtheoretical component that extended thinking about practice, forexample, comparing, taxonomizing, synthesizing, or deriving theoryfrom practices from more than one institution, or showing howeducational ideas have evolved over generations of presentation.

We also strove to set standards through a rigourous review processinvolving a distinguished international panel of researchers. Ourbrief required that papers should present documented research resultsor cogent analysis and contributions to theory. Papers should set thecontext for the work in terms of prior work, articulate thetheoretical basis of the research, provide clear description of themethodology used (whether qualitative or quantitative), presentclearly data that have been collected in a systematic manner, and tieany conclusions to evidence presented. The work presented here ismodest, as befits a young field, but the papers are diverse in topicand approach, and in presenting data on current practices andemerging trends they suggest many avenues for research in the future.

Overview of Papers

Vicki Almstrum’s research with over 850 students and thePropositional Logic Test showed ways in which one can predictmisconceptions and misunderstandings about logical operators. Herrigorous research methodology, which is situated well in previouswork, provides excellent research data and well-supportedconclusions. Her work challenges other researchers to continue toexplore the role of logic in the undergraduate Computer Sciencecurriculum and provide answers to the many other related researchquestions that her work has generated.

Cheng-Chih Wu, Janet Mei-Chun Lin, and Kai-Yin Lin provided anexcellent example of textbook analysis within the Computer Sciencefield. Their research, based on the situation in Taiwan, includedcontent analysis on four problem-solving steps. It encompassed 16textbooks and over 50 examples per book. This methodology, which isoften used in other disciplines but is rare in Computer ScienceEducation, provides us different perspectives on the structure of thediscipline as well as research within it.

Helen Sharp and Jacqui Griffyth explored the effects of softwaredevelopment experience on the acquisition of object-orientedconcepts. Their work included statistical analysis of more than 500responses to questionnaires and surveys encompassing students inmultiple presentations of a course over two years. They showed thepositive transfer from structured approaches and verified intuitionon the acquisition of fundamental beginning concepts. They clearlyidentified difficulties with the analysis and design phases. Theirwork provides research data situated in real context and adds fuel tothe current debate about how transfer develops within the objectparadigm.

Mary Granger and Susan Lippert examined peer learning andcollaborative skills in the undergraduate information systemscurriculum. Their extensive literature review provides us with aglimpse of best practices in this field. They extended this toinclude related classroom examples across the curriculum that supportthe peer learning methodology. Future extensions and implementationof their foundational work should provide excellent insights intopractical, meaningful ways to develop collaborative and social skillswithin the context of the content of the discipline.

Kenneth Modesitt, Bruce Maxim, and Kiumi Akingbehim employed a newmethod, just-in-time learning, in a software engineering course.Their work, which included multiple offerings of a course at oneuniversity, involved use of questionnaires and grades to compare thejust-in-time approach to more traditional approaches. This approach,which is a current new methodology being employed in Computer Sciencecourses, may offer one way to address current industry demands forpractical, pragmatic approaches to software engineeringeducation.

Kuo-En Chang, Kang-Yuu Wang, Chien-Yun Dai, and Teh Chung Sungalso investigated the use of peer learning and collaborative tools.Their work involved the use of Web-based tools by students learningrecursive concepts in introductory Computer Science courses. Theirclassical experimental design included pre- and post-tests, analysisof records of dialogue, and examination of processes used by thebeginning undergraduate students. Their work illustrates the use ofthe Web as a vehicle to extend previous results into a newtechnology, a trend that is often mandated by increasingly dynamicfield of Computer Science Education.

Nell Dale conducted a pilot study applying the cloze technique, amethodology extracted from other disciplines, in constructing testitems for undergraduate CS1 courses. Her research included thestandard experimental design of two treatments with swaying of thegroups and presentation order. She developed grading rubrics for thetest items and based her conclusions on the analysis of these items.Her groundbreaking work on extracting this technique, which is nowcommon in other fields, points clearly to areas where more extensiveresearch is needed.

Sally Fincher examined patterns and pattern languages for teachingacross the Computer Science curriculum. This “principledarmchair analysis” is a sound conceptual analysis that iswell-based in the literature. She points to directions for the futureand verifies the ‘power of patterns’ as a useful,fundamental contextual framework for future teaching andpractice.

Conclusion

The eight papers in this issue represent the collective effort ofmany individuals to produce a volume that illustrates clearly thevalidity and significance of research in Computer Science Education.During the past two years, this has been the passion of the editorsand many concerned, dedicated educators and researchers in the field.

The editors are deeply appreciative of the many individuals whocontributed to this work. When this volume was first conceived, weknew that we must rely on our colleagues in solicit contributions,prepare research manuscripts, and then serve as peer reviewers. Wewere blessed with a supportive community that responded to our callsfor action. Our review team was exceptional. Their insights providedus with the data that allowed us to make the critical decisions andshape the volume. Their careful, informed observations formed thebasis for the extensive revisions that took place after the selectionprocess was made. Our authors were patient and professional. Theyused this critical examination of their work to refine theirmanuscripts and provide us with the body of work that is presented inthis journal. They worked under short, tight deadlines and did sowith the utmost professional spirit.

Underlying the entire project has been the support of the AACEstaff and editors. JCMST Editor, Ed Dubinsky, has been a shininglight in this project. He nurtured it from inception to adulthood,providing us with careful guidance and wisdom many times along theway. We are deeply indebted to Ed for his support and assistance andfor placing his confidence in us to produce this issue. SarahWilliams, of the AACE staff, was responsible for the final productionof the issue. Sarah has worked tirelessly with us and the authors incompiling the manuscripts and materials and bringing it to press.

This issue represents the commitment of the JCMST to includeComputer Science Education as one of its basic components. We hopethat Computer Science Education researchers will recognize thisjournal as a worthy arena for their work and that there will besignificant contributions in this area. Our goal for the specialissue was to provide examples to help delineate this emergingsubdiscipline. Our challenge to the community is to keep thismovement alive by extending and documenting the growing body ofresearch within our field.

Return to Contents

The PropositionalLogic Test as a Diagnostic Tool for Misconceptions About LogicalOperations

Vicki L. Almstrum

Department of Computer Sciences

University of Texas at Austin

Autin, TX 78712, USA

almstrum@cs.utexas.edu

The concepts of logic are key building blocks in the knowledgethat computer science students must acquire. This paper describes asimple 16-item instrument, the Propositional Logic Test (PLT), whichis based on Piagetian theory. The PLT shows promise as a diagnosticinstrument to help identify students who systematically misinterpretlogical statements. This paper describes a study that explored theuse of the PLT in diagnosing student misunderstandings. Analysis ofindividual student responses revealed a number of systematic errorsin interpreting the logical operations in the PLT items. While thisstudy did not reveal strong correlation between PLT score andstudents’ final grades, other studies have revealed significantrelationships between PLT score and success in computer sciencecourses. Several areas for continued research are proposed, includingan invitation to other researchers to apply this instrument in theirlocal institutions.

Return to Contents

A Content Analysis ofProgramming Examples in High School Computer Textbooks in Taiwan

Cheng-Chih Wu, Janet Mei-Chuen Lin, and Kai-Yin Lin

Department of Information and Computer Education

National Taiwan Normal University

No. 162, Ho-Ping East Road, Sec. 1, Taipei, Taiwan106

{chihwu, mjlin}@ice.ntnu.edu.tw

In this study we examined 16 high school computer textbooks usedin Taiwan with an attempt to analyze the nature and the presentationstyles of programming examples in them. The presentation styles weredifferentiated based on the presence or absence of four majorproblem-solving steps: problem analysis, solution planning, coding,and testing/debugging. Our findings revealed that presentation ofprogramming examples in most of the analyzed textbooks lackeddetailed explanation of some of the problem-solving steps, especiallyproblem analysis and testing/debugging. With respect to the nature ofthe problems solved by the programming examples, most fell into oneof the following four categories: math problems, graphics problems,syntax problems, and real-life problems.

Return to Contents

The Effect of PreviousSoftware Development Experience on Understanding the Object-OrientedParadigm

Helen Sharp1

Centre for HCI Design, City UniversityNorthampton Square

London EC1V 0HB, UK

h.c.sharp@soi.city.ac.uk

Jacqui Griffyth

Advanced Information Technology

AIT Group plc, The Malthouse

45 New Street, Henly-on-Thames

Oxfordshire, RG9 2BP

This paper discusses results of a survey to gain insight into therelationship between the acquisition of Object Technology (OT)concepts and previous software development experience. A survey ofprofessional software engineers who had just completed a postgraduateOT course was undertaken, eliciting data about their previousexperience of software development, and their perception of thedifficulty of course topics and the level of understanding theyattained. Five hundred and ten responses were received over atwo-year period. These were analysed to establish which course topicswere the hardest and the easiest, and any correlation between variousdimensions of previous experience and the successful acquisition ofOT concepts. The results indicate that the fundamental concepts of OTand the management of OT projects are not difficult to understand,while analysis and design topics are harder. In addition, experienceof “traditional” structured and procedural approaches todevelopment have a significant positive transfer effect whichoutweighs the effect of any object-oriented orobject-oriented-related experience, and is substantially more helpfulthan having no development experience at all.

Return to Contents

Peer Learning Acrossthe Undergraduate Information Systems Curriculum

Mary J. Granger and Susan K. Lippert

Management Science Department

George Washington University

Washington, DC 20052

granger@gwu.edu

lippert@gwu.edu

Students concentrating in Information Systems not only needtechnical skills, but also communications and collaborative skills.Additionally, they need to learn to help themselves learn. Peerlearning techniques facilitate the combination of the required skillsand create an appealing educational environment. This paperconcentrates on the use of in-class peer learning exercises inseveral of the typical Information Systems courses: Innovation andTechnology, Management Information Systems, Structured Programming,Systems Design, and Database Design. There is an additional summaryof three semester-long peer learning projects in StructuredProgramming, Systems Design and Database Design. These InformationSystems courses involve many problem-solving tasks, for which peerlearning techniques are appropriate. Therefore, peer learningtechniques are introduced and combined with the standard lecturemethod. These exercises, introduced in many classes, are implementedthrough in-class small team collaboration.

Return to Contents

Just-in-TimeLearning in Software Engineering

Kenneth L. Modesitt, Bruce R. Maxim, and

Kiumi Akingbehin

Department of Computer and Information Science

University of Michigan-Dearborn

Dearborn, MI 48128, USA

bmaxim@umich.edu

Just-In-Time (JIT) Learning is a type of experiential learning inwhich students are motivated to learn about concepts and how to applythem as they are needed to complete a meaningful project. Use of JITby students in Software Engineering (SWE) can be an effectivetechnique for maximizing retention of core SWE principles. This paperdiscusses the use of JIT in a junior-level SWE course taught by theauthors. Having students work on projects for real-world clientsmotivates JIT. Analysis of student evaluation questionnaires andgrades in subsequent courses indicate JIT can be as effective astraditional SWE instruction. Both students and clients express highlevels of satisfaction with the experience.

Return to contents

Learning RecursionThrough a Collaborative Socratic Dialectic Process

Kuo-En Chang1, Kang-Yuu Wang1, Chien-Yun Dai2, and Teh-Chung Sung3

1Department of Information and Computer Education

2Department of Industrial Education

3Department of Educational Psychology andCounseling

National Taiwan Normal University

162, Hoping E. Rd., Sec. 1

Taipei, Taiwan, ROC

kchang@ice.ntnu.edu.tw

Web-Soc is a collaborative system for learning recursiveprogramming using a Socratic dialectic process. Web-Soc is built onthe World Wide Web (WWW). This study investigates the effects ofstudents’ learning of recursion and students’ communicationmodes when learning collaboratively using Web-Soc.

Return to Contents

A Pilot Study Using theCloze Technique for Constructing Test Questions

Nell B. Dale

Computer Sciences Department

University of Texas at Austin

Austin, TX 78712, USA

ndale@cs.utexas.edu

This paper describes a pilot study that investigates the use ofthe cloze procedure to design test questions in algorithmcomprehension. The cloze procedure is a prose-based technique used tomeasure the readability of text and students’ readingcomprehension.

Return to Contents

Analysis of Design: AnExploration of Patterns and Pattern Languages for Pedagogy

Sally Fincher

Computing Laboratory, University of Kent atCanterbury

Canterbury, Kent, England, +44 1227 824061

S.A.Fincher@ukc.ac.uk

This paper describes and delineates patterns and pattern languagesfrom different knowledge domains, and attempts to generate anunderstanding of why they are a distinctive and powerful way ofsharing knowledge. It surveys several examples within the genre ofPattern Languages and analyses the unique characteristics of theform, with reference to the instances surveyed. From this basis, itmakes the argument that a Pattern Language of Pedagogy is possibleand achievable and that for historic and disciplinary reasons CS issingularly well positioned to create such a tool and a particularlyfertile ground for its use.

Return to Contents