Read: Has college gotten too easy?
Popular kids, of course, are not the only groups that middle and high schoolers look to when deciding how to be and what to do—they also take cues from their parents, their friends, the media, and so on. Prinstein and his research collaborators looked at how much of an influence these various constituencies held, and, he said, “what we find is that the popular kids aren’t the strongest influencers—it’s actually, really interestingly, that kids’ own attitudes and behaviors tend to fall almost exactly in the middle of what their parents tell them to do and what their best friends tell them to do.” He says that when he gives lectures to crowds of parents, he’ll often tell them, “You’ve got 50 percent of influence to use every day.”
Close friends, then, seem to be a more powerful influence than popular kids. But one recently published study arrived at a finding that adds some nuance to this: The research, by Tim Malacarne, currently a visiting professor of data science and sociology at Mount Holyoke College, indicated that the behaviors of students at the center of a high-school social network were no more influential than those of a randomly chosen peer when it came to predicting how likely any given student was to drink alcohol or smoke cigarettes.
Malacarne, though, was using a different definition of popularity—one based on who was friends with the people who had the most friends—because the data set he was using didn’t record students’ subjective assessments of who in their grade was popular. His statistical model of popularity was an attempt to approximate those subjective assessments, though it more directly measures social connectedness than “coolness” or “popularity.”
Interestingly, Malacarne found that the habits of the most well-connected people in that model seemed not to have an outsize impact on what their peers did. Perhaps instead, as Malacarne speculates in the paper, “central individuals are influential, but as judges of behavior rather than models of it.” He also wonders, he told me, if his analytical method would detect different patterns when it comes to behaviors not accounted for in the data set he was working with—say, what clothes kids choose to wear.
But social connectedness is not necessarily the same thing as popularity, and Prinstein pointed out to me that it matters a lot how a researcher defines who’s popular. “Friendship reflects a close, dyadic relationship that includes companionship, emotional intimacy, et cetera,” he said. “Popularity is not a relationship-based construct—it is reputation-based. You can be very popular and have only a few people believe they are friends with you.”
Usually, the two main determinants of “reputation-based” popularity in high school, according to Prinstein, are aggressiveness (“unfortunately, in order to make [themselves] seem high on the totem pole, a lot of kids and adults sometimes try and belittle others”) and physical attractiveness. But he’s also found that a community’s own value system matters a lot. “In communities where it’s very religious, then maybe your parents’ level of status at the church, for instance, might be an important determinant [of popularity] … Whereas it might be the kid who is really high-achieving in a very academically oriented community,” he said.