Friday, January 15, 2016

Dr. Amelia thinks: Just because we are all shallow doesn't make it ok.

Two articles out this week show that we are all shallow, unfair jerks when we evaluate each other.

This one for proffies argues that students evaluate female professors lower in a way that is impossible to adjust for. This one for students argues that prettier students get higher grades.

And I will admit that this past fall semester, on the day I was going to give course evaluations I stood in the bathroom thinking if I should take time to curl my eyelashes so that some recency effect of appearing, I don't know, more awake, might tick those boxes up a notch. And then I thought, "You know, [male colleague in the office next to me] doesn't think about this kind of stuff at all." And then I curled my eyelashes.

A friend posted the first article on Facebook and one comment was, "With the exception of a modest number of students, these are popularity contests, based on two things: how cool do you think the teacher is and were you given too much work by your standards?"

I think that is right on the money, but I do think I will stilll be having mirror debates.

I am inspired to use the blinded grading mode on the LMS for the students, but there's not such a thing for course evals, unfortunately.


  1. Thanks for bringing these up, Amelia. I had always thought that with all their flaws, at least student evaluations were good for measuring things like whether the professor graded assignments in a timely way. But the online study showed that even that is influenced by perceived gender:

    In promptness, for example, the instructors matched their grading schedules so that students in all groups received feedback at about the same rate. The instructor whom students thought was male was graded a 4.35 out of 5 for promptness, while the instructor perceived to be female received a 3.55.

    1. I think that getting feedback to students promptly is one of my greatest challenges/flaws (though my guilt over this is somewhat moderated by the fact that I have what is, by the guidelines of one major professional association -- it used to be two, but the 2nd one caved to reality/actual practice -- completely unreasonable). Every few semesters, however, things go more smoothly than usual, and I'm very efficient about getting work back quickly, to a particular class, or to all or most of my classes. I always think that will help with my evals (which do, in fact, show some dissatisfaction -- I'd say justified dissatisfaction -- with the time it takes me to get work back to students). As far as I can tell, it has no consistent/predictable effect on my overall teaching scores (I haven't checked whether students think their work is being returned more promptly, or whether I'm up against a completely unrealistic imagined deadline -- e.g. within minutes/hours of turn-in -- in their minds).

      This news is not going to keep me from continuing to try to find ways to be more effective/efficient in my feedback, but it probably puts the final nail in the coffin containing any hopes that doing so would improve my eval scores, which generally aren't terrible, but are definitely not stellar, either.

  2. I've had students make very specific complaints that were wholly untrue (took weeks to get a paper graded--not true, the college mandates 72 hours and I always meet that requirement; didn't respond to e-mail for days: untrue, I always meet the 24 hour rule). Since I teach online, these complaints are easily disprovable by looking at the course data, but once they appear on my evals, the damage is done. I'm an adjunct, so that kind of thing could lead to me not getting course assignments.

  3. If administrators actually paid attention to such studies, this would presumably be good news for a middle-aged, gray-haired, fat, introverted female such as me. Unfortunately, administrators do not seem to pay attention to research that inconveniently challenges their preferred management methods.

    I must say, the use of evals on the department level at my school is actually pretty fair (they're treated as one data point among many, with consistently high ones combined with teaching materials that show rigor being -- appropriately, I'd say -- rewarded, and consistently low ones being treated as a red flag that something might be going on in the classroom that isn't detectable via curricular materials or occasional class visits, which also seems like a reasonable approach). However, administrators see the eval spreadsheets, but not all the other material (though it is summarized for them in letter form), which always leaves the possibility that we might find ourselves being managed/evaluated via spreadsheet.

    Given the overrepresentation of women in low-paid, teaching-intensive positions, maybe it's time for a class-action lawsuit? Nah; we'd probably find we signed our rights away in our contracts.

    1. P.S. I do find it interesting to teach students whose genders I don't know, which happens fairly frequently in my online classes (a few names which are gender-ambiguous in the culture(s) I know well, and quite a few more that come from cultures I don't know well enough to have gender associations with the names common in those cultures). If a student interacts with me fairly frequently via email, I usually form an impression -- sometimes unconscious -- of their gender, and I'm more than occasionally wrong in that impression, especially if the student is an immigrant or first-generation American. I haven't had enough experiences for them to rise, in aggregate, above anecdata, but I'm inclined to treat them as evidence that differences in language, expression, degree of emotionality, etc, etc. vary at least as much from culture to culture as from gender to gender.