February 7, 2010

Peer reviewing, redux and redux

I regularly attempt some form of student peer review process for short writing assignments and am regularly frustrated with the inadequacy of the process--inevitably someone falls down on the job either in terms of participating in peer review or submitting draft work or providing advice that is counterproductive... and leaves an awkward gap and a reasonable question by students: "Do I follow my classmates' advice if I'm not sure it's wise?" That's the reaction that gets me to suspend peer reviewing for a semester or two and restructure. Yet I keep returning to it because there is value in having students look at a range of actual work and think about their own work in comparison.

In odd moments over the past few weeks I've been thinking about a different structure, focused less on providing written feedback and more on just comparing a sample of assigned writing. To give the exercise some meaning (and motivation), I've been thinking about incorporating student judgment into the grades for the assignments. (The idea is that if student ratings or rankings make a difference to peers, they will take the process more seriously.)

That's nice in theory, but how do you do that in practice in a way that is fair and does not provide perverse incentives (with students' having a reason to rate peers either high or low in some fashion)? (Please don't tell me not to grade: I'm not at UC Santa Cruz or New College!) That took me in some interesting directions, through some psychophysics literature (Thurstone's "law of comparative judgments"), some folks named Bradley, Terry, and Mallow (and Kenneth Arrow's preference/voting paradox), and eventually into BCS football rankings and the international chess fideration ranking system. Because I did some of my reading late at night, and some of this is far out of my comfort zone, my brain was hurting at several points, but I think I have stumbled onto something that is simple enough to administer, consistent with how I'd like to look at the inclusion of student judgment, and gives students a structure that is easy to respond to and gives them a reason to be honest in expressing judgments.

Again, that's in theory. It'll probably be the end of 2010 before I know whether I'm completely bonkers or if it's workable.

Listen to this article
Posted in Teaching on February 7, 2010 3:00 PM |