"Gaydar," the ability to detect sexual orientation as if through radar, is real. And some people can guess with 80 percent accuracy, according to a new study.
In a study of college students at the University of Washington, most people were able to identify a person's sexuality with greater-than-chance accuracy based on black-and-white photo flashes of them for just 50 milliseconds—faster than the blink of an eye. Photos used in the study were digitally altered to remove hair, jewelry, and other "self-presentational" aspects of a person, suggesting there may be such a thing as a "gay face," according to researcher Joshua Tabak, a doctoral candidate at the university and lead author of the paper.
"The world's best lie detection experts top out at around 80 percent accuracy," he says. "Being at 80 percent accuracy on this type of judgment—just seeing a picture of a face in grayscale—that's pretty amazing."
But not everyone's "gaydar" is exactly on point. Some weren't able to guess sexuality at all, while others scored above 80 percent accuracy at identifying 96 randomly ordered gay and straight people. Overall, people were about 57 percent accurate at identifying the self-identified sexual orientation of subjects.
Although more studies are necessary, Tabak says his results suggest that guessing sexual preference is something that is likely learned via exposure to a wide variety of gay and straight people. Put in a real-world scenario, adding "hair, jewelry, clothing, gait, and posture," people might be even more accurate—or it might just contribute excess noise.
"The more you add, you might be adding conflicting information," he says. "Most likely, the differences in accuracy are driven by social factors. We can speculate that some people are more motivated to make these judgments or have more experience making these judgments."
Gaydar has been an oft-studied phenomenon, with several studies debunking it and several others suggesting that people couldn't tell sexual orientation based on body shape, walking style, or voice. Tabak says he believes it's a real thing—at least in a college population.
"I believe this is real in the population I'm testing," he says. "Grandma grew up in a different time, probably not knowing she was interacting with both gay and straight people. But if you're growing up now, you're exposed to a lot of different people."
His study suggests, but doesn't prove, that there may genetic facial components common in gay people. Surprisingly, students were more accurate at detecting lesbian women than they were at detecting gay men.
"With their prevalence in the media and the stereotypes, you'd think that since we have more exposure to the 'gay man,' we have reasons to believe we'd be more accurate in judging men," he says. "In fact, it was the opposite."
It also happens extremely fast—and perhaps unconsciously.
"It's judged so rapidly and efficiently," like other characteristics such as gender and race, he says. "It suggests that we may actually be judging sexual orientation without intending to in everyday life."
Tabak says he hopes his study makes people consider whether they may be discriminating against others without even realizing it.
"One primary argument against nondiscrimination protection, the stereotypical Republican response, is that if gay people kept it to themselves, it would be impossible to be discriminated against," he says. "Well, people might be making these judgments even when they're not trying to."
Jason Koebler is a science and technology reporter for U.S. News & World Report. You can follow him on Twitter or reach him at firstname.lastname@example.org