Facebook knows all about you, more than you probably know about yourself. A new study reveals that Facebook “Like” button data can reveal a surprising amount of information. Studying more than 58,000 volunteers who participated in the “myPersonality” app on Facebook, the research shows that the magical “Like” thumb can be used with over 80 percent accuracy to predict if you’re Democrat or Republican, your intelligence level, sexual orientation, personality traits, and if you drink or smoke on the weekends.
The Facebook data was compared to online personality tests taken by the volunteers. Compiling the Facebook data into an algorithm, researchers could predict whether someone was white or African American with 95 percent accuracy, whether he or she was gay with 88 percent accuracy, and which political party that individual leaned toward with 85 percent accuracy. Gender could be predicted 93 percent of the time, while age had 75 percent accuracy. People who used drugs could be anticipated with 65 percent accuracy, and identified as drinkers 70 percent of the time.
While some preferences were obvious, such as gay men liking the No H8 Campaign, others were a bit surprising. Gay men also tended to like “Oz the Musical,” but intelligent men tended to like curly fries and love thunderstorms. People who like Austin, Texas, tended to be drug users, as were people who related to the statement “Relationships Should Be Between Two People Not the Whole Universe.” Like sliding on the floor in your socks? You probably don’t own a bong. Straight men tend to be Wu-Tang Clan fans and wake up from naps feeling confused.
Naturally, computers are not prone to stereotypes, yet patterns tend to emerge. How each group relates to each “like” preference, remains to be seen, and could simply be a statistical coincidence.
“Although some of the Likes clearly relate to their predicted attribute, as in the case of No H8 Campaign and homosexuality, other pairs are more elusive; there is no obvious connection between Curly Fries and high intelligence,” the paper said.
Personality could also be predicted, and that could be a problem, though a fixable one. The study found that likes were connected with certain personality aspects in a similar ratio to the results of psychological tests. The study mentions one case where retailers used data to predict pregnancy, and in one instance, a pregnant teen was inundated with ads related to baby products though she had yet to tell her family, putting her in an awkward situation. Michal Kosinski, director of operations of the Psychometrics Centre at Cambridge University and lead author of the study, said that technology should provide some protections against privacy invasion.
“Our results [also] show that these predictions could be potentially very intrusive,” Kosinki stated in the study. “There are technical ways to make sure that individuals have full control over their data, and technology can be designed in such a way that data cannot be abused.”