I attended a webinar before Christmas that was hosted by Eric Mazur, a well-known Harvard physics professor.  Setting aside how exciting it was to have a guest speaker talking live and taking questions from his office in the US, the subject of his talk was really fascinating.  Mazur has developed a series of techniques that can change the way we teach and here he was discussing “peer instruction”.

In a traditional lecture, the lecturer stands up at the front of the class, speaks for an hour or so, then dismisses the class. This process is repeated a dozen or so times, then the students take a test and the lecturer gets to find out how much they learned.  Often there will be (i) opportunities to discuss the course with the lecturer, (ii) midterm assessments, (iii) feedback sessions, and/or (iv) tutorials, all of which help the lecturer find out how the students are doing.  The trouble is that this requires students to engage (which they sometimes don’t) and it provides only occasional snapshots of understanding.  Furthermore, even if there is an attempt to remedy a lack of understanding (e.g. through recapping material) there is no guarantee that this will help.

Peer instruction is different, and is illustrated below:

At some point in a lecture (it can be used to start the session), Mazur asks his class a question.  These questions (which he calls “ConcepTests”) are questions that probe difficult concepts to test understanding. At this stage, students spend a few minutes thinking about the question and come up with an answer. Mazur then reviews all the students’ answers (given either using clickers or a show of hands). If there is a difference of opinion, Mazur will ask the students to discuss the question in groups and attempt to reach a consensus on the answer. He then reviews the students’ answers again to determine whether greater understanding has been reached.  At that point, he can either move on or give more information to help.

Peer instruction seems to me to have five key strengths:

1. There is near-constant evaluation of student learning, while traditional techniques tend to focus assessment on examinations.
2. Peers explain concepts to one another, which means that students are conversing using their shared vocabulary and experiences.
3. It breaks the monotony of lecture classes, where plenty of studies show a lack of engagement over periods longer than 15-20 minutes.
4. It works in large lecture classes, where interaction is complex. Students often underestimate just how difficult it is to teaching upwards of 200 students at the same time, but peer instruction offers a way to do just that.
5. It is evidence-based, with plenty of literature to back up Mazur’s claims of enhanced understanding with peer instruction is used.

This is certainly something that I am going to be trying soon.  Have you tried it, and, if so, how did you find it?