2017 Oersted Medal Presentation: The changing face of physics and the students who take physics Free

It is a great honor to receive the Hans Christian Oersted Medal, and I thank the AAPT, the awards committee, and whoever nominated me for this award. I also would like to thank my many friends at AAPT whom I have gotten to know so well particularly while I was Editor of the American Journal of Physics. AAPT has been a significant and rewarding part of my life. In addition, there is no way I would have received this award without the stimulating collaborations with many of you, particularly Harvey Gould and Wolfgang Christian. I also want to acknowledge my partner Hillary Rettig, who is here with us today.

Finally, I acknowledge my home institution, Kalamazoo College, which has been a wonderful place to work. I have been well supported by both my faculty colleagues and the administration, and have had some opportunities to work in the administration, which has been enlightening. Kalamazoo College is a remarkable place, if for no other reason than that it has now produced two recipients of the Oersted Medal. I am not the first. The first was John Wesley Hornbeck in 1951, who is shown in Fig. 1.

I wish to speak about two topics, how physics has changed and how our students are always changing. For both topics I want to challenge us to think creatively about how these changes should affect our teaching of physics.

How has the culture of physics changed in the past 20 years or so? Research groups are bigger. Very few papers are single authored. Because many of the simple problems have already been done, it takes a much larger number of people to tackle most current problems of interest. Physics is more international. Globalization has affected all aspects of society including science. Many projects are truly international.

Physics is becoming even more interdisciplinary. Biological physics is one of the hottest fields of physics right now, but astrophysics, geophysics, chemical physics, among others are also very big fields. And fields such as econophysics and complex systems more generally are drawing upon physics.

However, the biggest change is due to the omnipresence of computers. Computers have been used for a long time to do numerical work such as root finding and inverting a matrix. This use of the computer did not fundamentally change physics. As computers became more common and powerful, they were increasingly used to collect data and control experiments. Again when used on small table-top type experiments, this use mainly freed up graduate student time from some tedious chores. Or, as vividly shown in the recent movie Hidden Figures, digital computers replaced human computers, and many people, particularly women, who used to do calculations by hand, then became computer programmers. Then in the last part of the last century, computer simulations became much more common, and computational physics as the third approach to doing physics took hold. This area is where I did most of my work. I came in near the beginning of this approach to doing science and have been using simulations for nearly 40 years. I have written research papers based on simulations, co-authored textbooks on it, co-edited columns on it, and I believe it is an important way of thinking about scientific problems.

We are now in the middle of big changes in how science is done that builds on the changes I have just mentioned. Traditionally, a typical experiment was designed so that all variables except one or two were held fixed, and then we made a number of measurements on specific quantities as we changed one or two variables. For example, we might look at the scattering intensity of X-rays scattered by a sample as a function of energy and angle. Or we might measure the properties of a material as a function of temperature and pressure. Today we are increasingly interested in more complex phenomena where we collect lots and lots of data—sometimes referred to as big data—and then look for the patterns. We need computers to collect, store, and categorize the huge amount of data, and then we need them again to do the analytics to find the patterns in that data. For example, how did LIGO detect a gravity wave? Their computers collected data from the interferometer signal almost continuously, and then checked the signals against a bank of known signals and only considered those that are possible candidates for a gravitational wave. Figure 2 shows what they found. If we ever detect a dark matter particle, it will be done the same way. Computers will take in constant streams of data and look for signals that are unlike any of the known reactions from already known particles. To give you an idea of the scale, CERN generates of the order of 100 terabytes of data per day whereas the human DNA is about 1.5 gigabytes of data.

However, looking for patterns in the data and even discovering a new pattern is not enough to generate understanding or any sense of causality. We also need to generate models and then implement those models usually with computer simulations. For example, with the LIGO detection we were able to say that the signal came from the collision of two black holes because we were able to simulate such an event on a computer using the equations of general relativity. Without a model our understanding boils down to merely an advanced form of categorization, with little in the way of insight.

What does all of this mean for physics education? At present it has meant very little in our undergraduate and graduate courses. Instead students are exposed to this way of thinking only if they are fortunate enough to engage in research projects. Because our courses provide little preparation for such research, my experience is that many students end up approaching such computational research in a mechanical and sometimes naive way. They are told how to run a piece of software, what to look for in the numbers, and sometimes they are told how to organize the data, although usually much of this is done by the software. Sometimes they are unsure what is even being measured. Sometimes this uncertainty is because so much preprocessing is done that what is actually measured is totally obscure to the student.

Here are some ideas of what is being done in the curriculum and what could be done. Toy problems are a staple of undergraduate and even graduate physics. We give problems that are very unrealistic all the time, usually ignoring forces such as air friction because the analytical computations are beyond the mathematical abilities of many of our students. Instead, we can teach our students some basic ideas of computer modeling. The intro text, Matter and Interactions by Chabay and Sherwood,³ for example, does this, and the PICUP group⁴ is pushing to incorporate more computational physics into the undergraduate curriculum. A number of us have also written textbooks on computational physics⁵ and there are a number of courses being taught throughout the country. An example of what I have in mind is illustrated in Fig. 3. We also should be teaching students how to search through data for patterns and then use modeling to try to understand those patterns. For example, we could generate lots of data of particle positions from molecular dynamics using Newton's laws. This data could be given to the student, and then the student could be asked various questions about the data. Can they learn, for example, what the inter-particle potential is? If the data contained some kind of defect, could they say something about the location of the defect and its properties? In an astronomical context could they determine whether a planet is missing from the data on planetary trajectories and roughly where the planet might be? We would need to provide some tools to the student to help them answer these questions, but I believe it is doable and it might be much more interesting than what they are doing now. Moreover, if structured appropriately this approach could lead to deeper understanding of the physics.

We also need to think about how research is done and model some of that activity in our courses. The days of individuals sitting in their offices or labs and working alone are pretty much over. Big groups and constant interaction with peers are the norm. How do we encourage that? How do we access student ability to do that? Most schools ask students to work on group projects or work in labs in pairs or larger groups. But the nature of the interactions are not modeled by the faculty for students to see, and are not taught or accessed in any rigorous way as far as I know. Shouldn't they be? One idea would be to have upper division courses where younger students work on projects run by more advanced students. The advanced students would need to explain the physics and teach the techniques to the younger students. Thus, they would gain practice in explaining physics, and the younger students would learn how to do things at an early stage and then repeat it later when they are in the advanced courses. I have not heard of anything like this being done in a course setting, and I think it is worth trying.

Now I want to talk about the students themselves.

The present set of colleges and universities in the United States evolved from European universities that originated over a thousand years ago. Up until recently, they mainly educated relatively privileged white males. In the U.S. that population is now the minority. The majority of college students are women, and a large percentage of students are not white. Many are international students, and for many English is not their first language. Thus, it is not a surprise that many of our students find that the institutions they find themselves in may not be the most conducive to learning. The large lecture classes that are common in introductory science classes are probably not too different from typical courses given hundreds of years ago. There is much evidence from physics education research and other work that this approach is not very effective for most students. I would guess that there are other institutional structures that exist today that are barriers to learning for many groups of students. In addition, the expectations we have for our students are usually based on our own experiences, even though most of us became professors or educators precisely because we were successful in the particular institutional structure that already exists. We are a very small minority, and thus what worked for us may not work for most others.

There are a number of insights that I have garnered over the years that might be of general interest. Some are relevant to any college course and some just to physics. First, we must realize that taking courses is a relatively small part of a student's life. This is particularly true at a place like Kalamazoo College, where nearly everybody is active in some extra-curricular activity such as a sport, musical activity, play, social activism of some sort, or a job to make money to afford to stay in college. Second, most students, particularly those from advantaged backgrounds, have had everything done for them by their parents; they are not very independent.

Student behavior frequently puzzles me. Many appointments I make with students are not kept. In many cases, the students do not even acknowledge missing them. Almost everything students do in a class must have a consequence. They have a difficult time seeing the connection between the reading and exercises that we ask them to do, and the assessment they get on examinations. Thus, it is almost useless to give ungraded homework to students. There are all kinds of strategies that people have come up with to get students to read the text before class, because just asking students to do so does not work.

And although many students have strong academic interests in a particular area, learning for its own sake is not a widely held value. Thus, in our introductory physics class, our biggest difficulty is that students see no reason to learn what we are helping them to learn. Even when we show clear applications to medicine or other fields that they think interests them, they show little interest. They can be wowed by various demonstrations, but rarely are they curious enough to want to know why these demonstrations work. On the other hand, students do things on their cell phones that are sometimes awe inspiring to me, and they are certainly far more adept at using most digital devices than I am, even though I use computers every single day. However, are these skills helping or hurting their ability to learn physics?

These generalizations might sound like student bashing. It is not. It is culture bashing. We live in an anti-intellectual culture that pervades almost every aspect of our lives. The media and entertainment are so pervasive and so enticing. As we saw in the last election, facts meant very little, and it was impossible to have a reasoned discussion about anything of importance. There also is some explicit anti-science rhetoric that we must contend with.

As physics teachers we have to work with our colleagues in other disciplines to work on all these issues. We need to work against the anti-intellectualism that pervades our society. We have to make our teaching methods inclusive for all students who enter our classrooms. That is a big task, and it would take more that a few minutes to offer anything of general significance. Instead, I want to focus on physics itself.

At Kalamazoo College, we use a studio physics or workshop physics format in our introductory physics sequence. This course has been team-taught by Liz McDowell, Arthur Cole, Tom Askew, myself, and a variety of adjunct professors. Our only introductory sequence requires calculus, but really relies mostly on algebra and trigonometry. Students are quizzed during every class on the previous class objective. They receive a grade of mastery or developing; there is no partial credit. If they do not master the objective, they have an opportunity to show mastery through an oral reassessment or on a final exam. We do pre- and post-testing using the Force Concept Inventory, a math diagnostic test, the University of Maryland survey of student attitudes toward physics, the Lawson test of logical reasoning, and in the second term pre- and post-testing using the Conceptual Survey of Electricity and Magnetism. In each class, we usually have two faculty members and several undergraduate teaching assistants circulating around the room talking to students as they work on physics problems, mini-labs, and simulations. In addition, by our oral reassessments we typically have a detailed interview of nearly every student. Conceptual gains by our students compare well with those using the best PER practices, and we are happy to report that learning gains are strong independent of the initial conceptual understanding of the students. On course surveys students state that they believe they are learning even though they express displeasure at the way the course is taught. The result of all this is that we have come to learn how our students think, what skills they bring to the class, and how their attitudes and skills are correlated with their performance on the various assessment tools.

With all this information on students, I still wonder why is physics so difficult? To me chemistry and biology are much more difficult. There is so much terminology to remember. In physics there is very little to remember. You just need to figure things out. Why is that so difficult? Are the concepts really that hard? PER has given us many tools to help students improve conceptual understanding, yet physics is still not that popular or easy. I am not sure I know the answer, but I think I know some of the ingredients to the answer. Every discipline has terminology to learn, and once we understand the terminology we are a long way toward understanding. In chemistry, for example, a big focus is on chemical reactions. There is terminology about different kinds of molecules and reactions, various rules for how these reactions work, and a small amount of math to obtain some quantitative results. In physics we have much less terminology, but understanding requires the ability to move between at least four languages or ways of thinking, which are used to build models of physical phenomena. The first language describes the phenomena itself. “A block slides down an inclined plane.” Here not only must the student understand what a block is and what an inclined plane is, but they must also know to think about whether or not there is sliding friction. Words such as “smooth” may be used to tip off the reader. The second language is that of physics, which uses words such as “force” and “energy.” The problem with this language is that these words are the same as the words in our native language such as English, but have very specific technical meanings. The third language is mathematics. At the introductory level, this language is mostly algebra. Most of our students have a reading knowledge of algebra, but are not fluent at it. They are very slow at doing algebraic manipulations. They use up so much of their cognitive resources doing the algebra that there is little left to think about the physics. There is another language issue going on here. I am not sure whether to call this issue another language or not, but my observation is that symbols in equations seem to have no meaning to students. We use subscripts and superscripts and arguments in parentheses to indicate functional dependencies and to distinguish one symbol from another. All of this use seems like noise to many students, as illustrated in Fig. 4. As a result, mathematics is not only difficult, but is tedious and prone to error, and obvious errors are missed because the meaning is simply not there. There are other languages such as plots and figures. Because most students have rarely, if ever, used graph paper, they cannot draw their own sketches, and plots have significantly less meaning to them than to us. When drawing figures they usually miss key points, for example, making certain objects much bigger or smaller than they should be. Or drawing curves that should be straight or angles that look about the same when they are quite different.

The point is that we require the ability to move from one language to another almost seamlessly even in introductory courses. We require constant translations. Not only do we require this, but it is not obvious to the student that we are asking them to do it. It is not obvious to me how to fix this when some of the languages, mathematics in particular, are so poorly developed in most students. Even many physics majors have poor math skills.

Maybe we need to rethink what we are doing. When I was a graduate student, special functions seemed esoteric, and I never really became very fluent in using them. I thought my professors were fluent and we were just ignorant. They seemed much better at mathematics than I. Nowadays this does not seem like much of a handicap, because so much research is done using computational tools. The focus on finding analytical expressions almost seems old-fashioned. Are we just moving in that direction even more so? Rather than teaching students to do algebraic manipulations, should we be teaching them how to visualize data on a screen? Rather than worrying about what the analytical form of a plot is, should they learn how to intuitively know what those forms mean just by looking at it, and perhaps how certain kinds of processes lead to certain kinds of visual effects? Was most of the mathematics we learned needed only because we did not have computers to grind out more accurate calculations? Instead of requiring so much math, should we be teaching students more about algorithms?

We have powerful digital devices for doing math from calculators to computers. My concern is not that students use these, but they seem to be under the illusion that by doing so they understand what the math is all about. Thus, they think they understand arithmetic when the calculator does it for them, or even worse, that they do not need to understand arithmetic because the calculator does it for them. We need to figure out what it is about arithmetic that they need to know. The same goes for algebra, geometry, and other parts of mathematics. Clearly, some level of understanding has been lost through the use of digital devices. The questions include what has been lost? Is what has been lost essential for understanding physics? What can we do about it?

Obviously, we are not going to change things overnight. We live in a world where globalization, urbanization, environmental destruction, the growth of wealth inequality, and the elimination of many jobs and the transformation of many others lead to great uncertainty. The best we can do is provide the critical thinking tools to help our students navigate such a world. I believe physics has an important role to play. It is a quantitative way of looking at the world. It respects data. It is open to bringing ideas in from one area to look at an apparently completely different area. It generates its own tools (experimental, mathematical, and computational) when needed. Although all areas of thought bring various forms of bias and prejudice, physics may be less susceptible than most. For all these reasons, we should be proud to be teaching physics, and encourage more of it to be taught.

I will end on a personal note. From a very early age I have wanted to be a professor. The idea of spending my life teaching and learning about the world seemed like the best of all possible worlds. As I have taught more and more students, I have constantly struggled to figure out what works best. I have never been fully satisfied. Teaching is very difficult. We live for those rare moments when a light bulb seems to light up in a student's mind, and we worry when those light bulbs remain dim or flickering for so long for so many. Nevertheless, I find the challenge of teaching to be as intellectually challenging as ever, and in this day and age teaching is perhaps also more important than ever. For this reason, I am glad that AAPT exists and I am honored to have been able to contribute to it.

Thank you for your attention.