The argument that colleges and universities should reward faculty for their teaching as well as their research is falling on some responsive ears, as has become evident at the meetings of the American Association of Physics Teachers. Last fall's issue of this Newsletter devoted an eight-page sections to new approaches to teaching physics and their evaluation. This past summer, at its meeting in Lincoln, NE, the Association rewarded with its Robert A. Millikan Award a physicist who has enhanced both his research and his teaching by focusing his research on physics teaching: Edward F. "Joe" Redish of the University of Maryland. "Building a Science of Teaching Physics"
Robert Millikan's most enduring legacy to physics has been his insight into the quantization of electric charge, Redish noted, alluding to the strategy Millikan employed to yield the first measurement of the charge on the electron. But Redish also noted that another Millikan work, Laboratory Work in Physics, was more pertinent to the topic of his talk. In this 1903 work, according to Redish, Millikan criticized laboratory work as degenerating into manipulative exercise rather than enhancing concept learning. This insight, Redish pointed out, has been lost and rediscovered many times, while Millikan's insight into the quantization of electric charge has continued to endure.
Redish began his 6 August 1998 address on "Building a Science of Teaching Physics: Learning What Works and Why" by observing that we each make our own map of the world. When we do science, in turn, we exchange these maps and come to a consensus, like the convergence of a mathematical sequence. We need to do the same for science education as we have done for science, he said. He listed a series of general principles yielded thus far by cognitive research, and he focused on three of these.
The first one was context. Suppose one were confronted with the following symbols on four cards -- K, 7, A, and 2 -- and asked to determine which ones had a vowel on one side and an even number on the other. Which would we have to turn over? Then suppose that one is confronted with cards whose face-up sides contained "Coke," "16," "gin and tonic," and "50." Which cards would we have to turn over to determine who is violating alcoholic beverage laws? Most of us could answer "16" and "gin and tonic" more easily in the latter case than "A" and "2" in the former, we realized, though the two problems are in actuality equivalent. The reason, Redish pointed out, was that the context of the latter problem made it easier to deal with. Another example of the role of context presented by Redish was a splotchy picture from which a dog could be ascertained.
The next principle influencing learning was constructivism. Redish was quick to point out that the constructivism he was advocating was not antithetical to content. It was "guided" discovery rather than "free" discovery. He stated that such cognitive researchers as Lillian McDermott (whose work was reported on p. 5 of our Spring 1988 issue), Priscilla Laws (whose work was cited on p. 18 of our Fall 1997 issue), David Sokoloff, and Ron Thornton had all reported that guided discovery is indeed effective.
Finally, Redish presented evidence that most students learn better in social situations than alone (though, he acknowledged, some physicists often violate this). First, he cited his own experience in replicating Thornton and Sokoloff's results that a microcomputer-based laboratory approach was more effective than elaborate lecture demonstrations. In fact, at another session at the meeting, David Braunschweig of Madison, WI, reported the results of polling his student teachers: the only demonstrations they remembered were the ones that didn't work or the ones that blew up.
Redish then went on to cite the research of Richard Hake (Amer J. Phys., 66, 64-74 (1998)) on the percentage gains of students on the Force Concept Inventory (D. Hestenes, M. Wells, and G. Swackhammer, Phys. Teach., 30, 141-158 (1992)), which has become an accepted standard reflecting understanding of forces and their relationship to motion. The ratio of actual gain to maximum possible gain ((postest - pretest)/(100% - pretest)) was found to be approximately 20% with traditional instruction but ranged up to 40% when student recitations, tutorials, or group problem solving were added. In a separate report at the meeting, Jerry Loomer of Rapid City, SD, reported a 33% percentage gain with his high school students who used the Modeling method (described on p. 15 of our Fall 1997 issue); and Priscilla Laws' Workshop Physics has achieved percentage gains of 30-50%.
Mindful of the teaching vs. research controversy, Redish concluded his presentation by asking whether physics education is research physics. It requires serious rethinking of our understanding of physics, he responded, adding that physics is more than creating a new map -- it's about creating new understandings of the world. Physics education research requires collaboration of researchers and physics teachers, he maintained, and education departments have broader issues to deal with. To this end, Redish has become the founding editor of the Journal of Physics Education Research.
Home Fall 98 Full Screen
The TEACHERS CLEARINGHOUSE FOR SCIENCE AND SOCIETY EDUCATION