Who's better looking: artists or scientists?
Well, artists, obviously. (I'm including things like design with art, and things like engineering with science.) Most of my readers know already what the typical group of graduate students or science professionals looks like, and regardless of how this compares to a random sample from the population, it's nothing like a group of artists or designers.
Since scientists tend not to hang out with artists, perhaps a few visual aids will help. Consider the first season of Project Runway, the fashion design reality show. I'm looking just at the first season because they were the most talented -- each season thereafter included people who tried out for, but weren't chosen for a previous season. The point holds for the later seasons, but you really see it in the first one. Here is a gallery of their pictures. To see them more in the flesh, you can watch the entire first season by clicking on the "littlespot" channel in this search. Around page 11 of his channel are the first season episodes. (An added bonus of this season is that the models are the best looking group of all four seasons.)
For the females, Alexandra looks like a bikini model, and Nora and Starr are also quite attractive, with Kara Saun somewhat less so. Vanessa has a manly face but is at worst average, and Wendy is a 40 year-old overweight mother, so she may not look as attractive as she may have once been. She appears to have dieted and had some kind of plastic surgery after the show, and looks slightly above-average. Remember: plastic surgery can't perform miracles, so this is probably what she looked like when she was younger.
For the males, I can't judge, so maybe female readers can rate their physical attractiveness. But I can still tell that Mario looks pretty good (you may have to see him on the show), and perhaps Robert too.
If you want a better idea, a crude but effective way to see this is to go to hair salons where the men's haircuts start at about $25 -- not necessarily the most expensive, but some place where the stylists will be selected for talent. A similar percentage of the female hair stylists there will be good-looking. Compare the percentage of lookers within this group to their counterparts at an engineering firm or research lab of a similar level of talent. Or if you live near a good design school (RISD, Parsons, Pratt, etc.), you can observe the looks of its students compared to a good engineering school (MIT, Carnegie Mellon, etc.).
What accounts for the better looks of artists? Continuing two previous posts on attractive songbirds and brainy fashion models, I suggest that there is greater cross-assortative mating between talented males and attractive females within the artistic vs. the scientific community. We know that rockstars and musicians are far more likely to mate with attractive females, compared to male scientists, so something similar must be going on with male visual artists and good-looking girls.
Both artists and scientists are hardworking, creative, and smart -- specifically, in a more visuospatial than verbal way, so it's not like the case with writers or lead singers, whose verbal skills may aid them in seducing women or attracting groupies. But artists on average have more of a bad-boy personality that attractive females seem to prefer. In this way, the children of the hot mom and talented dad will tend to possess the qualities of each parent, so that over time artists will look better than scientists.
There is actually an empirical test that this idea could be put to. First, check that the looks-talent correlation really is due to something like cross-assortative mating. You can check that by seeing if the better-looking siblings within a family also tend to be the more talented ones. You don't need a scalar measurement of talent or looks (as we have with height), just a rank correlation would do. The prediction is that looks and talent will correlate hardly at all within a family -- the siblings are choosing genes from the same parents, but it's at random, so a sibling can't choose to get the best from each parent. The opposite finding would suggest that a property of individuals were at work, such as genes that have effects both on looks and talent, or sheer chance (i.e., the sibling with the least amount of bad luck would therefore be better looking and more talented).
Then you could test the preferences of attractive females -- if forced to choose, would they rather have sex with an artist or scientist? You may have to work out some wrinkles in the wording ("have sex with," "have kids with," "marry," etc.), but that's a little detail. The prediction is that attractive females would prefer artists to scientists. There's no need to test the alternative that attractive but untalented males pair up with talented but unattractive females -- that's not what happens in general, let alone among artists and scientists (since any ranking of artistic or scientific talent will show a preponderance of males).
There's another explanation: that being attractive helps your career more as an artist than a scientist. Perhaps you have to meet with clients more as an artist or designer, and people trust the beautiful more than the homely, so that better looking artists climb the career ladder more easily. This assumes that there is no real correlation between looks and talent -- the apparent correlation would just be due to the politics of the art world.
I don't buy that, though, since as I showed in the "brainy models" post, the correlation between looks and smarts is real. It's not as if models are chosen based on their college-level intelligence -- you have the right frame or you don't, you can walk well or you can't, and so on. The ability to pass college-level science or economics courses is not involved. Maybe a Machiavellian personality could help you make up for a lack of "model looks," but that's a personality trait, not a facet of fluid intelligence (puzzle-solving).
So if someone has the free time, resources, and interest, there's a little project that's already pretty much set up.