Female Science Professor has been talking about student evaluations lately. (here are some other Student Eval posts - post 1 post 2 post 3). I had some ideas on student evaluations, and here they are.
A friend of mine likes to say that student evaluations should just be one question: "Do you like this instructor?" Maybe that is the only reliable information you can get from a student. Perhaps that can even be useful. Here is an indication of the problem. We have on our evaluation form (which is filled with useless questions) the following question:
Agree-Disagree: Instructor starts and ends class on time (1 - Strongly disagree....6 - Strongly agree)
I pick this question because it has a definite answer. Essentially, everyone should agree on this answer (including the instructor). For the past few years, I have been very careful to EXACTLY start and end class on time (and I mean down to the exact second). It has kind of become a game of mine. So, everyone should agree with the starting and ending class question, right? Well, no. I don't get 6's on this question. The scores I get are similar to the scores for the more subjective questions. Odd.
My friend from above suggests that many students treat the evaluation as though it were just one question.
Here is the other problem. The average scores for the University are on the order of 5 out of 6. Why is this a problem? This is bad because it gives very little room for discrimination between scores. It makes it really difficult to get an above average score when the average is 5.2 and the max is 6.
This could possibly solved by giving some explicit instructions (and training) to students. How about a 1 means the instructor is in the bottom 2% (or some other value) of all instructors this student has seen and a 6 means top 2%. That would make the 6 and the 1 scores mean a lot more.
I have to evaluate faculty that go through the tenure processes. We are supposed to use the student evaluation scores, and surprisingly I do. The important thing to realize is that the student evaluation scores are just one way to try and assess someone's teaching ability. The problem is when departments use this as just about the only measurement. For me, I really think of the evaluation scores as an indicator. If a faculty member has really low or really high scores, then I need to look at that class more carefully. Could be good, could be bad. It is just an indication that something could be wrong.
