Educational Computing and Problem Solving

W Michael Reed, John K Burton

• Publisher: CRC Press
• Year: 2019
• Print ISBN: 9780866567817
• eBook ISBN: 9781000723885
• Topic: Business
• Subtopic: Information Technology
• Index: Business

First published in 1988. Professionals who are on the cutting edge of educational computing discuss, in this provocative new book, one of the most exciting prospects of the field--harnessing the power of the computer to enhance the development of problem-solving abilities. Here is everything that educators will need to know to use computers to improve higher level skills such as problem solving and critical thinking. Current aspects of problem-solving theory, a philosophical case for including programming languages in the curriculum, state-of-the-art research on computers and problem solving, and a look at problem-solving software are included in this comprehensive volume. The research and its application to instruction are grounded in problem-solving theory--making this book a unique and critical addition to the existing literature.

RESEARCH ON COMPUTERS AND PROBLEM SOLVING

One of the most intensely investigated topics in educational computing research is the speculated link between learning programming languages and the development of problem-solving skills. One of the most consistent findings from such research has been virtually no support of such a link.

This section begins with John Burton and Sue Magliaro’s critical review of many of the studies to date. Their premise is to determine whether such studies have been as soundly constructed as they should be. Some of the weaknesses they have identified are (a) inadequate length of treatment, (b) whether a language was, in fact, actually taught or learned, (c) lack of ideationally rigorous treatment, (d) use of insufficiently sensitive and/or inappropriate dependent measures, (e) lack of a theoretical, problem-solving basis for conducting the studies, and (f) whether the samples were developmentally capable of learning the language to the extent that problem-solving skills might be affected.

The next three research reports reflect a series of studies that investigate the purported link between problem-solving skills and programming languages. The first study, by W. Michael Reed and David Palumbo, centers on the effect of eight weeks of BASIC programming language instruction on problem-solving skills and computer anxiety. They used the guidelines posed by John Burton and Sue Magliaro when constructing their study, with the intent to construct a design that was more sound than the designs of some of the previously conducted research. They found significant problem-solving gains after eight weeks or approximately 81 hours of working with the language. Additionally, they found significant decreases in computer anxiety from pretest to posttest and a significant negative relationship between problem-solving skills and computer anxiety. They also determined relationships between “internal” factors such as positive relationships between prior experience with BASIC and (a) problem-solving skills, (b) debugging competence, and (c) programming performance and negative relationships between computer anxiety and (a) prior experience with BASIC, (b) debugging competence, and (c) programming performance.

The second study, by W. Michael Reed, David Palumbo, and Aletha Stolar, poses two questions: Does learning a programming language affect problem-solving skills? If so, which language – BASIC or Logo – affects problem-solving skills more? As in the first study, they found that both BASIC and Logo do positively affect problem-solving skills, after eight weeks and approximately 81 hours of working with both languages. They also found that one language does not significantly affect problem-solving skills more than the other.

The third study, by David Palumbo and W. Michael Reed, is an attempt to determine at what point problem-solving skills might be measurably affected. They have added another factor to the notion of rigor of treatment identified by John Burton and Sue Magliaro in defining the term, intensity of treatment: (a) length of treatment, (b) rigor of treatment sessions, and (c) proximity of treatment sessions. Their treatment focused on BASIC instruction and paralleled the treatment of their previous two studies: approximately 81 hours over an eight-week period. They used the two-stage, problem-solving concept as a design framework: (a) Time must first be spent to establish a linguistic base and to overcome potentially debilitating factors such as anxiety toward computers; and (b) once a linguistic base has been established, the problem solvers can begin to apply and fine-tune strategies. They decided that an appropriate measurement point to determine if problem-solving skills were affected would be the midpoint, or after 40 hours over a four-week period. At midpoint, problem-solving skills had not significantly improved, although computer anxiety had significantly decreased. Such a finding supports their notion that time must be spent simultaneously to build a linguistic base and to overcome potentially debilitating factors. From midpoint to posttest, problem-solving skills had significantly improved, whereas computer anxiety had not decreased significantly. Such findings support the belief that, once a linguistic base is established, students can then use the language for solving problems that might affect problem-solving skills. Although they did not isolate the exact moment problem-solving skills significantly changed, the authors’ finding does indicate that many of the treatments in prior studies that have not shown a significant effect on problem-solving skills have either been too brief, not sufficiently rigorous, or both.

The next study is one on adolescents’ chunking of computer programs by Susan Magliaro and John Burton. They compared ability of student programmers of differing programming language backgrounds – novice, intermediate, and advanced – to recall lines and chunks of lines of coherent and scrambled programs. They found that all three groups of programmers recalled more program lines and chunks from coherent programs than from scrambled programs. Also, intermediate programmers were more successful at recalling lines and chunks of coherent programs than both novice and advanced programmers. One reason for the superior performance by the intermediate programmers over advanced programmers was that intermediates were being instructed in BASIC and the advanced programmers were being instructed in Pascal; the test was on BASIC. The advanced programmers outperformed the novices and intermediates on the scrambled-program task.

Mike Orey and David Miller’s study centers on computer programs constructed to aid in the detection and analysis of student bugs in arithmetic and computer programming problems. They give reasons why none of these systems is adequate in and of itself and explain how a superior system could be constructed through the combination of several of the strategies that are explored. They support some of their points with an analysis of mathematical decisions made by a student research-participant.

Leah McCoy and Mike Orey investigated the effect of computer programming on problem-solving ability and the relationship between ability and problem-solving scores with computer programming achievement. The 120 middle school and high school students received daily instruction on the BASIC programming language for one semester. The results conclude that (a) the problem-solving skills of both middle school and high school students increased significantly from pretest to posttest; (b) programming achievement for both groups likewise increased; (c) problem-solving ability was the only significant correlate of programming achievement for the middle school students; (d) both problem-solving ability and verbal reasoning, however, were significant correlates of programming achievement; and, (e) the only significant predictor variable for programming achievement was general problem-solving ability.

This section ends with another study by Leah McCoy who examined the relationships of six variables: gender, age, developmental level, mathematics background, ability to use mathematical variables, and mathematical problem solving. The research participants were 21 students, ages ranging from 10 to 17 years, enrolled in a two-week computer camp who received instruction on the BASIC programming language. She found a significant relationship between mathematics (higher mathematics courses, mathematical problem-solving ability, and the ability to use mathematical variables) and computer programming achievement. However, the only significant predictor of programming achievement was the ability to use mathematical variables.

The research reported in this section sheds some new light on the link between learning programming languages and developing problem-solving skills. As John Burton and Sue Magliaro point out in their review, researchers must soundly construct their designs before findings can be acceptable guidelines. The subsequent studies indicate that the treatment (or instruction) must be rigorous, have depth, and be applied for a certain period of time. One reason for the promising results of the research in this section may be the developmental capability of the research participants; it may veiy well be that student programmers must be of a certain developmental age before they can learn a language to the extent that problem-solving skills might be affected. Although much work on the link between programming languages and problem-solving skills still needs to be conducted, these studies are significant contributions to this area.

John K. Burton
Susan Magliaro

Computer Programming and Generalized Problem-Solving Skills:In Search of Direction

JOHN K. BURTON is Associate Professor of Educational Psychology, Education Microcomputer Lab, Virginia Tech, Blacksburg, VA 24061.

SUE MAGLIARO is a doctoral candidate, Educational Psychology; Center for Reading Diagnosis, Evaluation, and Remediation; Virginia Tech; Blacksburg, VA 24061.

Over the course of the last three decades or so, a new field of study, cognitive science, has emerged and developed. As Wilkinson and Patterson (1983) have characterized it, cognitive science is “a field of study at the intersection of linguistics, artificial intelligence, and psychology” (p. 4). Although it may be argued that other disciplines are also at this intersection, particularly anthropology and philosophy, there can be little argument with the notion that the new area represents an interdisciplinary synergy. As such, cognitive science has always been grounded in the basic theories, paradigms, and evidentiary proofs of the elemental disciplines that contribute to it. Not surprisingly, cognitive science has also had a great impact on the conceptual symbols and explanatory patterns found in the parent disciplines, most particularly, on the information-processing framework of cognitive psychology.

Coinciding with the last decade of this development, commercially available microcomputers began to make their way into public school classrooms. Due to the relative low cost of the micros, schools began gathering any and all available resources to permit these devices to be included in the curricula. Since the beginning of the subsequent movement to the present, there has been much said and written about the use of microcomputer programming instruction as a device to teach higher level cognitive skills (see, e.g., Papert, 1980). In fact, for many of us now engaged in educational cognitive science, this possibility, as well as the potential to model such processes through expert systems, has become a major reason for our decision to forsake the more traditional thrusts of our respective disciplines. From powerful ideas (e.g., Papert, 1980) to fifthgeneration prospects (e.g., Feigenbaum & McCorduck, 1984), the claims vary widely but Feurzeig, Horwitz, and Nickerson (1981, as modified by Pea & Kurland, 1984b) offer a reasonable summary of proposed outcomes:

1. rigorous thinking, precise expression, recognized need to make assumptions explicit;

2. understanding of general concepts such as formal procedure, variables, function, and transformation;

3. greater facility with the art of “heuristics,” explicit approaches to problems useful for solving problems in any domain, such as planning, finding a related problem, solving the problem by decomposing it into parts, etc.;

4. the general idea that “debugging” of errors is a “constructive and plannable activity” applicable to any kind of problem solving;

5. the general idea that one can invent small procedures such as building blocks for gradually constructing solutions to large problems;

6. generally enhanced “self-consciousness and literacy about the process of solving problems”; and,

7. enhanced recognition that for domains beyond programming there is rarely a single “best” way to do something, but rather different ways that have comparative costs and benefits with respect to specific goals, (p. 8)

Obviously, the critical words and phrases are “the general idea,” “domains beyond programming,” and “in any domain” which imply transf...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• ABOUT THE EDITORS
• EDITORIAL
• PREFACE
• THEORETICAL AND PHILOSOPHICAL PERSPECTIVES TO PROBLEM SOLVING
• RESEARCH ON COMPUTERS AND PROBLEM SOLVING
• PROBLEM-SOLVING SOFTWARE