SCIENCE EXPERIMENTS AND THE RELATIONSHIP TO

LEARNING STYLES

CARL F. BERGER

Background

Work relating problem solving, learning, and teaching styles was first reported by David A. Kolb and his colleagues, Irwin Rubin and John McIntyre (1979). This work came from Kolb's earlier work (1976) on learning styles. Gathering data from thousands of adults in varying occupations, Kolb found that people could be categorized into four styles representing two dimensions (see Fig. 1).

Figure 1 Learning Style Type Grid

(Adapted from Kolb, 1976)

Across one dimension was a pattern in learning style of active to reflective; across a different dimension was a pattern from concrete to abstract. These dimensions sound familiar to those of us acquainted with Piagetian psychology. Concrete operations and formal or abstract operations are two of Piaget's developmental states. Further, we are all familiar with students who learn actively or reflectively or, as Kolb puts it, the "doers and the watchers."

Using the very simple and straight forward "Learning Style Inventory," which takes about 5 to 10 minutes to complete, Kolb (1976) found striking differences in the learning styles of adults in different occupations. He also found marked differences related to the basic personalities of the people tested. Striking as these differences may be, they fit remarkably well into reasonable theories of personality and occupation selection. For example, in a large sample of MIT seniors, business majors were highly active and concrete, whereas physics, chemistry, and mathematics majors were highly abstract and somewhat reflective. Political science and history majors on the other hand, were highly reflective and concrete. Engineering students were active and abstract. In studies identifying elementary and secondary teachers, colleagues of Kolb found that elementary teachers were highly concrete, and secondary school teachers were more abstract; both were equally active (see Fig. 2).

Figure 2. Student majors grouped on the Learning Style Grid

Kolb, Rubin, and McIntyre extended these learning style investigations into the field of problem solving. In their book, Organizational Psychology: An Experimental Approach (1979), the authors noted problem‑solving steps that they identified with locations on the learning style chart. Starting from the 3 o'clock position and proceeding clockwise, as shown in Fig. 3, they defined a series of two steps in each quadrant that could be thought of as steps of a problem‑solving process.

Figure 3 Problem Solving Cycle from Kolb, Rubin and McIntyre (1979)

These steps are:

1) Select a problem

2) Consider alternative solutions

3) Evaluate the consequence of solutions

4) Select a solution

5) Execute a solution

6) Chose a model or goal

7) Compare it with reality

8) Identify differences (problems).

Study

David Keller of the Akron Public Schools and I noted that these steps were similar to steps in a scientific method (neither of us believe there is THE scientific method). We generated another similar series of steps that might be more recognizable to science teachers. They correspond to the above steps and are listed below:

1) Select a problem related to a model

2) Develop hypotheses

3) Evaluate to find the best hypothesis

4) Select and experiment to test the hypothesis

5) Carry out the experiment

6) Draw conclusions

7) Compare conclusions with the model

8) Identify new problems and/or models

These steps are very similar to many scientific methods or problem solving models such as one reported by Dorothy Cox (1980). Armed with this new information that could relate to science teaching, David Keller and I wondered if science teachers had a particular problem‑solving style.

Method

Study 1

We tested 76 teachers who were working in a demonstration project for gifted students in the Akron Public Schools. Using Kolb's Learning Style Inventory, .we first checked to see if we could identify the same dimensions from our data as Kolb had from his. We were quite pleased and not a little bit surprised to find that our results compared very well with those of Kolb. We could independently produce the same dimensions of "concrete‑abstract and active reflective" that Kolb had found in his original data. We did this through a complex statistical process known as principal component analysis. Not only could we reproduce the dimensions, but also there appeared to be no single learning style or problem‑solving style that clearly dominated our population of teachers (see Fig. 4).

Figure 4 Scores of 76 teachers on a Learning Style Grid.

Thus, it is no small wonder that any single curriculum will not be appropriate for all teachers, and perhaps this is the source of the expression, "But it doesn't work for me. " It appears from Fig. 4 that these 76 teachers fit, and each may be comfortable with, varying parts of a scientific method. Some might emphasize more the hypothesis portion of science, others the experimentation segment, still others the application part, and some the model generation portion. For fun we could place Mildred Fetisque, Walter Smedly, Alma Curlew, and Oswald Minicule on the chart, not by their learning style, but by their teaching style. Kolb's research on teaching styles indicates that there is a close match between learning and teaching styles (1976).

Study 2

Are teaching and learning styles so closely related to our personalities and occupations that they are set and immutable? The research on this question is not clear, so David Keller and I tried an experiment. We attempted to find out if the problem‑solving style of teachers could be modified by the kind of activity the teachers were involved in during a workshop. We generated three simple experimental activities that involved different parts of a problem‑solving or scientific method process. Each of these experiments, we hoped, would require the use of a problem‑solving segment found more in one quadrant of the model than in another. We could then determine if the teachers modified their problem‑solving style by shifting toward a particular quadrant depending upon the kind of experiment. Our first experiment was designed to be both in the active and abstract quadrant. In this experiment, teachers had to support a book 5 cm above the table using only a file card. We thought that this activity would encourage teachers to reason abstractly and then actively build a structure with the file card to support the book. Experiment number two was designed to be in the reflective and more abstract quadrant. In this activity, teachers, working in groups, had to generate a food web, remove one or two links from that web and infer changes that would occur in the food web. Next, they were to compare the changes with a similar activity done with an energy web. Our third experiment was designed to be in the active and concrete quadrant. It involved the use of a simple drop reaction timer to measure reaction time; teachers were to modify their reaction time through a series of experiments. We hoped that the data from the Learning Style Inventory administered just after each activity and reflecting on the problem solving during the activity, would show shifts toward the quadrants emphasized in the particular experiment. The results of our work with these 76 teachers exceeded our expectations. As shown in Fig. 5, in each activity the means of the group fell in the appropriate quadrant.

Figure 5. Means of Three Experiments on a Learning Style Grid.

Further, there were strong significant differences among all three means, well beyond the traditional expected levels of statistical significance. Again, the teachers covered the entire spectrum of quadrants, and while the overall mean of each group did shift in the expected direction, strong individual differences were still apparent.

Conclusion

While teachers maintain a strong individual problem‑solving style, that style can be modified, depending upon the particular activity involved. We hope to do further research to determine if teaching styles also change in given activities. My own work in the late 60's and early 70's shows marked evidence of the relation between teaching behavior and teaching style perception, and that differing activities encourage wider ranges of teaching styles (Berger, 1977)

Resources

Berger, Carl F. "Investigation of Teacher Behavior: Interaction with New Curriculum Materials." In M. Piper, and K. Moore (Eds.), Attitudes Toward Science investigations. Columbus, Ohio: ERIC/SMEAC, The Ohio State University, 1977.

Cox, Dorothy A. Early Adolescent Use of Selected Problem‑Solving Skills Using Microcomputers. Unpublished doctoral dissertation, University of Michigan, 1980

Crowfoot, James A. Personal communication, 1979.

Gephart, W., D. Strother, and W. Duckett. "On Mixing and Matching of Teaching and Learning Styles." Practical Applications of Research. Bloomington, IN: Phi Delta Kappa, 3(2):1‑4, 1980.

Karplus, R., A. Lawson, W. Wollman, M. Appel, R. Bernoff, A. Howe, J. Rusch, and F. Sullivan. Science Teaching and the Development of Reasoning. Berkeley: Lawrence Hall of Science, University of California at Berkeley, 1977.

Kolb, David A. Learning Style Inventory Technical Manual. Boston: McBer and Co., 1976.

Kolb, D.A., I. Rubin, and J. McIntyre. "Learning and Problem Solving.” Chapter 2 in Kolb et.al. Organizational Psychology: An Experimental Approach (3rd ed). Englewood Cliffs, N.J.: Prentice Hall., 1979.Tyler,. R. Basic Principles in Curriculum and instruction. Chicago: University of Chicago Press, 1949.