Socioeconomic disadvantage and psychotherapy
Kids With High IQs Live Longer
Älykäs pysyy terveenä – ja elää muita pitempään
Breeders: Daddy Makeover: Parenthood can change your personality. How having a child changes everything.
Type A Personalities Have the Edge in Procreating
‘Women are getting more beautiful’ – Getting the story right
Having your study publicized by the media is nice. Having your study misrepresented and misinterpreted in the process is not. The media coverage of my paper on physical attractiveness and having children had a bad start and even worse follow-up. The origin of the problem: Times Online news article sexing up the finding a bit too much (I wasn’t interviewed for this article at all and heard about it only after it had been published). Then things got worse with other journalists copying & slightly modifying the Times Online piece. Naturally, things were further muddled by the If-I-were-a-movie-critic-I-would-rate-movies-without-seeing-them-and-just-by-relying-on-discussions-overheard-in-a-pub columnists, the I-haven’t-read-the-paper-but-here’s-my-take-on-it-anyway bloggers and the ever so alert This-research-is-nonsense-I-want-my-tax-money-back-even-if-the-research-was-not-funded-by-my-tax-money readers.
Here are some clarifications and corrections to the press coverage. If you feel too exhausted reading it all, just try to focus on the words printed in bold. The original article can be found here.
1) The main point of the study was to see whether attractiveness predicts fertility in a contemporary American population, not whether people are becoming more or less attractive over time. The evolutionary extrapolation that people are becoming more attractive is just that – an extrapolation that depends on assumptions not tested in the study, e.g., whether the association holds over time. Basically, the argument is clear: given that attractiveness is partly heritable and it is associated with reproductive differentials, the mechanism of sexual selection is expected to increase the mean of the trait over generations. This possibility was only briefly mentioned in the article:
“The results indicated a weak positive directional selection gradient for attractiveness in women (β=0.06) and men (β=0.07). This is slightly less than half the median gradient (β=0.16) observed in a review of natural selection in 63 studies of nonhuman animals (Kingsolver et al., 2001). If the estimated gradient of ~0.065 were to hold across time and environments, and assuming a heritability of 0.60 for physical attractiveness (McGovern, Neale, & Kendler, 1996) and no genetic correlations with other traits, one would expect attractiveness to increase by ~0.02 standard deviations per generation (R=0.602×0.065).”
So how much is 0.02 standard deviations? Perhaps it helps to consider the effect magnitude in another context: if natural selection were to favor tallness with the same strength as observed for attractiveness here, height would increase by ~0.20cm (or 0.08 inches) per one generation. Such a slow process would be observed only over several generations, say, at least 5-10 generations to get an observable effect. In other words, the finding says nothing of comparison of people’s attractiveness in the 1950s vs. the 1970s vs. the 2000s!
2) The association between attractiveness and fertility was observed in women and men. Most stories have mentioned only women or have even claimed that there wasn’t an association in men (“Women more beautiful but men remain Neanderthal, study finds”). This is incorrect! Put simply, the main difference in the number of children in women was between those above average (50%) vs. those below average (50%) attractiveness, the former having more children than the latter. In men, the main difference was between the least attractive men (25%) vs. men who were at least moderately attractive (75%), the former having less children than the latter, but men’s attractiveness made no difference among those who were at least moderately attractive. Insofar as there is natural selection for attractiveness, this applies to men as well as to women.
This is what the study synopsis says
“In women, attractiveness predicted higher reproductive success in a nonlinear fashion, so that attractive (second highest quartile) women had 16% and very attractive (highest quartile) women 6% more children than their less attractive counterparts. In men, there was a threshold effect so that men in the lowest attractiveness quartile had 13% fewer children than others who did not differ from each other in the average number of children. […] A linear regression analysis indicated relatively weak directional selection gradient for attractiveness (β=0.06 in women, β=0.07 in men).”
Notice, too, that it was the “second-most” beautiful women – not the most beautiful women – who had the greatest number of children.
3) According to some news reports, my study found that attractive women were more likely to have daughters than sons. This is not true! TimesOnline presented the earlier work of Satoshi Kanazawa together with my study and it was Kanazawa’s study which found evidence for such an association. The claim that my study builds on Kanazawa’s earlier study is equally false! I included offspring sex as a secondary outcome of interest, because Kanazawa’s initial finding has been criticized on methodological grounds. There was only tentative evidence for Kanazawa’s hypothesis in my results, as the association was observed only in men and was statistically weak. This is how the finding is reported in my article:
“Analysis of offspring sex ratio suggested a tendency for more attractive men to have more daughters than sons, but this association was not statistically significant. […] In light of random variation observed in other studies of offspring sex ratios (see Gelman & Weakliem, 2007; Palmer, 2000), the potential attractiveness–offspring sex ratio association is best interpreted cautiously before more data are available.”
And for those who are still confused: No, I am not Satoshi Kanazawa and this study was not carried out by Kanazawa. So please, do not criticize my work because Kanazawa’s work has been criticized. And please, do not refer to me as the “Ann Coulter-loving scientist”, I hadn’t even heard about the lady before the headline.
4) Several commentators have dismissed the study because they think that the standards of beauty vary so much over time and across cultures. Yes, there is evidence that some measures of attractiveness vary across cultures. For instance, men seem to prefer heavier women in societies where food is scarce and fat/high BMI is an indicator of health/physical well-being. However, there is substantial agreement between individuals and between different cultures on some aspects of attractiveness, e.g., facial attractiveness (which was the measure of attractiveness in the present study). For instance, two reviews of attractiveness research reach the following conclusions:
“Contrary to conventional wisdom, there is strong agreement both within and across cultures about who is and who is not attractive. […] Beauty is more than just in the eye of the beholder” Langlois, Psychological Bulletin, 2000, vol. 126, 390-423. “A long-held view in the social sciences is that standards of beauty are arbitrary cultural conventions (Berry 2000, Etcoff 1999). Even Darwin favored this view after observing large cultural differences in beautification practices (Darwin 1998/1874). However, two observations suggest that some preferences may be part of our biological, rather than our cultural, heritage. First, people in different cultures generally agree on which faces are attractive (Cunningham et al. 1995; Langlois et al. 2000; Perrett et al. 1994, 1998; Rhodes et al. 2001b, 2002; but see Jones & Hill 1993 for weaker agreement). Second, preferences emerge early in development, before cultural standards of beauty are likely to be assimilated (Geldart et al. 1999; Langlois et al. 1987, 1991; Rubenstein et al. 1999, 2002; Samuels et al. 1994; Samuels & Ewy 1985; Slater et al. 1998, 2000).” Rhodes, Annual Review of Psychology, 2006, vol. 57, 199-226.
These findings clearly rebut the argument that attractiveness cannot be assessed reliably. In the article, I do acknowledge the possibility that the attractiveness-fertility association changes over time:
“It is even possible that the attractiveness–fertility association has changed in the United States since the 1960s and 1970s as a consequence of changes in rates of marriage, divorce and fertility. […] Humans living in modern industrialized societies have ample opportunities to enhance their looks with readily available make-up products and exercise programs, so it is possible that physical attractiveness no longer functions as a reliable cue of individuals’ health status.”
There were some other, more minor misconceptions presented in the news articles but I will not bother to comment them all. On the other hand, there were also some news articles that were quite accurate in reporting the finding. It is worth adding that the few people who actually read my article were quite quick to notice that it did not make the claims that were circulated around in the media. I got several e-mails from these people because they were wondering if they had somehow misunderstood the original article!
On the more amusing side, the media flurry did have one funny unintended consequence. The Fox News covered the story by telling the viewers that evolution is driving women to become even more beautiful. A note to future historians: when tracing back the turning point at which conservatives begun to believe in the theory of evolution, please cite my article.
Oh well, my article had its 15 minutes of fame. Not the fame it would have deserved, but anyway. My guess is that nobody remembers the article anymore; it has already been 7 days since the Times Online report. I don’t believe this commentary will be read even by 1% of those who have read the news articles of Times Online and others. It is merely damage control targeted to people who come across my study in blog & newspaper archives sometime in the future. At the moment, there seems to be more pressing news in the horizon, reported with the accuracy and moderation so characteristic to professional journalism:
MJ, 3rd August, 2009
PS. You can leave your comments here.
The original article: Jokela M. Physical attractiveness and reproductive success in humans: Evidence from the late 20th century United States. Evolution and Human Behavior.
Some links to news & blog coverage
The science of sexism (why science of sexism when even the blogger seems to realize that it’s not about the study but how it has been presented in the news?)