Cognitive Styles and Implications for the Engineering Curriculum

Donald E. Beasley
Cecil O. Huey
Clemson University
John M. Wilkes
Keith McCormick
Worcester Polytechnic Institute

Abstract:

In a effort to assess the effects of cognitive style and the structure of an engineering curriculum on student learning, Gordon's Cognitive Style typology was used to characterize faculty and students in the Department of Mechanical Engineering at Clemson University. Gordon's typology provides measures which characterize those aspects of learning that involve creativity, problem solving, and problem identification. Originated to examine scientific creativity and research styles, it was subsequently found to be a predictor of scientists' choice of career fields and has been correlated with performance of engineering students on tasks of different types. The results of the study using this typology were used to assess the effect of design activities and selected courses on performance and retention of students of different cognitive styles.

The results of cognitive style typing for Clemson Students were examined for success in different curriculum areas and for attrition by cognitive style. The results suggest that the existing curriculum disadvantages one cognitive style more so than others. They also indicate that faculty/student differences are significant issues that deserve further study.

Introduction

``Everyone is different. That's why everyone should be treated the same..''....

Aphorisms such as the one above abound, but practical measures to insure fairness are difficult to develop and implement. In the case of the engineering curriculum, the matter goes beyond simple fairness to the very heart and soul of the profession. The vitality of engineering requires eyes that see different things in similar scenes and hearts that respond to different challenges or ones that respond in different ways to similar challenges. Accordingly, we need engineers of every stripe. We should easily satisfy this demand since we are certainly supplied with students of every stripe. Yet, we likely do things that disproportionately disadvantage students who possess particular cognitive and perceptual characteristics. Further, we often embrace these practices and tout them to others as measures that insure ``quality'' or enforce standards, assure rigor or something equally meritorious when they actually filter the student population and yield a more homogenized student cohort. A couple of examples will help clarify the issue.

Back to the Basics

Engineering curricula often immerse entering students in coursework that is heavily weighted to the sciences and that emphasize activities that reward certain attributes of students. In terminology to be introduced shortly, such activities accommodate students who can be classified as problem solvers. Later in the curriculum, activities shift to the advantage of other groups. This sequential aspect of the curriculum advantages or disadvantages different segments of the student cohort sequentially. Those students who are disadvantaged early are disproportionately discouraged and drift away from engineering in greater numbers before being given a chance to play to their strengths. It would be far better to have parallel experiences that stretch some students while confirming others.

The authors feel that a common refrain heard in engineering education circles is a symptom of this problem. Instructors in senior design courses often complain about the students' inability to ``put it all together.'' They note with dismay the ``good'' students who flop when they get to the ``real engineering.'' Perhaps the apparent magnitude of the problem stems from the absence of students who would excel at integrative thinking, but who can now be found in the business school because they were weaker at pre-structured problem solving and disaggregative analysis

Many curricula address the problem by including rigorous filter courses at the freshman or sophomore year to prevent students who are unsuited to engineering from going further or discouraging students who ``don't want to work.'' The courses are effective or they would not be so common. However, they might do more harm than good if they result in an unnaturally homogenized student cohort coupled with a curriculum that is tuned to the dominant characteristics of the group who make it through the filter. A curriculum so implemented is unlikely to produce the full range of practitioners required to sustain verve and vitality in the profession. It will, however, produce hordes of failed students who forever think the profession is typified by such courses.

Informed but Misguided Curriculum Reform

Many have called for reform in the curriculum in response to the issues cited earlier and to others. These calls often include a theme of sorts that can be expressed variably as the following.

These are worthy goals and they should be pursued. Yet, they are often advocated in fashion that seems only to substitute a new ideal for an old one; a new student model to be cloned endlessly, in place of the old one. In many ways, this suggests developing a refined, sensitive, new-age engineer to replace the conventional version. The profession demands diversity, but not necessarily a full measure in each individual. It demands individuals who differ in their strengths and weaknesses to produce balanced and effective teams. Accordingly, we must expand our students' awareness and sensitivities and their mutual respect for differences but without compromising individual strengths and talents.

In All Things, Moderation

A curriculum should help students grow comprehensively, including in the areas that are existing strengths. To that end our curricula should challenge students in their weak areas and help them to blossom in their strong ones. It should do these things simultaneously, not sequentially, to avoid the problems cited earlier. It is to this ultimate end that the subject study was undertaken at Clemson University. Before discussing the study and its implications a brief overview of Gordon's Cognitive Style Typology is in order.

Gordon's Cognitive Style Typology

Gordon's Cognitive Style Typology is designed to distinguish between individuals having a propensity for problem identification and formulation and those whose strength is solving problems, especially those that are well-defined. It uses two measures, remote association and differentiation, to establish four cognitive styles. The first measure reflects the ability to produce novel combinations of associative elements. It assumes that the more mutually remote the elements of the combination, the more creative the associative process. Remote association is mental quality that enables some to rapidly conceive of numerous novel possibilities and to almost instantly see that some are superior to others and that one is a solution to the problem. Individuals scoring high on this measure often display impatience with others who are slower in apprehending a solution. There is no evidence to suggest that remote association is the kind of ability that can be learned, and this fact alone has implications for the curriculum that will be addressed later.

The second measure is termed ``differentiation'' and is related to the ability to discern subtle differences and make fine distinctions on the basis of relatively nebulous criteria, including subjective standards. Differentiators are not bound by preconceived notions and are attuned to subtle inconsistencies and anomalies in a situation, regardless of expectations. Indeed, they can get hung up on them. They also accept unanticipated outcomes readily.

The two measures are very weakly correlated with each other so it is possible to use a ``high'' and ``low'' scale for each and to define four distinct cognitive styles as indicated in Table 1.

There is likely to be differences between the four types involving the following:

It is important to note that in engineering, the extreme opposites as defined above, i.e.,. High-Low and Low-High types, can each excel but they do it in substantially different settings. For example, the High-Low type dominates in solving unstructured, unconstrained problems. The Low-High type are quick and efficient in handling defined, structured tasks. Engineering practice abounds with problems of both types.

The Clemson Study

This paper focuses on one of the principal goals of the Clemson study which was to test the hypothesis that there is a cognitive bias in the early part of the program in favor of problem solvers. It was feared that problem finders who would later have a great deal to contribute as masters of ill structured problems were being lost before they got to show their stuff. The cognitive styles of Clemson mechanical engineering sophomores and seniors were determined. In this preliminary investigation, a comparison between a course believed to be highly structured, freshman chemistry, and a design course was undertaken. The design courses are where skills suited for ill-structured problems seemed most likely to surface. The principle observations are summarized using Figures 1-6.

Figure 1 shows the distribution of cognitive style for both students and the Mechanical Engineering Faculty. The numbers shown are totals including both seniors and sophomores. There are disproportionately more problem finders among the students and more of their cognitive opposites, the problem solvers, among the faculty. This is especially true in certain faculty specialties. However this particular aspect of the relevant cognitive distributions will not be addressed here.

In courses examined thus far, the cognitive types that did best in courses having differing demands varied, as was expected. Sometimes there was a clear cognitive bias. For instance, as indicated by Figure 4, the freshman chemistry course is clearly a ``problem solver'' task environment .

The results depicted in Figure 5 were a bit surprising at first but have proved to be quite interesting. We expected the HL group to excel since this course is structured around open-ended projects. However, no significant portion of the grade is based on the problem formulation stage of the project. The LH group, most nearly matched by the faculty, achieved a high percentage of A's and B's. The vast majority of C's earned in the course were awarded to the problem finders, HL's. Grades seem to reflect how soon the students pick up on conventional wisdom and followed it. It is clear that being high in remote association is an advantage in this course, since both HH's and LH's earned only grades of A and B.

Figure 6 shows attrition by type. The evidence that the early part of the program favors the problems solvers among the students is pretty clear. As one compares the distributions of the students from the different cohorts the mix changes dramatically as the attribution seems to be concentrated on the two low remote associating types, especially the problem finders. This initially dominant type drops steadily as a portion of the pool as one compares freshman and sophomores to the upper classmen which yields a more balanced cohort. Which courses are the main contributors to this pattern of attrition is under investigation.

Summary

A pilot study into the influence of cognitive styles on learning and curriculum structure has been undertaken, and some preliminary results from a small student sample have been examined. The Clemson students in mechanical engineering who are of the same cognitive types as the majority of the faculty are significantly advantaged in the early part of the curriculum. Because the profession clearly requires a variety of cognitive abilities, further investigation into means to enhance the success of all cognitive types is warranted,

References

  1. anon., ``Diversity and Learning Theory,'' The National Teaching &Learning Forum, Vol. 1, No. 5, 1992.

  2. Felder, Richard M. and Silverman, Linda K., ``Learning and Teaching Styles in Engineering Education,'' Engineering Education, April 1988.

  3. McCaulley, Mary H., Godleski, E. S., Yokomoto, Charles F., Harrisberger, Lee, and Sloan, E. Dendy, ``Applications of Psychological Type in Engineering Education,'' Engineering Education, February 1983.

  4. McCaulley, Mary H., Macdaid, G. P., ``Myers-Briggs Type Indicator and Retention in Engineering,'' Int. J. Appl. Engng. Ed. Vol 3, No 2, pp 99-109 1987.

  5. Miller, Judith E., Wilkes, John, Cheetham, Ronald D., and Goodwin, Leonard, ``Tradeoffs in Student Satisfaction: Is the ``Perfect'' Course an Illusion?,'' Journal on Excellence in College Teaching, 4, 27-47, 1993.

  6. Beasley, D. E., Bishop, E.H., Huey, C.O., ``A Process for Curriculum Renewal and Innovation,'' Annual Conference Proceedings, ASEE, pp 1407-1418, 1994.

  7. Wilkes, John, ``An Overview of Gordon's Cognitive Style Typology,'' Paper distributed with presentation at Southeastern Regional Conference of Association for Psychological Type, September, 1992.

  8. Shadish, W.R. and Fuller, Steve, ed., ``Characterizing Niches and Strata in Science by Tracing Differences in Cognitive Styles Distribution,'' Chapter 10, Social Psychology of Science, Guilford Press, NY, 1994.





mort@etp.com
Mon Oct 9 13:50:36 PDT 1995