Thursday, January 15, 2009

Teaching to the evaluation

Via The Edge of the American West: Texas A&M, aka The Yale of College Station, will now be giving $10,000 bonuses to faculty based on student evaluations. Generally, I support initiatives to help faculty focus on better classroom practices. I've been a college student an incredibly long time now, and I've had instructors who teach badly, often for a variety of reasons. But I've also taught, and I think most teachers understand why student evaluations are a bad metric for giving away money. Among the more compelling reasons not to do this:
  • Positive evaluations correspond to higher grades. Classes that frequently reward students with substantial high marks receive higher student evaluations. This is troubling for several reasons. First, and this has been pointed out elsewhere, making student evaluations a sole serious metric for bonuses would lead to grade inflation. And if universities need anything right now, it's less incentive for grade inflation. Second, and this is more crucial, I think, is that teachers who structure their classes around frequently rewarding students will be perceived as better teachers. Classes that grade based on more cumulative criteria--say, for example, freshman composition courses, which measure a student's development as a writer over the span of an entire term--give students less of a grade-based incentive for enjoying the course. So depending on the requirements of a given course within a given department, the timing of grades varies.
  • Race, gender, and "attractiveness" play a role in student evaluations. Interestingly enough, male teachers that students call "attractive" get better evaluations that female teachers that students call "attractive." Also, the placement of men and women in teaching posts varies from department to department, as does the gender makeup of the students from department to department.
  • Electives will, on average, fare better among students than required courses. Required courses have a stigma built in before students even enter the first class meeting: they are the lima beans of the university experience. Take these courses before you can take the courses you want.
  • Student evaluations are arbitrary and vary not only by department, but also by course. When I teach composition, I not only keep regular office hours, I require individual conference meetings with students. I also respond to all class-related emails quickly (in my opinion; and I should note, anecdotes are never great evidence). Still, every quarter a handful of students claim I was unavailable outside the classroom. Yet I received nary a peep from these students during the quarter about needing to meet in my office. From what I can tell, the students who claim I wasn't available outside class are those students whose grades aren't very good.
There's a broader point beyond that anecdote. Departmental evaluations are written to cover a very broad spectrum of courses that serve students in very different ways. Our English department, for example, offers courses (and degrees) in literary study, journalism, rhetoric and composition, creative writing, and editing. How do you objectively compare student evaluations as metrics for rating these courses?

For more on this, follow along at Edge of the American West, which has given Ezra Klein, who supports the bonuses, the corporal punishment he deserves.

No comments: