The New York Times education section had a long article about teaching evaluations last Sunday. It starts with the story of an adjunct who did not have her contract renewed because she did not have over 85% good or exellent marked on her student evaluations.
It uses this to launch into an extended discussion about the worthiness of student evaluations.
The article dwells on the vast differences between student perceptions, especially in classes that challenge students assumptions about their world. It also includes an upper-class white male student bashing segment, accusing them as a group of ruining a professor's class on "whiteness."
In the end, there are no definitive studies that show that they tell us much of anything. We have had our own debates over evaluations over the years here, so this made for an interesting read.
I have never been much of a fan of student evaluations, because they don't really seem to tell us much. I occassionally look through the written comments, as those are the only things that hold potential interest for me. Student evaluations have something to tell us about the shape of a class, but perhaps more about the students than the professors.