Peer Instruction

Peer Instruction


This lecture was part of the Center for Teaching and Learning's Transformational Speaker Series and the video of the lecture and the workshop will be available on their website soon.

Peer Instruction Model
1) Lecture for 7 minutes or less
2) Ask a conceptual question (usually multiple choice but could be free response)
3) Give people 1-2 minutes to think then everyone must commit to an answer using their clicker
4a) If less than 30% of students answer correctly the concept is revisited with lecture or an easier question
4b) If more than 70% answer correctly instructor goes over the correct answer and moves on
4c) If 30-70% answer correctly students pair up with someone who answered different from them and they each try to convince the other student that they are correct
5c) Everyone again commits to an answer for the question whether they changed their mind or not
6c) Instructor goes over the correct answer and moves on

After the second clicker poll students are shown the histogram of student responses. Students are NOT shown the histogram after the first poll because then the histogram would skew the student discussions that follow the first poll.

Peer instruction has been show to double or even triple student gains on instructional tools like the force concept inventory. By comparing student performance on calculation-based exam questions after Mazur started using peer instruction to years before he discovered peer instruction he found that peer instruction also improves general problem solving strategies which extends to computational problems.


Misc. Notes

  • Learning consists of two parts: Information Delivery (easy) and Information Assimilation (hard). Traditional lectures focus on the first part and leave students on their own to accomplish the second part. Peer instruction attempts to shift the delivery part to the students at home (pre-lecture reading) and the assimilation part to the instructor during class time.
  • A problem is a situation where you know the desired outcome but you don't know how to get there. Most questions we give on assessments are not problems, they are situations where the student knows how to get there but doesn't know the desired outcome.
    • Ex) If a student looks at a circuit diagram and immediately knows they need to use Kirchoff's Laws then where is the physics? The physics part of any question is determining the method and the approximations/assumptions necessary to reach the solution, the rest is just math.
  • Open book exams help shows students that the class isn't about memorization but rather understanding and applying concepts.
  • Five days before lecture students are assigned a reading from the textbook and are required to answer three online questions before lecture. These questions are graded on effort not correctness and count for 10% of the student's semester grade. Two questions are content questions and the last question is always "what did you find most confusing?"
    • Often times the best students list many specific things that they are confused about while students at the lower end tend to say everything makes sense.
  • Mazur reviews student responses before lecture and uses their statements of confusion to write clicker questions. In lecture he will even post anonymous quotes from students about their confusions to show students they aren't alone in being confused and to show them he's listening to them. This technique is often referred to as Just in Time Teaching
  • When students answer free response questions in class or on exams Mazur will write down the most common incorrect answers to be used as multiple choice clicker questions in the future.
  • The goal of clicker questions is to get students engaged and thinking. With this goal questions without a correct answer still work as clicker questions.
  • Mazur designs his syllabus backwards starting with the desired student learning objectives then determining what would be considered acceptable evidence that a student has achieved those objectives and then finally end with the determination of what topics should be taught to reach those objectives. This is known as Understanding by Design
  • Mazur has written a book on Peer Instruction