Women in science have plenty of anecdotal evidence that looking feminine and “looking like a scientist” at times seem mutually exclusive. They’ve revealed their professions at parties to be met with surprise been asked where the professor is in their own university offices thought maybe they shouldn’t wear a dress to that conference. Now there’s scientific evidence as well. A recent paper in the journal Sex Roles reveals that people viewing photos of real researchers are more likely to associate women with non-science occupations if they appear more feminine. Lead author Sarah Banchefsky of the University of Colorado Boulder tells us more.
Sarah Banchefsky: There is a lot of anecdotal and qualitative work out there showing that women in STEM feel pressure or have been explicitly told to stop wearing make-up bright colors and dresses—basically to assimilate—if they want to be taken seriously in their careers. One study found that the majority of men and women in STEM report that the tension between femininity and science is a serious problem for women in the field. That said we wanted to perform a rigorous experimental study to see whether it was really the case that feminine female scientists (but not feminine male scientists) are viewed as less likely to be scientists. This phenomenon had not been scientifically established. A big plus of our studies is that they ruled out the possibility of actual differences in scientific potential of the people being judged—all targets were all qualified top scientists so it’s clearly not that the feminine women just aren’t as scientifically capable. We also instantiated femininity in a compellingly modest way: through naturalistic variations in facial appearance—not for example by presenting them in high-heels or with obvious make-up or hairstyling. This means this finding likely generalizes to everyone not just hyper-feminine women but those who are very subtly and naturalistically more feminine.
In the first study each participant evaluated 80 photographs (40 men 40 women) of real scientists at top universities in the United States. Participants were not aware of this and instead thought they were doing a study on the accuracy of first impressions. They rated various aspects of each person s appearance including masculinity-feminininty attractiveness and also how likely each person was to be a scientist or an early childhood educator. The second study was similar but made several design alterations to strengthen our conclusions. First we removed all appearance ratings so participants were not explicitly asked to consider masculinity-femininity. Thus to the extent we saw career judgments in in the second study relate to masculinity-femininity ratings made in the first this indicates that people were using the same appearance cue to make career judgments (femininity) without being asked to consider the target’s appearance. We also added an additional career option journalist to ensure participants didn’t think well I said they were likely to be a scientist so I’d better say they’re unlikely to be an early childhood educator . Finally we presented faces either blocked by gender or intermixed randomly by gender. To still see the same effects when faces are presented intermixed by gender means the impact of feminine appearance on career judgments is strong enough to influence judgments even when the most obvious difference between the faces—their gender—is being highlighted.