Social media study shows how we choose friends
Examination of MySpace profiles supports alliance hypothesis for friendship
It's like Confucius says: Have no friends not equal to yourself.
Professor Peter DeScioli's research on the online friendships of 11 million people shows we may be taking the ancient Chinese proverb literally.
With the help of web crawlers, which collected data from profiles on MySpace, the social networking site, that used the "Top Friend" application to rank their pals, DeScioli found that people are most likely to consider as best friends those who reciprocate the feeling.
It supports what is called the "alliance hypothesis for friendship." In contrast to the long-held belief that friendship was a trade relationship, one based on exchanging favors, the alliance hypothesis suggests we depend on our friends in conflicts. The study's findings were recently published in the journal Perspectives on Psychological Science.
"It's crude," says DeScioli, a Florence Levy Kay Fellow in psychology and behavioral economics. "We think we care about our friends as people, not because we're getting something."
In part, we're right. To do well in a trade relationship, humans would have to be good accountants, but we're not, according to DeScioli. Instead, we look for people who will side with us - and against others - in a dispute, which also offers a good explanation for jealousy. The fewer people ranked higher than you, the greater the chance that said friend will side with you.
"Other animals generally dispute one-on-one, but humans are different. ‘Can you believe what so-and-so did to me?'" DeScioli says, mimicking a common conversation between friends. "We're good at recruiting allies."
Alliances are cultivated in advance so that when the conflict arises, we can predict our supporters with ease and accuracy. The same holds true on a larger scale. Nations aren't just concerned about their direct relationship with an ally, but also with whichever other nations it supports.
"It's not conscious, you generally just feel closer to those people," DeScioli says.
Conducted with Robert Kurzban, associate professor of psychology at the University of Pennsylvania, the study examined the age, sex and location of each person using the application to rank their friends, as well as each person's overall popularity - in other words, how often others ranked them as a top friend.
"This is the largest data set on friendship I know of that's ever been collected," DeScioli says.
The results show your ranking of a friend depends on how they rank you. Comparing first- and second-rank friends, 69 percent designated as best friend the person who ranked them higher.
That correlation proved to be a much stronger predictor than geographic proximity, which is "the [theory] everybody talks about, the one in text books," DeScioli says.
Additional research in this area, such as manipulating friendships in a lab environment instead of collecting data on online friendships, might help people understand how to deal with relational aggression, or social bullying like gossiping or spreading rumors.
The results of the study "makes you realize that sometimes we're more jealous than we need to be," DeScioli says. "Sometimes you just need to reassure a friend, and it's helpful to know that it's not just totally arbitrary craziness."