×

Dartmouth Study: Friends Think Alike



Valley News Staff Writer
Tuesday, February 13, 2018

Hanover — Researchers at Dartmouth College have applied hard data to once-philosophical questions about the nature of friendship, using brain scans to track how friends think alike.

Using magnetic resonance imaging technology, a trio of scientists recently mapped the social networks and neural responses of an incoming class of students at the Tuck School of Business. The results, published in Nature Communications last month, show that the closer people are socially connected, the more they think alike.

“We thought this would be a great tool to see if people really do see the world more similarly when they’re friends,” said Thalia Wheatley, an associate professor of psychology who co-authored the study with Adam Kleinbaum, an associate professor of business at Tuck, and Carolyn Parkinson, an assistant professor of social psychology at University of California Los Angeles who recently completed her graduate work at Dartmouth.

Wheatley said she and her collaborators took inspiration from Princeton University’s Uri Hasson, whose lab recently demonstrated that brain scans of storytellers and their audiences are remarkably similar while the story is unfolding.

For this latest paper, the Dartmouth researchers in 2016 conducted a social network survey of roughly 280 first-year Tuck students, asking them to list their friends and then drawing a map of interconnected dots, where each point was a person and each connecting line a mutually expressed friendship.

Then, the scientists took a smaller cohort of about 42 participants and showed them a variety of videos, featuring such subjects as politics, science, comedy and music, while using an MRI to track which areas of their brains lit up and when.

The results showed that similarity in people’s thought processes is proportional to their social distance.

Not only were the scans most similar between people who identified each other as friends, but the association also decreased in a proportional way when compared to friends of friends, and even friends of friends of friends.

Those findings remained true even after the authors controlled their test group for other attributes that can predict social connection, including age, race and gender.

The finding has implications in countless circumstances, including in the corporate realm, where Adam Kleinbaum, a co-author and an associate professor at Tuck, studies social connections in business.

Kleinbaum said the research he did in offices was “less intrusive” than a brain scan and more often involved personality and social network surveys.

Still, he said, mapping social interactions can be revealing, because it determines who people go to for help and how they make decisions about who to collaborate with.

Another realm that brain scans could touch — someday — is romance. If the MRI can predict friendship compatibility, why not use it to find a mate?

“We’ve sort of joked around that this could be the next big dating app,” Kleinbaum said — although, he noted, taking the idea seriously for a moment, MRI data is expensive to collect. It might cost $1,000 to have your brain scanned and compared to a database of potential partners.

This latest study is only the jumping-off point for further inquiry, the researchers said. Brain scans of people in social networks, for instance, don’t convey by themselves the substance of what people are thinking.

“What we don’t know, for example, is if two people are responding to the same slapstick comedy is whether they’re both enjoying it or they both think it’s stupid,” Wheatley said. “What we do know is they both see it in the same way.”

Another avenue for exploration, one that the academics say they’re likely to pursue soon, is to track how neural responses among friends change over time.

This study was “cross-sectional,” which means it measured how people’s brainwaves looked at one moment in time. If another study were to track the changes, the researchers might be able to answer even deeper questions.

“Do people come in and they see the world together and that makes them friends, or is it (that) you come together and through your shared experience you come to see the world a similar way? Or is it both?” Wheatley said.

Wheatley and her collaborators said they may soon return to Tuck to follow a first year class as its members grow to know each other and form friendships, all the while taking scans to determine what, if anything is changing.

Their instinct is that both assumptions are true: people who think alike become friends, and friends tend to grow closer in their thinking.

Kleinbaum said he was working on a related study, which hasn’t yet been published, that uses computational linguistics to show how people who are friends use similar language. His research supported the intuition, he said: people’s use of language grew more similar as friendships formed, and friendships, in turn, tended to form between people already alike in their speech patterns.

Dartmouth’s geographic isolation and social insularity make it a particularly productive environment to study social connections, Wheatley said. Whereas people studying at city universities often branch out and find friends in many different places, members of Dartmouth classes tend to stick together, making their networks easier to map.

Wheatley also noted that some people worried that this study had delivered a “depressive finding” by highlighting an underlying neural basis for groupthink and viewpoint polarization — a touchy subject in today’s politics.

If the next study proves that these patterns run in both directions, then that might add a little hope to the diagnosis, Wheatley said.

“If I’m assigned a roommate in college that thinks differently from the way I do, then maybe I’ll come to see the world a little more like them,” she said.

Rob Wolfe can be reached at rwolfe@vnews.com or at 603-727-3242.