Can Twitter tell us when our relationships are falling apart? According to a forthcoming paper from an international team of researchers, it looks like social media might know a relationship is over before either party is ready to admit it.
The study: Using public data, researchers Kiran Garimella, Ingmar Weber and Sonia Dal Cin tracked the relationship status of 661 couples on Twitter who broke up over the study period from Nov. 27, 2013 to April 27, 2014. The couples were identified by whether or not they mentioned each other in their profile info and then removed it, to remove any ambiguity. Twitter was able to tell just how long their relationships lasted based on their messages:
More than 3,200 pre-relationship tweets were included in their analysis to provide additional information. The 661 breakups were compared to Twitter's general population, and the researchers were able to identify several standout characteristics of the unhappy former couples.
Among the findings:
- Couples approaching a breakup sent tweets to each other less often and contacted other users more frequently.
- Breakups were followed by "batch unfriending," in which users tended to lose 15 to 20 followers shortly after the deed was done.
- "Stonewalling" or "communication asymmetries," known to the rest of us as getting the cold shoulder, was observed significantly more often when a couple was approaching a breakup.
- Twitter users who had just gone through a breakup used more "depressed" terminology in their tweets immediately after, with this behavior being more prevalent among jilted lovers than those who initiated a separation.
- Ultimately, couples who had communicated often on Twitter before the observed breakup tended to interact more often after, indicating that they remained friends.
The researchers also assembled word clouds of pre- and post-breakup messages between users whose love lives went sour. They pretty much speak for themselves.
Even the researchers were taken aback by this massive shift in tone, which they called a "surprising amount of public fighting and insulting happening after a breakup." Just look at all those swear words.
So what does this tell us? The team concluded that the wealth of information available on Twitter just concerning breakups indicated that the service could present "huge potential" for "studying sociological and psychological processes." Basically, Twitter acted as a massive public archive of the circumstances and characteristics of the various breakups, and the research team was able to obtain all of it rather easily.
Other social media companies have been developing advanced analytics related to their members' love lives for quite some time — like Facebook, which is likely now capable of predicting relationships and breakups based on usage patterns, such as how often couples posted on each other's walls. As Facebook data scientist Carlos Diuk writes:
During the 100 days before the relationship starts, we observe a slow but steady increase in the number of timeline posts shared between the future couple. When the relationship starts ("Day 0"), posts begin to decrease. We observe a peak of 1.67 posts per day 12 days before the relationship begins, and a lowest point of 1.53 posts per day 85 days into the relationship. Presumably, couples decide to spend more time together, courtship is off and online interactions give way to more interactions in the physical world.
Facebook users also go through a massive spike in their time on the site during and after a breakup, which the Facebook Data Science team attributes to "people receiving support [from] their friends in times where they need it, whether it comes in the form of private messages, timeline posts or comments."
But Twitter is, for the most part, public, and airing your dirty laundry on it could help ensure that one of your tweets ends up in a similar word cloud. Given how easily accessible this information is to advertisers, employers, friends, family, intrepid data scientists and the authorities, it might be better to nurse that broken heart in a more private setting.