I will confess that, before this semester, I had a lot of doubts about the utility of peer instruction. Like many faculty, I've used small group discussion in smaller classes pretty regularly; but often, it seemed more an opportunity for students to lose rather than gain focus on the course material. Still, I was willing to give peer instruction a go, particularly because it is the only kind of discussion possible in a 400-student class. In addition, I felt that I had enough training from learning specialists to find ways to make it more meaningful. Peer networks are at the heart of MOOCs and there's some good evidence to indicate that students can do a very effective job of teaching one another. They may not have the expertise that the instructor has; but once the content has been delivered by the expert, peer instruction is very effective in getting that content to stick and encouraging active engagement with and curiosity about that content.
On Monday in the Rome class, the students had been asked to watch a recorded lecture covering the 1st and 2nd Punic War (up to the Battle at Cannae). This is complex material and, since they have an exam on Wednesday that includes this material, I wanted to be sure that they got a chance to practice it and ask questions about it. I prepared a 45 minute review consisting of i>clicker and "talk to your neighbor" questions. Just before class, I got an email from the Echo guru at UT letting me know that about 25% of them had at least opened the lecture. Of course, this information is not a perfect indication of how many were prepared: I don't know how carefully they watched the lecture; I don't know if groups of them watched it together; I don't know how many read the textbook but didn't watch the lecture. Still, it is a good rough estimate of what I can expect when I walk into the class. So I knew that about 25% of them had some idea about the material.
I started with a few i>clicker questions, some easy ones and some more difficult ones. They were all intended to highlight "stuff you should be sure you understand." The class was quite good at narrowing the answer down to two options but on several polls, it was a close vote. In those cases, I had them break out into peer discussion for 1-2 minutes and then re-take the poll. Inevitably, there was a substantial shift to the correct answer. The students all oohed at this and I smiled at the power of peer instruction at work. I also made a point of pointing out to them how effective it was to talk to their fellow students.
By the end of the class, we had reviewed all the key points of the lecture. Students who had done no work prior to class were forced to engage actively at every moment. I kept my direct comments about the material to a bare minimum and used the Socratic method as much as possible. I also did everything I could to keep them on their toes, moving from i>clicker to peer discussion to short class discussion and back to i>clicker. In this way, those who had done some preparation before class got to practice the content and teach it to others. Those who had not done any preparation got some exposure and some sense of how well they will need to know the material by exam time. Most of all, though, everyone came out of the class knowing that they could turn to their neighbor (see Julie Schell's blog Turn to Your Neighbor for much more on the power of peer-based education).