If you think student evaluations of teaching (SET) tell you how much students learn from a teacher, get used to disappointment. SET scores are not correlated with learning (Uttl, White, & Gonzalez, 2017), they can create perverse incentives for teachers, and they are biased (Carpenter, Witherby, & Tauber, 2020). Criticisms like these have led some to suggest that SET should be abandoned. I argue that they should not. SET are biased, but most of this bias comes from the students, so there is no unbiased alternative. They do create bad incentives, but they create good incentives as well. Most important, SET measure student opinion. Knowing what the students think is not as good as knowing how good a teacher is, but student opinion does matter. SET are easy to misread, but useful if one knows how to use them. As Mike Tyson said, "It's good to know how to read, but it's dangerous to know how to read and not how to interpret what you're reading." 1 Why should you listen to me? I am a professor of cognitive psychology at Williams College, a liberal arts college that prides itself on good teaching. I recently helped design a new SET form for Williams and I have written about SET before (Kornell & Hausman, 2016). Also, defending SET goes against my own bias: I hate SET. My first year at Williams College, I was shattered by my low ratings. Ten years later, my ratings are average, but like a rat that has been shocked repeatedly, I still freeze when I think about students judging me.
Three Shortcomings of SET"The best argument against Democracy is a five-minute conversation with the average voter." -Winston Churchill 2Other articles have detailed the shortcomings of SET (