Wednesday, May 20, 2015


Universities supposedly have a bunch of smart people working for them.  Why do they continue to rely on such poor methods for evaluating teaching?  Is this really the best that can be done?  Why haven't we seen innovation in this area?  (Simply putting the paper evaluations online doesn't count as innovation to me.)  Below let me suggest questions to ask that might motivate the type of innovation I have in mind.

Let me begin by noting that often when I attend a workshop on teaching there is an evaluation performed at the end.  This is for a single session, which usually goes for an hour.  It suggests the possibility that something similar could happen in a course.  Indeed, I tried this once in my own teaching.  Technically, I relied on Google Forms, which was easy to use and allowed me to share the results with the students.  Content-wise, I didn't try to imitate the end of semester course evaluations at all.  Instead I concocted my own metrics for what would be it a good class session.  (The reader should note that this was in an honors seminar with only 17 students.)  Eventually the students tired of this and the response rate got so low it didn't seem worthwhile any longer.  But while we were getting good response the information the students supplied helped to shape subsequent sessions and the students could see their own influence in the course trajectory.  So one might imagine early on that this sort of more formative feedback would encourage student participation.  If anyone else tried something like this one might not want to do each session individually, but say have the survey done at some frequency - weekly, biweekly, monthly, I'm not sure which of these is best - or if the surveys should only be at the start of the course and then taper off.  What I'd like to convey to the reader here is not that I know what is optimal, I don't, but in determining what is optimal there is some tradeoff in getting timely formative feedback while not burning out the students in providing that feedback.

So envision some system of formative feedback provided by the students within the semester.  Now a bunch of related questions.

  • Should that feedback be used only within the class or should it shared with the department and others on campus to count for measuring teaching performance?
  • Should the type of questions asked in the formative feedback be standardized (within department, within college, on campus) to make cross course comparisons,  the way instructors ratings are now used for measuring teaching performance?  Would standardization of this sort lessen the effectiveness of the feedback?   Would it, in contrast, encourage more instructors to try the approach?
  • Suppose, hypothetically, that an outside observer were also to attend each class session.  Would that observer come up with similar impressions to the students, as measured by the questions on the feedback form, or differ from the students in significant ways?  Would it then matter who that observer was, whether a faculty member in the same department, a faculty member in a different department, a campus pedagogy expert, a student not taking the course but instead representing the student senate, or outsiders of campus on an accreditation visit?  
  • In that seminar class I didn't give exams.  Should there be formative assessment about the exams too?  (In a recent piece in the New York Times, Richard Thaler, one of the founding fathers of Behavioral Economics, wrote that students care about raw scores on exams even when there is grading on a curve, so rationally they shouldn't care. But they seem to prefer when the outcomes are near 100 in raw score, which seems to convey the message that mostly they understand - even if it really doesn't.)  If such formative assessment was given and if it indicated student disposition about the exams and if that proved to correlate highly with how the course was rated in the end of term evaluation, what would the campus response be?  After all, in this contingency the data would be showing that course evaluation is essentially irrational reaction to testing. Wouldn't that put pressure on campus to lessen its reliance on course evaluations for evaluating teaching?
I could go on, but I hope with this brief list that the reader gets enough of an idea of possible innovations with teaching evaluation and that much experimentation in this domain would be a very good thing, to point at innovations that really make things better

Let me close with one other point.  The piece linked below talks about instructors manipulating their course evaluations, with one technique baking chocolate chip cookies for the students.  So there has been some innovation in the way courses are conducted generated by need to boost teaching evaluations.  Might some of these innovations in instruction be useful if severed temporally from when the end of semester evaluations are given?

In particular, I like the idea of having a celebration in class somewhere around mid semester and I'm all for the instructor bringing in treats for this purpose.  It creates a little bond between teacher and students and it shows that at least in some ways they're all on the same team.   I came to understand this a few years ago when in the middle of the semester I spent 5 days in the hospital and had to miss a couple of classes as a result.  That hospital stay coincided with Halloween, which my wife and I missed that year.  As a result there was a bunch of candy in the house.  The first class where I returned to teaching I had each student come up to the front of the classroom, give me a high five, and take a piece of candy.  I know I was touched by that moment.  I think some of the students were as well.

The next year (I now teach one course a year) around the same time of the semester I received a small inheritance.  This was a different cause for celebration.  This time it was donuts and cider.  I ditched the high five, though now I'm not sure why.  Then last fall we had a celebration again, though this time there was no external event that warranted the celebration.  It just seemed to me a good thing to do based on the prior experience. 

Are there other things instructors might do as regular practice that they first tried just to boost their teaching evaluations?  I don't know.  I do know that we don't experiment enough with our teaching overall and if some experiments happened for less than ethical reasons, that should not itself condemn the practice, particularly if the ethical issues can be addressed.  So let's improve teaching evaluation.  And in the process, let's improve teaching as well.

No comments: