The Face Study - Additional Info & FAQ
I wanted to provide additional information about our facial attractiveness study since I've had a few questions & realized that the Gazette, Harvard Science, and Crimson articles only cover a small fraction of what we did in this study. There are some meaningful findings that I've explained here, and I responded to some inquiries. I think a Q & A format will work best to address a lot of what is not currently out there in the Harvard publications (and various other news sources). So, here goes:
Q. Why did you do this work?
A. I noticed there was a huge gap in the literature on facial attractiveness when it came to gays and lesbians. I thought this gap needed to be addressed. Also, I'm just interested in facial attractiveness. There's a lot we don't know... a lot we do know... and it's something that affects a lot of social behavior.
Q. What did you find?
A. The primary findings are best summarized in this chart that David put together. On the x axis (the horizontal axis) you can see the sex of the faces that were rated by the participants in the study. The Y axis (vertical axis) shows the proportion of masculine faces that were preferred by each group of raters, on average. We then used ANOVA and t-tests to determine if these differences were (a) significantly different from chance and (b) different from each other. Chance is 0.5. So, you can see that heterosexual women had no preference for sexually dimorphic OR non-sexually dimorphic faces. They were at chance (their error bar overlaps 0.5 - so that puts them at chance). When heterosexual women rated other women, however, it was a different story. They found feminine faces much more attractive. You can see how all the groups rated all the other groups in the chart below.
In addition to this, we found that men with higher degrees of sociosexuality (basically - people high in this trait are more open to one night stands, multiple partners, etc... you can google it and read more) were more likely to prefer sexually dimorphic faces.
Q. What do the findings mean?
A. Basically, men are men, regardless of sexual orientation. They want sexual dimorphism in a partner's face - gay or straight. Women, however, are more complicated. We know that birth control pills, ovulation, self-esteem, and numerous other factors can influence women's preferences. Men aren't as influenced by things like this (and they don't ovulate & normally don't take birth control pills). So, we had a clear explanation for men. However, the explanation for women still remains open to interpretation & we cautiously interpret the findings here because we used a random population sample of women. We didn't focus on women that were ovulating, those that had high self esteem, or those that were on the pill. So, when looking at the population at large, women seem to fall at chance levels. If we broke up the analysis - we would probably find other things. However, these findings have already been published by others so there was no use doing that.
Q. I don't agree with the study's findings. I like feminine (or masculine) faces & your study says I should like the opposite.
A. Our research looked at averages. No research looks at a single individual unless it is a case study (usually reserved for medicine or clinical psychology). We used basic statistical methods like t-tests and ANOVA to study if the average preferences of a group differed from chance. So, for example - did gay men like masculine faces, feminine faces, or were they at chance levels? It turned out they liked masculine faces and this was significant, based on the results of a t-test (which compares the distribution of scores made by gay men to random chance). Naturally, some gay men did like feminine faces. In fact, the overall average we found was only 0.6 - not 0.999. So, we are not saying all gay men like masculine faces, but statistically, on average - this is the case. Our t-test was significantly different from chance at the level of p < 0.001. Google "p values in statistics" to read more about what p < 0.001 means (it means that it is significant, but don't take my word for it).
Q. The faces you showed the participants in the research are ugly.
A. Participants indicated their preferences when looking at two faces that were digitally manipulated to be masculine and feminine. So, if they were looking at a face they found ugly, then they still chose the lesser of two evils. If you showed me a picture of two ugly couches, I could still tell you which is less ugly. We also didn't make the choice binary. We asked about the degree to which people preferred a masculine or feminine face. The presentation of images were counterbalanced to avoid bias and other experimental issues.
Q. I have a masculine/feminine face and now you're saying people are going to find me ugly?
A. No. Look at Brooke Shields' jaw angles. They're pretty masculine. Is she ugly? No. There are dozens (if not hundreds) of factors that make up an attractive face. We only studied one small thing here...just as a place to start. However, if you are very concerned about this, then you can always see a plastic surgeon. Although I am kidding...this IS actually done. This man had jaw implants.
Q. Why did you not study any other races but white people?
A. There are a number of reasons for this. (1) The central research question was not about race. It was about sexual dimorphism. There are only so many variables you can study simultaneously and still be able to analyze your data reliably. (2) We based our studies on previous literature using similar Caucasian faces. In order for this study to be accepted by the field, we had to use methods that were already proven and not introduce new phenomena. Our goal was to elucidate what gays & lesbians were attracted to when it comes to sexual dimorphism - not race. Surely, another study can look at race - but one study cannot cover everything. We couldn't study issues like symmetry and baby faces, etc...really, research like this moves along slowly - 1-2 variables at a time. (3) My research collaborators in Scotland invested a significant amount of time, money, and energy to take photographs of hundreds of locals. I honestly just didn't have the money to make my own facial stimuli. Using their faces was the most practical & proven method.
Q. Why did you not study bisexuals?
A. We only had 15 or so in the sample. Statistically speaking, drawing conclusions from 15 people would cause some serious problems. Furthermore, our focus of the study was gay men & lesbian women...not bisexuals. The literature on bisexuals is not well developed and therefore it is hard to make any robust study that focuses on them. There are a few papers on bisexuals out there which show some interesting findings - but they tend to conflict. Maybe in the future, someone can take this up for a study of their own.
Q. Was your sample representative?
A. Yes, we studied over 900 people recruited from a vast array of online websites and advertisements. Over 1000 studies in social psychology simply test undergraduates and draw conclusions about human behavior from that. So, our study was representative. We had people from (almost) all 50 states participate as well. (I can't remember the exact number right now...it was pretty diverse)
Q. Why does sexual dimorphism even matter?
A. In the introduction of our paper, we have covered this & also cited an extensive amount of previous research on the issue. Sexual dimorphism has been studied by dozens of researchers and very robust findings have come out about its significance. Here is an excerpt from our literature review on sexual dimorphism:
Studies of preferences for sexual dimorphism in faces among heterosexual raters suggest preferences evolved to maximize potential benefits of mate choices (for reviews, see Feinberg, 2008; Fink&Penton-Voak, 2002; Jones et al., 2008; Little& Perrett, 2002). Similar interpretations of preferences for sexual dimorphism in voices (for review, see Feinberg, 2008) and bodies (Little, Jones, & Burriss, 2007) have also been proposed. Masculine facial characteristics are positively associated with indices of male health (Rhodes,Chan, Zebrowitz,&Simmons, 2003; Thornhill & Gangestad, 2006) and dominance (Mueller & Mazur, 1996). While masculinity in male faces may cue dominance and health, masculine men are less likely to invest in offspring and relationships than are relatively feminine men (Boothroyd, Jones, Burt, DeBruine, &Perrett, 2008; Burnham et al., 2003; Gray, 2003; Gray et al., 2004; Gray, Kahlenberg, Barrett, Lipson, & Ellison, 2002). Moreover, masculine men are more interested in pursuing shortterm relationships than are relatively feminine men (Boothroyd et al., 2008; Rhodes, Simmons,&Peters, 2005). Female preferences for masculinity in male faces are modulated by a large number of factors, including menstrual cycle phase and state hormone levels (Frost, 1994; Johnston, Hagel, Franklin, Fink,&Grammer, 2001; Jones,Little et al., 2005; Penton-Voak et al., 1999; Welling et al., 2007), hormonal contraceptives (Cornwell et al., 2004; Feinberg, DeBruine, Jones, & Little, 2008; Little, Jones, Penton-Voak, Burt, & Perrett, 2002), body morphology (Penton-Voak et al., 2003; Scarbrough & Johnston, 2005), self-perceived attractiveness (Little, Burt, Penton-Voak,&Perrett, 2001; Little&Mannion, 2006), and sex drive (Welling, Jones, & DeBruine, 2008). Variations in female preferences for masculine men (i.e., those with masculine face and body shapes and those with low vocal frequencies) may function to maximize the benefits of female mate choices (e.g., by increasing offspring health or reducing mate search time) (Feinberg, 2008; Jones et al., 2008). Potential adaptations are also evident in male preferences for female facial femininity (Jones, DeBruine, Little, & Feinberg, 2007; Perrett et al., 1998; Welling et al., 2008). Facial femininity is positively associated with indices of female reproductive health (Law-Smith et al., 2006; Penton-Voak et al., 2003). Although male preferences for sexually dimorphic face cues are relatively stable in comparison to female preferences, men alter their face preferences based on social feedback (Jones, DeBruine, Little, Conway et al., 2007; Jones et al., 2008; Little, Burriss, Jones, DeBruine, & Caldwell, 2008). Individual differences in male preferences may maximize the benefits of choices by reducing mate search time or by increasing the efficiency with which men allocate their mating resources (Jones, DeBruine, Little, Conway et al., 2007; Jones et al., 2008; Little, Burriss et al., 2008).
Q. I thought attractiveness had to do with symmetry, average features, etc...
A. Yes - attractiveness has to do with a lot of things. Sexual dimorphism is just one small piece of a big puzzle. The effects we found do not explain all of facial attraction. They simply were a place to start. Future research can look at other topics.
Q. Aren't your findings kind of obvious?
A. It depends what you define as obvious. I think a lot of social psych research falls into that category. Sometimes, what we think is obvious has to be tested, empirically, to see if it is really "true" or just a hunch. A lot of social psychology research actually finds that our "obvious" inclinations turn out to be false. I've found when I tell people the findings, they say "duh - no kidding" - but if I ask them to predict in advance (by telling them the design of the study), they can't predict. So, I'm not sure how obvious it is. We still have no explanation as to why lesbians, on average, liked more masculine female faces - but heterosexual women were at chance. If gay men & straight men operate similarly, why don't lesbians and straight women operate similarly? I don't have an explanation really. This was not a finding I thought was obvious. I'm still confused by it.
Q. Who reviewed this study? What was the process to get it published?
A. Our study was put through peer review (twice) at the Journal - Archives of Sexual Behavior. Over 25 pages of reviewers comments were provided to us & we responded to all of them to the satisfaction of our peer reviewers & the editor of the journal. If we hadn't met the bar they set, we would not have been able to publish this study.
Q. How long did this study take? What was the process to do this work?
The study began with an idea I had in the summer of 2007. I took this idea to David Feinberg who was a post doc here at Harvard at the time. He lent me his expertise and his stimuli from his colleagues in Scotland (the faces) for the study. It took several months to collect data from over 900 people. It then took several more months to analyze over 100,000 cells of data in SPSS (we didn't simply look at facial preferences in this study). It took another few months to do a lit review & write up the paper for publication. The paper received a "revise and resubmit" from Archives of Sexual Behavior with an extensive critique. We responded to the critique and sent it back. They asked for more corrections, and we responded again. The paper was then published. From start to finish, it was a 2 year process.
Q. How has your study been received by the academic community?
A. When I presented it to the Department of Psychology at Harvard University in 5/2008, it was well received. Both of my advisors approved the study without requesting revision. I was personally very honored / speechless when Stephen Pinker himself told me that he was glad I did this work & was impressed by it. So, within academia (where people see & can understand everything that goes into work like this) - it is well received. In the media (where a 1/2 page summary tries to encapsulate 2 years of work) it's been sort of been overly susceptible to confusion.
Q. Are you continuing with this research? I am a student & I am interested to work with you in the future on similar studies. Are you hiring research assistants?
A. Unfortunately not. I already got my masters degree & am getting a PhD in organizational behavior now. However, my coauthors are highly respected, highly published experts in this field. I would recommend visiting their websites and checking out their work: David Feinberg, Ben Jones, Anthony Little Lisa DeBruine. Two of them also run the site www.faceresearch.org. n press). Sex-Dimorphic Face Shape Preference in Heterosexual and
Q. Where is the paper itself?
It's here. For now, I do not think I can provide the PDF here since the copyright is owned by the journal & they require a fee for each copy. However, you can see the first page. If you have university library access, then you can easily get the full article.
