I shouldn’t bite the hand that feeds me, but here goes. It’s teaching evaluation season again. Students fill out forms at the end of class rating their teachers on a range of qualities, and we carefully tot up the numbers (or rather, some computer does). I think this is nice for the students, and, so that I get something useful out of it, I ask them specifically to comment on issues concerning teaching style and topics in the course (I had one topic in my contemporary moral issues course this term that I definitely thought didn’t work, and was interested to see if they agreed). My department prides itself on maintaining reasonable teaching standards, and we take the evaluations pretty seriously when it comes to merit raises. I should preface these negative comments by saying there is no sour grapes here: my evaluations tend to be good, in fact better than I think I deserve and better than any other mechanism of evaluation would produce for me. Here are some observations.
I do not think the central problem with teaching evaluations is the ‘popularity contest’ problem. Students are reasonably savvy about attempts to ingratiate, and are frequently irritated if they feel that they are not learning much in a course. I am sure that an easy A makes for a better evaluation other things being equal, but it does not automatically produce a good evaluation; lots of students do want to learn too. (I realize that hard grading probably does damage evaluations, by contrast. I have no direct experience, because my grading standards are pretty constant across courses. Experimenting on this would be unethical, no?).
The main problem with evaluations is that students don’t reflect a great deal on teaching quality, and do not have a great deal of information on which to base their judgments. So they often make a judgment about the instructor that is really triggered by features of the experience over which the instructor had no control. The two key variables I notice are i) the quality of the TA and ii) the quality of the students.
The quality of the TA can work either way. A really bad TA can ruin a course, beyond the control of the instructor, and evaluations of both suffer. A great TA (and I have had a run of them) can make the instructor look really, really, good, if the instructor is good enough; but make the instructor look bad if the instructor is not good enough. (Relationships between TA and Instructor may have an impact here: I speculate that I do well from having really good TAs, and until this moment thought it was because I was good enough that they don’t make me look bad, but it may just be because I seem to be on good terms with them). In a department like mine, in which we get no choice over who TAs for us, we do not deserve the halo effect, or the reverse.
The other variable is the quality of the students in the course. My observation is this: if a course has a critical mass of good-natured, smart, and vocal students, it works well, and my evaluations are the better for it. Without that critical mass the course does not work, and my evaluations suffer. I suppose it is within my power to prevent a critical mass from emerging, but it is not within my power to ensure a critical mass. My worst ever evaluations were for a course at 8 am; I was lively and full of energy, but none of the students were; they took it because they were the group of students too off-the-ball to register in time for more reasonably scheduled courses. My best evaluations were last semester for the best course I ever taught. I co-taught with a much more experienced professor in another department, and our styles complement each other well. But what we were rewarded for was the once-in-a-decade accident of having not just a critical mass of serious, smart, and lively students, but a classroom full of them. All we did was refrain from wrecking it. But the students, who have not (in most cases) been taking 10 years of courses, attributed the fantastic classroom atmosphere not to their peers, but to us.
I don’t have a conclusion; I just think evaluations should be eyed with skepticism. Maybe, also, a preamble explaining to students what they re for, and emphasizing that they are being asked to make judgments, not merely to express their own reactions.