We’re back with another fun OKCupid paper. This time with the ever productive Ed Dutton.
- Dutton, E & Kirkegaard, E. O. W. (2023). Gentle-men prefer faces: Intelligence, life history strategy, and sexual preference for attractive faces over attractive breasts, buttocks, and legs. PsychReg
We utilised data from the OkCupid dating site to study preferences for body parts in attraction judgements. We found that most people (71%) have a stated preference for faces, with non-Europeans finding other parts more important: lower parts (“butt/legs”; 8%), chest (“chest/breasts”; 3%), and Other (18%). We regressed these preferences on likely causal variables. We found that slower life history speed, female sex, and higher intelligence predict preference for faces and negatively for buttocks/legs/chest, replicating previous research. We further found that most non-European ethnicities have a stronger preference for buttocks and legs, in particular Africans, Hispanics, and Amerindians (Native Americans). We explored how these patterns may help to explain cultural differences in mating display for different body parts and how they may be explicable in evolutionary terms.
It’s the usual sort of study. We used the OKCupid dataset, which is composed of some 60k+ people who filled out detailed dating profiles on a dating site about a decade ago. Some people filled out 1000s of questions (in order to get better matches), and we used a scraper to get all this data for research purposes. You can download it too (click supplementary files). You can also read about the drama this data release caused back in 2015. It has become somewhat of a staple in “data ethics” books to give it as a “don’t do” example.
Using the dataset, we looked at people’s self-reported preferences for body types, as well as their life history speed (r-K theory), and the usual covariates. The simple results are:
So most people say they prefer faces, but women prefer them more than men do. Considering the prevalence of semi-naked boobs and butts in media, one might think that perhaps self-report isn’t too accurate here. But people were only asked “Which is most important to you in determining attractiveness?” and they could only choose one option and could not rank order them. We will do a better study later, but wanted to get these initial results out.
Life history speed was measured in the usual way, with questions such as “(q41953, n = 48,614) asked “About how long do you want your next relationship to last?” with options “one night” (2%), “a few months to a year” (29%), “several years” (27%), and “the rest of my life” (42%). The second (q20021, n = 40,763) asked “Which of these options most closely describes what you’re looking for in your next relationship?” with options “Someone to come home to” (59%), “Someone to go out with” (39%), and “Someone for tonight” (3%).” which formed an index based on factor analysis.
In terms of race differences, there were some:
(Note that the sample is about 70% male, so these are mainly the male values.) Even in a dataset of 60k+ people there aren’t that many Middle Easterners or Indians, so that’s why the error bars are large for them. The main groups of interest are the larger ones: the Whites (Europeans), Asians, Hispanics, and Blacks (African Americans). As can be seen by comparing the turquoise bars, many races were more into butts than Europeans were. The question is whether this can be explained by simple demographics, or other things we could measure. So we used regressions to try to figure that out:
(We combined boobs and butts to one category because the proportions were fairly small.)
Here “G” is general intelligence. We see that even adjusted for our somewhat poor measures of life history speed and intelligence, sexual orientation, and age, race keeps predicting preferences for boobs/butts over faces. Blacks, Hispanics, and Native Americans are more into these. The first two are not surprising, and I have no particular theory for why Native Americans are also into them. Maybe I haven’t watched enough of their dance shows to form a theory (commenters speculate away!). In retrospect, we should also have added mental illness as a predictor, as this can also be measured in this dataset (as we did in prior research).
Interesting is the very strong effect of intelligence in predicting not-preferring boobs/butts, beta = 0.35. Maybe that’s a signaling effect where smarter people know that it’s bad to signal low class on a dating site, or maybe it’s more internal. From the opposite perspective, we see that various things predict preferring faces: slower life history speed (gentlemen type behavior), and intelligence (beta 0.15). Hence the title.
Not shown in the table is a strong interaction between sex and intelligence. For men, being smarter makes you prefer faces, but among women there’s no detectable effect, probably because they already mostly prefer faces.
Also in the table we see some that unusual preferences, i.e., not preferring either faces or boobs/butts, but some other body part (hands? feet? shoulders?). This was largely predicted by odd sexual orientations as the people who are very into this were the bisexuals. Bisexuals are also the highest mental illness category and most criminal, so this finding isn’t so surprising. It would be better to have people rank a bunch of different body parts for their importance and use that.
Conclusion
We confirm that life history speed, race, and intelligence relates to preferred body parts in determining attractiveness. This is probably the first and only such study. Later this year, we hope to do another related study but with some better measures, so we can examine whether George Francis’ theory that preferring boobs over butts predicts intelligence. This question could not be properly analyzed with this current dataset because of the “choose 1” nature of the question. If you haven’t checked out George Francis’ theory and findings yet, be sure to do so now (Intelligence and the Evolution of Breasts).
