One of my maxims in evaluating faculty is this: The more stock a professor takes in his student evaluations, the more concerned I am about the integrity of instruction. Mark Steckbeck, discussing professor ratings now being online, gets at the same concern:

A diploma certifies that a student has some base level of knowledge sufficient to represent the objective(s) of the institution. Therefore, what I am supplying in the form of education is actually an obstacle to what some students ultimately seek: that diploma. I look at most � not all � disgruntled students as individuals who I blocked from obtaining their desired end. My job is to preserve the integrity of the institution by preserving the signal of the diploma. This means evaluating each student�s ability at the end of each class and flunking those who in my estimation have not proven competence in a subject. Their negative remarks are often part of their outburst of anger that results from their failure.

Although some students may have constructive comments about their professor�s performance, the incompatible objectives discussed above lead to too much noise in student evaluations of faculty to make them meaningful. Worse still, if administrators weigh evaluations for tenure and promotion there is a resulting perverse incentive for faculty to degrade the integrity of the institution.

SCSU's contract allows professors to define their own packages for promotion and tenure as they see fit; I cannot compel anyone to include material in their packets, nor can anyone in the administration. But I do not comment on a faculty member's student evaluations in my recommendation for promotion and tenure unless I also have some assessment that education has occurred.

A testable hypothesis: There is a positive correlation between student grades and student evaluations. If you could test this holding knowledge or skills added constant, you'd have an interesting paper. Call it 'evaluation inflation'.