Imagine what kinds of personal traits might be associated with high intelligence. Chances are, having a taste for curly fries didn’t come to mind. According to a University of Cambridge study, however, liking the “Curly Fries” page on Facebook is a trend among those with high IQs.
The study published by Proceedings of the National Academy of Sciences analyzed the Facebook likes of over 58,000 volunteers to see how they correlated with various traits among Facebook’s user base. These traits include age, intelligence, race, sexual orientation, political views, emotional stability, satisfaction with life, drug and alcohol use, and even whether or not the user’s parents had separated by the time they turned 21.
Some of the correlations were fairly obvious; Liking Bill O’Reilly’s page was mostly associated with republican users, while the Joe Biden page was mostly liked by democrats. Other correlations were more ambiguous, however, such as “Curly Fries” being one of the top 5 liked pages among users with high intelligence.
In a TED Talk titled “The Curly Fry Conundrum”, computer scientist Jen Golbeck theorized how these kinds of associations exist. One theory, she says, is based on the concept of homophily. Homophily refers to the idea that individuals often associate with those who are most similar to themselves. In other words, “birds of a feather flock together.”
“If I were to give you a hypothesis, it would be that a smart guy started this page, or maybe one of the first people who liked it would have scored high on that test,” Golbeck said. “And they liked it, and their friends saw it- and by homophily, we know that he probably had smart friends- and so it spread to them, and some of them liked it, and they had smart friends, and so it spread to them, and so it propagated through the network to a host of smart people.”
While curly fries have no real connection to high intelligence, Golbeck says the spread of trends through homophily could have resulted in the average liker of the “Curly Fries” page having a high IQ. This is how researchers are able to make assumptions about a user based solely on their Facebook likes.
In fact, the findings of the University of Cambridge study were startlingly accurate. Conducted by Michal Kosinskia, David Stillwell, and Thore Graepel, the results show how the study was able to determine a user’s race in 95% of cases, and their sexual orientation in 88% of cases.
Of course, this method of data compiling could pose privacy issues. For instance, even if a person doesn’t like LGBT-related pages on Facebook, liking pages such as Kathy Griffin or Mac Cosmetics (some popular pages among homosexual males) can help researchers make conclusions about the person’s sexuality. In extreme cases, this could lead to a closeted gay person being outed by targeted ads and content.
One solution, Golbeck says, is to develop mechanisms that warn users of the risks of liking a page or even give users the option to make their information invisible to third parties. She says the same kind of research that is used to gather information can potentially be used to prevent information from being gathered.
“As a scientist, my goal is not to infer information about users,” Golbeck said. “It’s to improve the way people interact online.”