Facebook can predict if romantic relationships will last
Believe it or not, Facebook can predict with astonishing precision the status of your love relationship, as well as if you're going to get dumped, according to a new study from Cornell University.
The study, conducted in conjunction with Facebook, was able to make these predictions at any given time based solely on your social connections, measuring how many mutual friends you and your significant other have in common on the social network.
However, there’s more to determining the status of a couple’s love relationship than the number of mutual friends they have. It also involves using a new network measure that researchers refer to as “dispersion”, which can correctly predict who you’re dating up to 50 percent of the time – and up to 60 percent of the time if you’re married.
Think of "dispersion" as a way to measure how many mutual friends a couple have in common, as well as how many of those mutual friends are connected to each other as a result of being linked to you or your loved one's social network on Facebook.
Given that networks link mutual friends together, a romantic couple with low dispersion is typically a positive sign that not only reveals the couple has a lot of mutual friends, but that these common friends know each other too.
In this regard, couples in a relationship serve as a sort of social bridge that connects their individual networks together; thus, making it easier to introduce each romantic partner’s Facebook friends to the other, similar to how you might bring some of your friends to a get-together, and your significant other brings some of their own so everyone can get to know one another.
For the study, the researchers developed their own measure of dispersion, examining 1.3-million Facebook user profiles to detect not only the number of mutual friends a couple had, but to also look at their individual network structures. From there, they were able to make several observations.
If the researchers found a link between two people who did not have a lot of mutual friends in common, the couple was referred to as having “high dispersion”.
In a relationship with high dispersion, a man would be connected to all of his girlfriend’s closest friends, and her to his. However, those friends would not be friends with one another.
In comparison, a relationship with low dispersion would have more mutual friends connected to one another. After all, when someone is interested in you, they are more likely to share their Facebook contacts from a wide range of social circles, including friends, family, co-workers, etc.
In another fascinating finding, the researchers also managed to use their dispersion methodology to determine the health of a relationship and whether it was strong enough for the long haul. They did this by looking at how couples declared their official relationship status on Facebook.
Using an algorithm to measure the amount of dispersion between two people who declared they were in an official relationship on Facebook, the researchers found that the greater the dispersion, the greater the likelihood that the relationship would last at least 60 more days.
In contrast, when the algorithm failed, that was seen as a possible sign that the relationship may be doomed.
Indeed, the study showed that when couples declared themselves to be "in a relationship" on Facebook, but had low dispersion on the social networking site, chances were 50 percent higher that the couple would be switching their relationship status from “in a relationship” to “single” within 60 days, compared with couples with a high dispersion rate.
The researchers concluded that dispersion is “a structural means of capturing the notion” that some friends come and go, while others last through “multiple life stages, or because they have been systematically introduced into multiple social circles."
SOURCE: Cornell University and Facebook, Romantic Partnerships and the Dispersion of Social Ties: A Network Analysis of Relationship Status on Facebook, by Jon Kleinberg and Lars Backstrom, published October 24, 2013.