How quickly does hotness fade?

Ratemyprofessors.com allows students to grade a professor's clarity, helpfulness, ease and - just for fun - rate their appearance as "hot" or "not". A professor with more hot than not votes is awarded a chili pepper on the ratemyprofessors.com web site.

Hotness declines with age, but how quickly? To find out, I combined ratemyprofessors hotness scores with information on when professors acquired their PhD - the best available measure of a professor's age (this information was gathered jointly with my co-author, Anindya Sen, and his co-authors). 

The locally weighted smoothed scatterplot (lowess) line above shows that a male economics professor's hotness peaks shortly after he has completed his PhD, and declines steadily from there. An ordinary least squares regression on the 306 male Ontario economics professors in our sample finds that the relationship between years since PhD and hotness is statistically significant: a male professor's probability of being  rated hot by his students falls by 0.5 percentage points with each passing year (p=0.003 - the same results are obtained by a probit analysis). 

The relationship between hotness and age for female professors appears, at first glance, to be similar to the male graph:

Look closely, however, at the horizontal and vertical axes. At a young age, female professors get relatively high marks for hotness. Reading through the comments, it seems students evaluate such professors in part, on their physical appearance, with occasional comments along the lines of "Wears sexy skirts but terrible prof". In contrast, junior male professors are more likely to get comments like "the best teacher I have ever had, he's just so silly and great. I love him!"

Given the greater weight placed on women's physical appearance, one would expect female hotness to fall relatively steeply with age. Indeed, the lowess graphs above show a steeper drop off for women than for men, but it is hard to see this, because the horizontal axis on the male graph is more compressed.

A probit regression of hotness on years since PhD suggests that each additional year reduces a female professor's probability of being rated as hot by 1 percentage point, however with a sample size of 77 individuals, and p=0.156, the results fall short of statistical significance [I was able to regain significance, however, by dropping one older female professor, whose chili pepper seems to have disappeared since the data was gathered.]

The female professors who are still hot over the age of 50 tend to be outstanding instructors. For example, one gets this rave review "Totally love this prof. She's the most challenging prof ever. She will push and push you to become an active student..." Quality teaching and engagement with students come through in the reviews of hot older male professors as well. For example, one is described as "a pillar of authority, credibility and respectability." However older male professors have been known to get positive comments on their physical appearance along the lines of "Looks a bit like Tim Robbins."

Some might question the intellectual value of an analysis of hotness, and point out that sites like ratemyprofessors, where students can post anonymous degrading remarks, contribute to the challenges young scholars face. Is this post any better than the threads discussing the hotness of various female (and sometimes male) economists on www.econjobrumors.com? 

The fact of the matter is, universities are in the business of serving students, and students care. They're spending hours looking at a professor, and appreciate anything that makes those hours more enjoyable. Students believe they learn more from professors who are likeable, good-looking, well-dressed, and approachable. 

Some economics professors try to signal their brilliance by following Ariel Rubinstein's advice and wearing jeans and a t-shirt. Do student fail to get the message, and think the professor is simply scruffy or indifferent? I've looked up the ratemyprofessors evaluations of some male economists who are rarely seen wearing anything but jeans, and students don't seem to mention dress sense as an issue, so it's hard to know.

With increasing competition from on-line education and greater demands from the public for accountability, universities are facing pressure to step up their game. Since it's hard to measure how much students actually learn in university, I believe we will see an increasing focus on metrics such a student satisfaction. To the extent that professorial hotness is correlated with student satisfaction, it is useful to understand how it is generated.

Comments

[Giovanni]

Frances,

Do I detect a bit of a "Sean Connery effect" at the far end of your boy-economist hotness curve? Or is this just wishful thinking on my part?

Also, this work practically cries out for cross-discipline comparisons. Who`s hotter - economists or paleo-anthropologists? Molecular biologists or Middle-English specialists? Computer scientists or - well, just about anyone? Inquiring minds want to know!

(Seriously...interesting work.)

[kevin denny]

Hotness looks pretty correlated with productivity so maybe thats what students had in mind as in "Wow dude that AER paper was smokin' ". On anecdotal evidence (i.e. blind prejudice) mathematicians are the least hot.

[Frances Woolley]

Giovanni - the one thing I suspect from poking around in the data, but am not 100% sure about, is that biz skool profs are hotter than straight econ profs.

[Frances Woolley]

Kevin: "Hotness looks pretty correlated with productivity"

This is the sort of thing we're trying to look at in the paper (I'm not putting the potentially publishable stuff on-line!). My sense is that the hotness score is a better measure of overall awesomeness for male than for female econ profs.

[Shangwen]

Cue the reference to a Van Halen song....

Biz skool profs are more likely to wear suits (that Banker look the chicks dig). Anecdotally, I recall from my prolonged university days that students joked more about the profs who had attained that sub-clinical sloppy look (holes in pants, same sweater every day, stains, etc.), but if there were discussions about the hot teachers, I wasn't privy to them. Perhaps the ratings are influenced by unreported conversations among students, in which certain profs rise up the hierarchy.

One could argue that this is a valid issue to study because hotness--heavily mediated by how well you take care of yourself--is partly a function of underlying confidence, which would also contribute a lot to a person's productivity, social skills, clarity, and general approachability.

[Giovanni]

"Hotness looks pretty correlated with productivity"

If by "productivity" we're talking research output, I'd be surprised if hotness scores reflect this. One of the things that has struck me in looking at ratemyprofessors is how little students seem aware of or impressed by a professor's standing in their field. I've seen any number of "big wheels" get terribly poor ratings, often embellished with scathing (albeit, literacy-challenged) comments. (A couple of my old profs - both easily top-five in their fields in Canada - regularly get assessments that are sack-worthy.)

[Giovanni]

Shangwen,

Don't mean to be pedantic, but surely you mean ZZ Top ("Every coed's crazy 'bout a sharp-dressed prof...")

[Linda]

Re Frances' comment that "..biz skool profs are hotter than straight econ profs". Do you have any (relatively simple) way of checking to see if this affects the ratings of econ profs in places with separate business schools? If the biz schools do not offer their own econ courses, then many students in the first two years of econ courses might be in business, or trying to get into business..which seems to lead to lower ratings of econ courses/instructors.

[Frances Woolley]

Linda - we'd have to work out which profs taught which courses, which would be a lot of work. I guess that's something we could code up - I haven't thought of that.

[Nick Rowe]

I was really surprised to see that hotness initially increased with years since PhD. Eyeballing it, both seem to peak at around PhD+7 years. Which would be around...mid 30's, on average? And the students doing the rating must be early 20's, on average?

Hang on. Maybe there's some sort of vintage effect(?) going on here. "Years since PhD" means "2013 minus year completed PhD"? But some of those ratings must have been done several years ago, when the profs were younger, and won't represent "hotness in 2013"? I can't get my head around this. My brain hurts. I still can't see why that would make the curves initially rise.

[Frances Woolley]

Nick - as usual, your intuition is spot on - rmp scores are cumulative, and it takes a little while before a student comes around who loves you enough to give you a chili pepper (we're not allowing people to carry over chili peppers they earned while teaching as grad students in another university).

When I did the estimation, I tried a quadratic specification i.e. years since phd and years since phd squared. I didn't fit the data as well as the linear specification. So I think that the graph is a little misleading, it's just picking up a few people.

[Stephen Gordon]

Ah. So my initial inference - that peak hotness is obtained around the time a new professor gets tenure - was incorrect.

[Giovanni]

Depends. Self-assessed hotness probably does max out around then. Externally assessed hotness...in most cases already hurtling down the slippery slope to no-chili-pepper-land.

[Jacques René Giguère]

Formerly, it was coolnes. I was still cool in my early 50's but I never had the chili..
Couldn't they had both as they are different valuable qualities?

[Frances Woolley]

Giovanni ;)

Jacques Rene - ratemyteacher has cool sunglasses, ratemyprofessor has chili peppers. I guess the two sites are appealing to a slightly different demographic. I just looked up Carleton on ratemyteachers, and I could see hardly any ratings at all.

[Peter N]

Does it matter what sort of course is being taught? IIR from many years ago, courses for majors were rated a bit differently from survey courses for non-majors. One of the highest rated professors (for many years running) taught a double credit course 6 days a week at 8:00 AM. As you might imagine there wasn't a non-major to be found taking it.

[nottrampis]

We read you downunder. Indeed all of you.

however Frances we think you are not losing your hotness!!

I love these quirky articles

[Jon]

Slightly confused... the female data goes negative? Presentation would be more powerful perhaps with both men and women plotted on the same graph.

[Frances Woolley]

Not trampis - thanks for the link and the love from down under!

Jon - hotness is zero/one, and the line is just trying to fit a bunch of data that's all clustered around zero.

Peter N - ratings vary a lot from course to course. For example, always get far lower evaluations for intermediate micro than my other courses, no matter how hard I try with that course. One can see that in ratemyprofessors evaluations, too, for example: "During math econ, he was the mad, crazy, extremely hard, and messed up in the head greek. For Stats, he was cool, smart, and was a great teacher. I don't know what happened. Must have gotten layed during the break." Since RMP has no course info (and the info that's there is often inaccurate) this makes me think that I should be collecting info on courses taught, but it's not that easy to get (but it's not impossible either, class schedules are usually publicly available on-line).

[Joe Heath]

I've often heard the view expressed by colleagues that the extent to which you remind students of their parents has a strong impact on evaluations. The sweet spot is when you are older than the average grad student (so that they aren't feeling competitive with you) but younger than their parents (so you don't seem objectionably authoritarian). This means your 50s will be the nadir. However, once you get into your 60s you begin to remind them of their grandparents, and so supposedly you get a bit of a bump in your evaluations. I can see, however, why your hotness score would not recover.

[Frances Woolley]

Joe - I don't know if I agree. I figure there's a sweet spot when your students are almost exactly the same age as your kids because then you know every TV show and movie and song they listened to as kids.

[Ralph Musgrave]

Heat won't pass from a cooler to a hotter,
You can try it if you like but you far better notter,
You can try it if you like but you'll only look a fooler,
'Cos the cold in the cooler will get hotter as a ruler.
For more of this nonsense, see:
http://www.iankitching.me.uk/humour/hippo/entropy.html

[Ralph Musgrave]

To listen to the above "cooler to hotter" poetry put to music, go here:

http://www.youtube.com/watch?v=VnbiVw_1FNs

Starts 2 minutes into the video.

[Peter N]

If you could show the contributions of course hotness and professor hotness, it would get really interesting.

[The Keystone Garter]

When I was lonely in the 90s, no one who is here now was there for me. Which is fine; rewarding relationships aren't given. And I gave myself a mild mental illness so I would study better; I did this before puberty. I waited for a woman to ask ask me out to prove she might be utilitarian, and it never happened. And I used women for learning, grieving, and board. Now that I've lifting my mild illness, the first thing I'm noting here is I like making the world a better place much more than I like wasting time. And I'm strong enough, at least now, that I don't need TLC to be effective. There is one solution: I have to be here now, but you have to be there then. I'll never question my younger self again.

[The Keystone Garter]

...if it isn't clear, the answer doesn't matter if you matter.