Lots of debate over my endorsement of Texas A&M's decision to offer a $10,000 bonus to teachers who receive superlative evaluations from students. For some good counterarguments, see Rob Farley and Dana McCourt. The basic disagreement is the same one you find in most debates about pay-for-performance teaching: How do you measure performance? In this case, the metric is the subjective evaluation of students. And as McCourt says, "Providing a handsome reward for good student evaluations isn’t going to give professors an incentive to care about teaching; it’s going to give them an incentive to care about whether they get good student evaluations. There is not necessarily a lot of overlap." The question is whether there's enough overlap. Keep in mind that we're talking about a bonus for the best evaluations rather than salaries based upon evaluations (the latter would leave me much more skeptical). This is, as McCourt says, an incentive for professors to care very much about getting good student evaluations. I think, in the aggregate, that that's a good thing. He doesn't. Nor does Rob Farley, who notes that he gets "fantastic" student evaluations, but believes they "have almost nothing to do with teaching skill or effectiveness." He goes on to say that he "can think of half a dozen ways to pump evaluations, none of which have a positive impact on student learning." I'm sure he could also think of a half dozen ways to pump evaluations such that students would enjoy going to class more, and would learn more. A clear class outline put on the projector, for instance, so students could follow the verbal presentation and understand the structure of the argument. An animated lecture style. If these incentives compel some efforts that make teaching better and some that simply make class more enjoyable, I'd consider that a policy success. There are legitimate concerns, of course. Evaluations are connected to the ease of a class. They're connected to the attractiveness of the professor (though this is complicated: Attractiveness can correlate with social, and thus teaching, skills). But Farley, by his own admission, gets fantastic evaluations and presumably doesn't consider himself a grade inflater or a wildly easy professor. It's possible his rugged good looks are saving his evals, but I think it's more than that. I think he gets good teacher evaluations because he's, well, a good teacher. And some of that has to do with simply being friendly and interesting and accessible and young, qualities which aren't about student learning per se, but probably help students learn. Lastly, there were a lot of e-mails that accused me of using "anecdata" in the post. And it's true. I don't know whether, in their heart of hearts, most tenure track professors truly care about teaching. Nor does anyone else. I know what I've been told and what I've seen. Both have left me extremely skeptical that tenure-track researchers feel able to focus on teaching. But this is all anecdata, on all sides. So give the Texas A&M policy a try. If it's a failure, then remove it. But the idea that student satisfaction should have no role in teacher compensation is, to me, an odd one. The customer isn't always right. But they sometimes are. Update: Mistook Dana McCourt for Eric Rauchway. Fixed.
