Friday, September 29, 2017

Scoring Your Professors. Evidence on Gender Bias.

Many colleges and universities use students' evaluations of their professors as part of the evidence which determines the raises and promotions awarded to professors.  But increasing amounts of recent evidence suggest that such evaluations are not free of (probably implicit) gender bias.  Now yet another working paper suggests that students don't necessarily rate their professors only on the basis of objective factors*.

The study**, carried out with data from the Maastricht University, Netherlands, exploits an institutional feature of the university which divides the students in each of the analyzed courses into three sections, led by either male or female teachers.  Each course has one course leader who assigns the reading material to all three sections, and the students cannot self-select into sections based on, say, their preferences for female or male teachers in those.

The study abstract summarizes its findings:

Any study trying to measure students' biases, including possible gender biases, must grapple with the possibility that what looks like bias might actually reflect real quality differences between male and female professors.

Teaching quality cannot be directly observed, but the Maastricht data provides a partial indirect way for checking on possible differences.  Since the exams in these courses were mostly centralized, i.e., the same for all three sections (and not given by the section leaders), worse quality teaching in some sections should be reflected in either lower average grades for students in those sections or more reported effort by students in those sections, to compensate for worse-quality teaching, or both.  No such differences were found.  The researchers conclude**

Our results show that female faculty receive systematically lower teaching evaluations than their male colleagues despite the fact that neither students’ current or future grades nor their study hours are affected by the gender of the instructor. The lower teaching evaluations of female faculty stem mostly from male students, who evaluate their female instructors 21% of a standard deviation worse than their male instructors. While female students were found to rate female instructors about 8% of a standard deviation lower than male instructors.

What I found most interesting in the working paper is this**:

The gender bias against women is not only present in the evaluation questions relating to the individual instructor, but also when students are asked to evaluate learning materials, such as text books, research articles and the online learning platform. 

Strikingly, despite the fact that learning materials are identical for all students within a course and are independent of the gender of the section instructor, male students evaluate these worse when their instructor is female.  One possible mechanism to explain this spillover effect is that students anchor their response to material-related questions based on their previous responses to instructor-related questions.

Even text books and research materials can acquire girl cooties from close contact!

Assuming that the findings of the paper are correct, what's going on here?  

I would have loved to get more information on both those male students who were driving the observed bias in the evaluations and those male students who didn't show such bias.  Is the former group younger, more conservative, more religious, from the same parts of the country or the world as the latter group?  And so on.

But absent such information, my guess is that the results are driven by the common cultural (implicit) belief that experts and authorities are male, that a female expert must first be questioned and doubted.  Why that belief would show more among male students is, of course,  something that still remains to be answered***.

Studies of this type belong to the drip-drip theory of how disadvantage by gender and/or race might accumulate over someone's career.  None of the individual events (just a little lower teaching evaluations, just a little less credit for a co-authored article, just a few pertinent observations ignored at a seminar) are that earth-shaking, but if several of those drops hit someone every year, the cumulative effect may well explain who ends up with the full professorships.
*  For those of you who are interested in the economic theories of discrimination, what might probably be studied here is customer discrimination, first discussed by Gary Becker.  See this post for a little more information.

** These links are to the working paper.

***  Bias seems obvious, but not what type it is or what causes it.  Note, as an aside, that an alternative theory could be that female instructors demonstrate a bias against male students who therefore return the favor by giving them lower evaluations.  That's pretty unlikely, given that neither  the grades nor student's reported effort levels show such pattern of bias against them, and also because other students in those sections would complain about such behavior in the evaluations.

More generally, the bias against female instructors could be based on Becker-type dislike of certain people in roles of authority, on religious patriarchal teachings, or in the STEM courses on statistical discrimination:  Applying average prejudices about female inferiority in STEM to specific individuals.

Why any of these would be more common among men is not clear, though it's possible that female students are more aware of such biases in general, given that they may have been at the receiving end.