Experiments in the Classroom: Part II

As anyone who has wandered the corridors of an economics faculty knows, we are not a profession known for our sartorial splendour. Australia’s top-ranked economist (according to the RePEC database) is Adrian Pagan, who is rarely seen in anything but a t-shirt and shorts during summer. In the US, for every Ed Glaeser (who always wears a suit), there must be half a dozen Matthew Rabins (a splendid behavioural economist with a penchant for tie-die t-shirts).

In terms of the signal that dressing up sends to other economists, my guess is that it’s strongly negative. When I arrived at ANU on my first day wearing a suit and tie, I could immediately tell that several people thought they’d hired a dud. (My protestations about having been a lawyer in a previous life did little to assuage this.)

By contrast, there is some evidence that looks matter in the classroom. In a neat paper several years ago, Dan Hamermesh and Amy Parker showed that more beautiful academics get higher teaching ratings. But does primping pay?

To test this, I ran an experiment on my Masters students this semester. At the end of each class, I asked them to tear off a scrap of paper, and write a number on it, from 1 to 10, where 1 indicated that this was the worst lecture they’d attended in their life, 10 was the best lecture they’d attended in their life, and 5 was an average lecture. They could also write comments/suggestions on the scrap of paper. (This turned out to be a useful feedback tool for me, though I don’t recommend it for the faint-hearted!)

In order to gauge the effect of attire, I alternated from lecture to lecture between a suit and tie, and a collared shirt with beige pants. I figured that if the effect of dressing up was big enough, it would show in the data.

Turned out that it did, but not the way I expected. On lectures when I was wearing a suit, students gave the class an average rating of 5.8 out of 10. On lectures when I was not wearing a suit, students gave it 6.3 out of 10. Somewhat surprised by this, I looked back at the schedule, and realised that several of the suit-wearing lectures coincided with days when the students took an in-class test. After accounting for this, it transpired that the difference was a mere 0.2 points (in favour of the suit). In other words, primping doesn’t pay… at least if you’re an economics lecturer.

This entry was posted in Economics for Government Course. Bookmark the permalink.

3 Responses to Experiments in the Classroom: Part II

  1. hc says:

    For those of us whose livelihoods depend in part on student evaluations this is a serious issue. I know many academics at both La Trobe and University of Melbourne who wear coat and tie whenever they lecture partly (i) to improve their evaluations and (ii) to convince students that they are serious and mean business.

    The development is partly a reflection on students rather than academics. The bourgeois values of students these days shock even a conservative like me!

  2. conrad says:

    I agree with HC about the seriousness of this issue — there are big problems with all these surveys, unless you are an older but not too old white male. I’m surprised that where I work they are even allowed to give us the survey comments without screening — some of the ones I have seen have been blatently discriminatory — far worse than “I don’t like the clothes my lecturer wears”. I work with one person who got something along the lines of “is old an needs to retire” twice in one evaluation. You get the reverse if you look even slightly young, which I imagine is partially to do with average age of most university staff (I constantly get asked if I am finishing my PhD, even though some of the work I have done from years ago sits in the average textbook in one subject I teach. Perhaps I need some grey hair and a pot belly to fix this). I wonder what would happen if those comments were given to a lawyer?

    I was also lucky enough to get a hold of all the mean subject results last semester where I work (where the surveys are taken seriously) and that was interesting because students evidentally can’t pick “average” difficulty. Of all the second year subjects, I believe only two (if I remember correctly) were given above average difficulty (3.5) on a 1-5 Likert scale, so students evidentally use some absolute measure, not a comparitive one (that’s not surprising since they can’t do all subjects). The other implication is obvious (i.e., is everything you learn really that easy or is it dumbing down across the board which means nothing taught is hard anymore?).

  3. christine says:

    Isn’t the story about what contributes to the individual fixed effect, though? So given that you’re talking about a bunch of students who all think you’re a weird flip-flopper sartorially, this isn’t much help. They probably thought you had job interviews on suit days, or something. You’d need, I think, to try one semester wearing a suit and the next the tie-dyed t-shirt, with no overlapping students. I tried this a bit this year, and evals went up a bit, but not very much. Another friend had multiple sections of same course, and tried being nice to one lot and (a bit) nasty to another, and got much better evals among the nice group, including on organisation, even though it was precisely the same stuff – which I think says that personality, not teaching ability, is what matters.

    Also, depends on the student group. MA (Econ) students are probably relaxed. Undergrad business students, not so much.

    An ‘average’ student evaluation at my university is a 6 out of a highest possible 7 (on everything ranging from ‘is audible’ to ‘displays concern for students’, with I guess a 4 supposedly being average. And they complain about our evaluation! (I should note, I’m apparently below average, mostly, but goodness only knows how I can get up that high.)

Comments are closed.