By Alex Fradera
Vacant job roles should be filled on the candidate’s skills, experience and knowledge, not their identity. But that means dodging our deeply held stereotypes, such as men being a natural fit for decision-making roles like management and women for care-giving professions. Evidence suggests this also applies to sexual orientation, meaning, for instance, that CVs that indicate the candidate is homosexual (by mentioning college experience in a group promoting gay rights, for example) are likely to be seen by recruiters as a better match for care-giving roles. New research from the Journal of Applied Psychology adds to this, suggesting that merely looking gay is enough for a candidate to be treated in a biased way by recruiters.
Nicholas Rule and his colleagues from universities at Toronto and Stanford conducted several studies, each involving between 68 and 201 participants. In every study, participants read about a target job, in most cases either a nurse or engineer, and then estimated the likely success of 90 different candidates, based only on a photo provided of each one. These cropped photos were originally taken from dating sites, ensuring that sexual orientation data was known by the researchers for each candidate.
Participants repeatedly predicted gay candidates would have more success applying to nursing positions than heterosexual candidates, and they predicted the reverse pattern for the engineer position. This effect was also found when participants were asked to say who they would prefer to teach their child maths (straight men preferred) or English (gay men). In another study, there was also a relative preference for gay men applying to paediatric roles over surgeon roles, although heterosexual men were predicted to have more success than gay men for both medical roles. Participants who had previous hiring experience were just as influenced by sexual orientation cues as other participants.
What’s going on? Previous evidence from this research group suggests that people can rapidly detect sexual orientation from viewing faces, and that this “gaydar” goes beyond interpreting stereotypical cues like clothing or gesture, reflecting instead our ability to read something more phenotypic, coded into the face or body. However recent research that used photos from a dating site to attempt to replicate Rule’s general gaydar effect revealed a potential confound: photo quality tended to be higher for men who were gay. If this was true of the photos used in the current research, it would muddy the researchers’ conclusions.
Even if photo quality happens to be what we use to judge sexual orientation (the alternative hypothesis that we simply don’t trust people who take good photos to be good engineers seems implausible), we should still take heed of these new findings, especially when other people’s prospects are in our hands. Thankfully, there was a suggestion that the bias was assailable, at least to some extent. When each photo was accompanied by a GPA score showing the candidate’s success in studying for the profession, a low-GPA score seemed to trump any effect of sexuality cues in facial appearance.
As real life job picks always involve considering some relevant information, this is somewhat reassuring; however Rule’s team thought a single relevant score may be too obvious a guide for decision making, and that the effect of sexuality cues in facial appearance would hold their own when accompanied by more complex (and representative) supporting information, as would be found in real-life contexts. They conducted a further study that supported this view, suggesting facial cues to sexuality are likely to persist in spite of more relevant information.
However, I wasn’t entirely comfortable with the methodological design of this particular study, so I’m skeptical that sexuality cues would have such strong influence. For instance, it used photos (and profiles) procured from LinkedIn rather than from a dating site, and the researchers presumed sexual orientation based on the person’s workplace – LGBT organisation or otherwise – which seems a noisy and inaccurate criterion to use.
Another of the new studies provided more reason for hope. It showed that simply asking participants to be fair and objective in their judgments stopped them from favouring heterosexual men for the engineering role, although their bias for gay men for nursing positions remained. This suggests diligence on the part of recruiters could help correct this ‘insidious’ tendency.