Viewpoint

Kenneth Heller on Viewing Teaching as a Physicist

By Kenneth Heller

Kenneth Heller

As physicists, we are concerned about science education. We know that to survive, a modern democratic country must have a population that understands and appreciates not only the fruits of science but also supports the process of science. The key to this understanding is not new technology, or better curriculum, as useful as they may be. The key is effective teaching in universities and colleges, in K-12 schools, in museums and nature centers, on TV and radio, and in personal contact as parents or colleagues.

The first step toward a culture that promotes and supports effective teaching is the recognition, by both teachers and their critics, that teaching is neither easy nor natural. As in other complex human endeavors, effectiveness requires using techniques and ideas that may be counter intuitive. This is as true for teaching as it is for physics. Recognizing that teaching is a complex set of skills and not a personal attribute would move us beyond recrimination by critics and defensiveness by teachers. We need to expunge the notion of a "good" or "bad" teacher and replace it with the notion of using more or less effective techniques of teaching. Changing one's tools does not require a gut-wrenching mea culpa that past teaching was "bad," but simply the recognition that more effective techniques are now available.

As physicists, we can view teaching dispassionately as the operation that transforms people from their initial state to a desired final state. Effective teaching is then the operation that maximizes the fraction of students making the transition. To find this operation, it is clear that someone must first carefully characterize the desired final state and also determine the ensemble of initial states that we are given. When the final state involves physics, mathematics, or science in general, we physicists have a great deal of input through such projects as the work on national education science standards by the National Academy of Sciences.

Characterizing the initial state of students is the province of education, cognitive psychology, and the emerging fields of specific subject-matter education, including physics education. If every learner were in a completely unique state, it might be impossible to implement a finite number of operations to substantially populate any desired final state. Luckily, broad categories have been found which categorize the initial state of a large fraction of people. This characterization of the initial state of the learner has improved dramatically over the last 20 years and a more precise characterization will come with the increase of research support and the development of new diagnostics tools.

How can we construct the relevant operation: the teaching? As in physics, random guessing is not an efficient technique, although it sometimes works. Theory is needed as a guide, and it is provided by the fields of education and cognitive psychology. A theory does not have to be correct to prove useful. After all, caloric theory was useful in guiding the early and very fruitful development of thermodynamics, and Newtonian theory is still useful in many venues. A useful theory can encompass basic principles or be purely phenomenological.

A phenomenological theory which has proven useful and seems to appeal to physics faculty - probably because it is reminiscent of graduate school - is called cognitive apprenticeship. It begins with the observation that apprenticeship has been an effective approach to teaching complex skills in a small group setting, and then extends that approach into the realm of more abstract learning for large numbers of people. Effective teaching based on cognitive apprenticeship must incorporate modeling (showing exactly how to do the desired skill), coaching (correcting individual work in real time), and fading (independent work). This provides the necessary framework to teach a course, and the framework, in turn, provides a structure to help teachers incorporate other empirical observations.

Determining whether or not a technique will lead to more effective teaching is difficult because learning is a complex process, and may well be non-linear. This may account for the observations that simple "controlled experiments" varying a single quantity typically show very small learning changes. When large learning changes are reported, they are usually difficult to reproduce unless all parts of the learning environment are reproduced. It may be that human learning, which depends on many parameters, has resonances. Although changing each parameter in turn gives a very small effect, the parameters can be tuned to give a large effect.

As physicists we can apply the same standards to teaching as to our field. Our research is based on theory and past measurements. We don't often repeat work without good reason. When a new technique arises that enables us to attack problems more efficiently, we embrace it. Changing method, technology, or analysis technique does not cast doubt on personal worth. We do not dwell on the past, nor do we demand that every new theory or experiment be a breakthrough. We take pride in our past accomplishments and marvel at all we accomplished using the tools at hand. We look forward, with some trepidation, to using the latest techniques and probing the latest theories. Powered by this attitude, the technology and techniques used in physics continuously improve. Can the same be said for teaching? As we look around, do we see the continuous incorporation of improved teaching techniques, or do we hear a clamor for identifying good and bad teachers?

Kenneth Heller is the Morse-Alumni Distinguished Teaching Professor in the School of Physics and Astronomy at the University of Minnesota, and vice-chair of the APS Forum on Education.