|Talk About Teaching and Learning
February 16, 2010,
Volume 56, No. 22
Peer Instruction in Large Classes
Paul A. Heiney
As a high school student, I remember visiting a local college and being asked to try out the latest innovation: a computer program that guided the student through a series of physics questions. Selecting the right answer led to a different page from the wrong answer. My friends and I quickly figured out that this was more like a test than a game, and lost interest. This was my first experience with the “computer is a hammer, let’s find an educational nail” approach to educational technology. However, in this article I will describe a relatively new piece of technology that can be used in the service of what I consider to be a worthwhile pedagogical goal: making large lecture courses feel like small seminars.
Later, as a freshman at UC Santa Barbara, I had a much more stimulating educational experience. As an adjunct to the big introductory lecture course in physics, I was placed in a small seminar with four or five other students. In each meeting, we were challenged with tricky physics problems, and asked to go up to the board in front of the others and work them out. The experience was terrifying…but immensely productive, and after this “boot camp” I felt ready to handle anything else the undergraduate curriculum might throw at me.
As a professor, I have now taught versions of the big freshman physics course many times. Everyone pretty much handles these lectures the same way. The professor explains the new concepts and equations, and then works out a number of sample problems on the board in front of the class. This is in fact exactly what the textbook does, except that the book has better pictures and makes fewer mistakes. The students dutifully write everything down and hope that they will understand it before the exam. Student evaluations tend to reflect the clarity of the lectures and the extent to which good lecture notes can be produced; it also helps if the professor is personable and charismatic.
A few years ago, despite my skepticism towards most educational technology, I started experimenting with a different approach to lectures. The goal was to replicate, or at least come closer to, my freshman experience with being forced to think, while retaining the classroom of 50-150 students driven by course enrollments and faculty availability. The tool was the “Peer Instruction” approach original developed by Eric Mazur at Harvard. (E. Mazur, “Peer Instruction, A User’s Manual,” Prentice Hall, 1997; see also www.sas.upenn.edu/computing/node/194/ and www.ctl.sas.upenn.edu/tools/VideoDiscussions.htm).
In a typical implementation of this approach, I start with a brief explanation of the next concept or topic. I then present the students with a short puzzle or problem incorporating this concept. The students discuss the problem in small groups of 2-4, and then vote using polling hardware or “clickers” similar to TV remotes. After everyone has voted, the results appear on the screen at the front of the class. A well-chosen problem is one for which about 70% of the students get the right answer. If too many students pick one of the “distracter” questions, I spend a few more minutes addressing the issues that were raised. Sometimes, I will also ask a particular student (chosen at random) to explain his/her reasoning to the class. Otherwise, it’s on to the next topic. The whole cycle takes about ten minutes.
The obvious benefit to this approach is that I get immediate feedback as to whether the students are “getting it.” You don’t just have the same four bright students asking the questions; the confused ones in the back of the room are also forced to respond. The anonymous nature of the voting is important; when I have done the same thing with a show of hands instead of the clickers, the responses are much more homogeneous, indicating that students are unwilling to vote against the majority.
A less obvious but more valuable consequence of the peer instruction approach is that, roughly every ten minutes, the students are forced to stop taking notes, wake up, and start thinking. Again, for a well-chosen problem, there is more than one answer that seems reasonable, and the discussion is often vigorous. By explaining their reasoning to others in their group the students are also internalizing their own grasp of the concepts.
A common criticism of peer instruction goes something like this: “The students don’t know anything, that’s why they are paying lots of money to be in your class. They would get much better value for their money by listening to the professor.” The crucial difference is that this is a guided discussion. I still provide the context, select the problems, and adapt to the students’ responses. By engaging in a discussion with their peers, the students have become active learners rather than passive recipients of received knowledge. A real seminar course, with 5-15 students in a small room working at the blackboard, would be better…but this is clearly not feasible for Physics 101 (or Chemistry or Psychology or any of the other large-enrollment introductory courses).
Since the lecture hour is of fixed length, something is given up, usually most of the example problems. In my view this is a reasonable tradeoff; the students are better off thinking carefully about a smaller number of problems than hearing a more superficial exposition of more examples—after all, there is always the textbook.
Will this approach work in other disciplines? I expect that it would work naturally in any subject where questions can be posed in a multiple-choice format, with one “correct” answer and several answers that are wrong but not obviously so. Thus, the sciences, engineering, medicine, and the more quantitative parts of the social sciences. Indeed, my colleague Melissa Wilde in sociology has been using a very similar approach. Perhaps less so in the humanities, where questions tend to be more open-ended. (For example, I have no idea how to usefully recast “Why is Hamlet so indecisive?” as a multiple-choice peer instruction question). On the other hand, departments in the humanities seem to roster many more seminar-size courses, perhaps for exactly this reason.
How well does it work? The jury is out, but the indications are positive. Anecdotally, a number of students have told me that they enjoy this approach, but my overall student evaluation ratings have not substantially changed. I suspect that this is because my personal clarity, charisma, etc., have not changed much. When I have used the peer instruction approach opposite other professors in different sections of the same course, the students’ exam grades tend to be similar.
A peer instruction lecture is about the same amount of work to prepare as a traditional lecture. Relinquishing some control over what will happen in the classroom is a bit of a high wire act, and can be nerve wracking at first. On the other hand, it is extremely satisfying to see sleepy eyes and faces bent over notes replaced by animated discussions.
Paul A. Heiney is a professor of physics and astronomy in the School of Arts and Sciences and
was faculty director of the Benjamin Franklin Scholars program from 2004 until 2009.
This essay continues the series that began in the fall of 1994 as the joint creation of the
College of Arts and Sciences and the Lindback Society for Distinguished Teaching.
See www.upenn.edu/almanac/teach/teachall.html for the previous essays.