New Models of Physics Instruction Based on Physics Education Research: Part 1


Edward F. Redish

Department of Physics and Astronomy
University of Maryland
College Park MD 20742-4111

This paper will appeas in Proceedings of the Deustchen Physikalischen Gesellschaft Jena Conference (1996).

It is displayed in two parts. Go to part 2.(References are at the end of part 2.)


During the past fifteen years, physics education research has taught us many surprising things about the difficulties introductory university students have in learning physics. At the same time, the ongoing revolution in information technology has led to new tools for creating innovative educational environments. In response to these two developments, a wide variety of new models of physics instruction are beginning to appear. We review some of the findings of physics education research, putting them into the context of a theory of thinking and learning. Some of the most promising instructional models currently being developed in the US are discussed.

I. Introduction

Over the past few decades, changes have taken place that require a change in how we teach introductory university physics. First, a larger fraction of the population is graduating high school and going on to universities than in previous times. Many of these students are concerned about finding jobs in an increasingly technological workplace environment, and so are enrolling in technical curricula that require physics. As a result, a larger fraction of our students today appear inadequately prepared to take university physics than was the case in the past.

Second, for those of us in publicly supported institutions, the governments and the populace that employ us appear more likely to hold the educational system (and therefore its teachers and administrators) directly responsible for the students' learning -- or lack of it -- then they were in the past. In previous times, the individual student was often held personally responsible for their learning and less attention was paid to the effectiveness of teaching.

As a result, the task of the physics teacher today is to figure out how to help a much larger fraction of the population understand how the world works, how to think logically, and how to evaluate science. This is doubly important in democratic countries such as the USA and Germany in which a large fraction of the adult population is involved in selecting its leaders -- those leaders who will make decisions not only on the support of basic science, but on many items that depend intimately on technological information. Having a populace which cannot be fooled by the misuse of science and by scientific charlatanism would be of considerable value.

We may ask ourselves whether we are perhaps in the best of all possible worlds and are already achieving these goals. Does traditional physics teaching "work" in the introductory physics classroom? Unfortunately, detailed examinations by many physics education researchers have shown that it does not work well for a large fraction of our students. Many of our students fail to gain the skills that permit them go on to success in advanced science courses.

This can have strong negative consequences. When many students fail, faculties may be pressured to pass more students, with the result being a lowering of standards. This of course is ineffective in the long run. The lowering of standards simply postpones the time at which the unprepared student will be unable to meet the requirement either of more advanced courses or of a job in the technological workplace.

The nature of the difficulty, as illuminated by physics education research, appears to be a kind of "impedance mismatch". The professor sends out information and sees it reflected back in a similar or identical form, but little information has actually gotten through to the other side.

Fig. 1: The fact that something "comes back as we sent it out" does not mean that much has "gotten through to the student", especially if students possess a large inertia!

If we are to understand these difficulties we must treat the problem scientifically by observing carefully the phenomena we want to understand. From educators and cognitive psychologists we learn two important lessons.

A Model of Thinking and Learning

Over the past three decades, cognitive psychologists and educators have begun to build a model of learning that seems to provide a framework for this analysis. This framework is a relatively recent growth based on the ideas and experimental methods learned from psychologists Jean Piaget, Lev Vygotsky, and the gestalt school of psychology, among others. I refer to this as the constructivist model of cognition and extract four principles that help us understand the kinds of difficulties that take place in a physics class.1

  1. Constructivism
  2. Context
  3. Change
  4. Variability

Principle 1: The Constructivist Principle

Students "construct" their ideas and observations -- pulling together what they see and hear into a "mental model".

This construction is an active but in most cases, automatic and unconscious process. Think of language learning by a small child as a prototypical example. Children create their own grammars from what they hear. (The fact that they don't always create the same rules as their parents have is one of the facts that causes languages to evolve.) A nice example of the way the brain constructs is seen in Fig. 2.

Fig. 2: A picture of an animal. "Some assembly required."

The picture will be immediately obvious to some, more difficult for others. (See the footnote if you have looked at the picture for a while and can't make out an animal.2) The image is in your mind, created by your brain pulling together the loosely related spots and "constructing" the image. Once you have seen it, it will be hard to remember what it looked like to you when you couldn't see it.

For more complex situations, the brain constructs a pattern or "mental model" of the situation in order to understand and analyze it. When I say "mental model", you should not construct a picture of something machine-like. That isn't the nature of the phenomenon. Some properties of mental models (MMs) can be summarized in the following statements: MMs consist of propositions, images, rules of procedure, and statements as to the context in which they are to be used.

The last has a rather direct implication for a physics class. Students may well accept an idea within the bounds of a physics class or carefully constructed experiment, but not consider that it has any implications for real world events or for their personal experience. We elaborate this in our second principle.

Principle 2: The Context Principle

It is reasonably easy to learn something that matches or extends an existing mental model.

This has two corollaries.

This is illustrated with another visual image in the picture at the right taken from a cleverly designed greeting card by artist Jay Palevsky.3

Fig. 3: What is this? Are you sure?

When the two halves of the card are pulled apart, they reveal that they are part of a different picture than the one you originally perceived. The change of context changes the way our minds interpret the visual image. (To see the opened card, just click on the figure above. Click on it again to return it to its original form) The important point to realize is that the "context" in which a student interprets any information in a physics course includes not just the classroom or laboratory environment, but the context of all their previous learning and experience. The most important context is the state of the student's mind at the instant the information is presented.

Principle 3: The Change Principle

It is very difficult to change an established mental model substantially. This is nicely illustrated by the famous picture shown in Figure 4.

Fig. 4: A drawing of an old woman. Or is it a young one?

This ambiguous figure can be seen either as a young woman or an old woman.4 The interesting part of the story of this figure is that it was used in a psychology experiment at Columbia University. Two unambiguous figures were prepared -- one of the old woman and one of the young woman. Half of the class were given one of the figures, half the other. Then the figures were collected and the entire class shown the ambiguous figure. In subsequent class discussion, those who had seen one of the figures had great difficulty in seeing the second possibility, even when it was described in detail by someone in the other half of the class.

Students often have a similar difficulty in a physics class. If they have already misinterpreted previously given knowledge or previous experience, it may be very difficult for them to put the correct interpretation on what a teacher says. The fact that this problem is widespread and occurs in many areas of an introductory physics class has been well documented in the physics education literature.

Two of the consistent observations of this research are:

This is even the case if students are "warned" about common misconceptions

This means that environments in which students are encouraged to elicit and confront the mental models they have are more effective in changing those models than environments in which the "correct" information is simply presented.

Principle 4: The "Distribution Function" Principle

Since individuals construct their own mental states based on their own experiences and personal makeup, different students have different learning styles and responses.

This is by now very well documented by a large number of psychological studies.5 Some students respond better to visual information, others to symbology. Many students seem to learn better using "hands-on" activities as compared to listening to abstract reasoning. There are many variables, and a good knowledge of physics requires calling on a wide variety of different media and manners of coding and conveying information.

Two corollaries result.

Implications of the Cognitive Principles for Physics Teaching

I have taken my examples from the human visual response since they are dramatic, many people easily see the illusions, and they clearly illustrate the principles. However, the principles are general and have a powerful impact in the physics class as well. With the model in the back of our minds we must raise some concerns that are often neglected in a traditional physics class.

Learning about the difficulties: Physics education research

If we are to really understand what is happening when we try to teach physics we have to study it as a scientific problem. Human beings have a strong tendency towards "wishful thinking" -- to seeing what it is they want to see. This does not imply that we are duplicitous, just prone to "hopeful misinterpretation". It is this tendency that the scientific method, as carried out by an active and skeptical research community, is specifically designed to combat.

If we are to find out what is really going on in our classes, we will have to do research. In the context of physics education, this means the direct observation and interpretation of student behavior, especially detailed interviews. Our standard examinations, designed as they are for evaluation of student success rather than for understanding student difficulties, do not usually suffice. A research evaluation may be carried out through observational (as opposed to instructional) interviews, and occasionally, by means of other carefully developed testing instruments.

One approach to the linking of research to the development of instructional materials is the cyclic process practiced by Lillian McDermott and her Physics Education Group at the University of Washington.6 In this process, research on student understanding illuminates the difficulties in current instruction. The results of the research can be used to design new curricula and teaching approaches, which lead to modified instruction. This process cycles in a helix of continuous educational improvement.

Of course, to understand what one sees in a research situation one must have a model or theory of the system under investigation in order to know what to look for and to make sense of what one sees. On the other side, the experimental observations may cause us to refine or modify our theoretical model. I represent this process schematically as "McDermott's Wheel" in Fig. 5, with the model of cognition and learning serving as the axle about which the wheel rotates.

Fig. 5: "McDermott's Wheel" - illustrating the role of research in curriculum reform.


1 Edward F. Redish, "Implications of cognitive studies for teaching physics," Am. J. Phys. 62 (1994) 796-803.

2 The picture is of a Dalmatian dog with head down and to the left, drinking from a puddle in the road, seen through the shadows under a leafy tree.

3 This is taken from a greeting card created by the artist Jay Palefsky. When the card is opened, the two halves of the picture (vertically split) slide open to reveal that they are part of the larger picture as shown on the next page. See his book Metamorphimals (Kutzkies Artworks, Garrison NY, 1995) for a number of examples. Copyright Jay Palefsky. Reproduced with permission.

4 Hint: The young woman's chin becomes the old woman's nose. The old woman's mouth is the young woman's necklace. (This picture is given courtesy of the Harvard-Radcliffe Cognitive Science Society, Clay Budin, Pres. It may be used at will and distributed, provided this message remains.)

5 Howard Gardner, Frames of Mind: The Theory of Multiple Intelligences (Basic Books, NY, 1985); Noel Entwistle, Styles of integrated learning and teaching: an integrated outline of educational psychology for students, teachers, and lecturers (John Wiley, NY, 1981).

6 Lillian C. McDermott, "Millikan Lecture 1990: What we teach and what is learned Closing the gap," Am. J. Phys. 59 (1991) 301-315.

Go to part 2.
This page prepared 9. June 1996 by
Edward F. Redish
Department of Physics
University of Maryland
College Park, MD 20742
Phone: (301) 405-6120