Similar faces may be related to shared group membership.
By November 2, 2017 - last reviewed on April 17, 2018published
Have you ever had the feeling, scrolling through your Facebook feed, that certain circles of college buddies, teammates, or neighborhood moms bear a resemblance? You may be onto something. Studies reported in Personality and Social Psychology Bulletin found evidence that people with similar facial appearances are more likely to be part of the same social group.
Researchers collected and tightly cropped headshots of people from different fraternities at a Midwestern university, groups of friends, and major-league baseball teams and enlisted volunteers to rate the photos on characteristics such as attractiveness, youthfulness, and perceived intelligence and strength. In certain studies, the team also used a computer program to estimate various physical dimensions of the faces.
Using both the subjective impressions and objective measurements, the researchers created statistical models that were able to sort the faces into their specific groups, fraternities, and teams with above-chance accuracy—even when they controlled for common gender and ethnicity. Since research suggests that similar people are drawn to each other, and because we may guess at how similar others are by looking at their faces, "we were basically saying that what a person looks like influences whether he or she gets into the group or not," says Ryerson University psychologist Eric Hehman, the study's lead author. "We have indirect evidence that this might be the case."
Most surprising were the facial similarities among members of particular baseball teams, Hehman says. The model based on facial measurements correctly sorted baseball players into the correct team—one of six—nearly half of the time. (The odds of making the associations by chance are about 17 percent.) In these groups, membership is decided by management, not players, and the logical criterion is a player's stats, not how he looks. Then again, Hehman says, research on elections and other domains shows that "appearance shapes a lot of outcomes that it shouldn't."