Solving the Problem of Problem Sets

I love great conferences. The plenary speaker, Simon Bates (UBC) talked about a brilliant replacement for our traditional weekly problem sets, and as I ride home on this lovely train, I have to share his ideas with you!

We have a problem with problem sets.  We need our students to practice challenging quantitative problem-solving if we hope to support the development of critical thinking scientists and mathematicians.  Yet every week, we spend precious department dollars marking and giving feedback for... nicely copied internet solutions.  We return this great feedback to our students, who glance at the red number at the top of the first page and promptly stuff the papers in their bag.  Rich and real problem-solving? Critical thinking? Revised understanding from feedback? Learning? Not really.

I loved Simon's solution. He introduced us to PeerWise (http://peerwise.cs.auckland.ac.nz/), an exciting (and free!) tool that has the potential to provide much richer learning opportunities than weekly problem sets. PeerWise is a online social platform where students generate their own "problem set questions" (with accompanying solutions) on the week's material.

The prof begins by taking care to train the students in how to generate really good problems that exist sweetly in the students' zone of proximal development. Then the students are required to write 1 problem, answer 5, and comment on 3. Simon suggested keeping the number of problems that the students are required to write low, so that the quality remains high - which makes sense given that time is usually our biggest obstacle.

Then the magic happens. Students write questions. They're goofy, they're fun, but most importantly - they're really good problems. When your friends, not just your teacher, will be evaluating the quality of your work, there's considerable motivation not to take the easy way out of an assignment.  The web interface allows students to both answer and comment on the problem - similar to how you might comment on a facebook post. Sometimes, the author gets it wrong. But the instructor doesn't step in.  Instead, students debate their solutions through this forum, collaborate, and write a better problem together! Finally, the prof puts the marks where his/her mouth is. Some of these best problems appear on the final exam to add extra motivation for students to learn with their peers.

I'm so excited about this. Students have a natural investment in figuring out the answer because they created the question. I can't wait to try this with my class!