Peer Instruction is a teaching method invented in 1991 by Eric Mazur, a physics professor at Harvard University. Mazur found that, in spite of his award-winning lectures and his students’ high test scores, the students failed to grasp the concepts he was teaching.
They could crank out formulaic answers by rote, but couldn’t explain the principles they were using or apply them in new contexts. Peer Instruction, or PI, arose during a flash of inspiration (or a bout of desperation) during a lecture on Newton’s third law.
Frustrated by his own repeated failures to get the idea across, Mazur told his students to discuss the problem-at-hand with one another. What followed evolved over time into the extremely effective, evidence-based method of instruction PI is today.
Peer Instruction Described
PI begins with a brief explanation (in other words, a lecture) by the professor. This explanation should be in the neighborhood of 10 minutes. The professor then confronts the class with a multiple-choice problem to solve, also called the concept question. Each possible choice should seem plausible.
The students spend a couple minutes attempting to work out the solution individually. The professor then holds a vote: collecting each student’s answer as close to simultaneous as possible. Voting can be done using clickers, but less sophisticated means, like flash cards or show of hands, can be used as well.
The next step is determined by the distribution of answers. If most students were right, briefly affirm the solution and then move on. If the answers were predominantly incorrect, the instructor should provide a new explanation or hint, then repeat the individual phase.
If, on the other hand, the answers are more evenly divided, the special sauce is then added. The students break into groups of two or three and attempt to persuade one another of their answer. As professor Mazur found, and as numerous studies have since confirmed, the proportion of correct answers and the comprehension of underlying concepts rises dramatically.
Peer Instruction Flow Chart
Why does PI work?
A key reason PI is more effective than traditional lectures is what’s called the curse of knowledge: the inability of the expert to appreciate the obstacles faced by the non-expert. Paradoxically, the professor’s subject mastery becomes a barrier to effective teaching. Students who explain concepts to other students are keenly aware of the hurdles they just recently cleared.
Another force at work in PI, a constructivist mechanism, lies in the process of explaining. The very act of articulating ideas correctly fleshes them out and firms them up in the mind of the explainer as well as the listener. “Students learn best when active learning takes place” (Butchart, Handfield, & Restall, 2009).
Is Peer Instruction suited to Wicked Problems?
PI was invented in the physics classroom and has been applied successfully in numerous STEM fields over the years. Questions in math and science tend to be convergent: zeroing in on one definitive answer. What happens when the problem-at-hand doesn’t have a fixed answer? Can PI help us navigate the murkier waters of social science, humanities, and Wicked Problems?
Butchart, Handfield, & Restall (2009) insist “the opportunities for student discussion and active engagement offered by PI can be achieved even with open-ended questions which do not have a unique correct answer.” The authors present the following example concept question:
“Which of these outcomes is worse?
- Five people contract a fatal disease that can only be treated at great cost and with difficulty. They are not treated, and die of it within a year.
- A person murders a completely innocent stranger. The murderer feels no guilt, but never re-offends.
- Neither: they are equally bad.”
Because this question hinges on ethical values, the answer cannot be clear-cut. Nevertheless, using PI in such cases can prime the students for discussion. It can also get the constructivist juices flowing: prompting students to solidify the rationales behind their answers, however subjective they may be.
Using one of my own Wicked Problems as an example, here is what PI concept questions might look like:
What would have the greatest impact on cyberbullying?
- Nationwide laws and regulations to punish cyberbullies
- A grassroots, nonprofit advocacy and awareness campaign
- State-by-state funding for anti-bullying programs in schools
- Social norms will take shape naturally as technology becomes settled
What, if any, threat does Artificial Intelligence pose to mankind?
- None. Even strong AI will be confined to machines with no means to rule or enslave us.
- Little. It may affect our quality of life some but Terminator scenarios are overblown.
- Significant. Runaway AI is a unique and unprecedented risk that can be managed only with proper oversight.
- It’s over people. This genie isn’t going back into the bottle.
I’m excited for the possibilities Peer Instruction provides. Time will tell whether questions like the ones I’ve created above will advance the learning objectives of my TWP section.
Butchart, S., Handfield, T., & Restall, G. (2009). Teaching philosophy, logic and critical thinking using peer instruction. Teaching Philosophy, 32(1), 1-40.