This morning, you might have checked your Facebook account. According to StatisticBrain.com, over 1.7 billion people in the world have a Facebook account they log into at least monthly, so you’re certainly not alone.
While you’re logged into Facebook, you might come across a page you decide to like so that those updates show up in your news feed. Now let’s say one of your friends posted a meme you find amusing — so you like that as well. As you travel around the Facebook world, you leave a trail of likes. But did you know that trail is followed by big data companies, government agencies, and researchers? One researcher designed a way to know things about you, information that not even your spouse may know about you.
A doctoral student at Cambridge University by the name of Michal Kosinski devised a questionnaire app for Facebook called MyPersonality. The app offered Facebook users questions about their personality and was taken from a model called the Big Five, the definitive technique for psychometrics (the measure of psychological traits). Users who filled out the Facebook app were given a personality profile outlining their Big Five traits, such as openness, conscientiousness, or extroversion. The resulting profile was compared to the rest of their Facebook data: their likes, shares, posts, gender, location, and much more. The app went viral, and as Vice reports, “Suddenly, [Kosinski] owned the largest dataset combining psychometric scores with Facebook profiles” ever collected. Hmm…you could call it a cyber shrink.
Kosinski was sitting on a veritable data gold mine, and he quickly realized that the sample was so large, he could make extremely reliable deductions about Facebook users’ personalities simply based on what users were doing while logged into their accounts. Kosinski’s app could determine who was gay, who was married, who was single – even if the person had not stated so in their profile. Vice explains exactly how detailed the research models became:
In 2012, Kosinski proved that on the basis of an average of 68 Facebook “likes” by a user, it was possible to predict their skin color (with 95 percent accuracy), their sexual orientation (88 percent accuracy), and their affiliation to the Democratic or Republican party (85 percent). But it didn’t stop there. Intelligence, religious affiliation, as well as alcohol, cigarette and drug use, could all be determined. From the data it was even possible to deduce whether someone’s parents were divorced.
Eventually, Kosinski’s models were so accurate that ten Facebook likes could evaluate a person “better than the average work colleague.” Seventy likes were enough to know a person better than their friends. One hundred and fifty likes were sufficient to know them better than their parents, and three hundred gave them better knowledge than the target’s significant other.
Kosinski’s research has been used incessantly since then – Brexit, the 2016 Presidential election, and other events link back to Facebook data used as outcome predictors. Interestingly enough, evidence uncovered by VICE shows that Trump’s campaign hired a company named Cambridge Analytica – as did Ted Cruz. The company claims that they “were able to form a model to predict the personality of every single adult in the United States of America.” Its CEO, Alexander Nix, stated in a presentation at the Concordia Summit, that “pretty much every message that Trump put out was data-driven; some say, by Michel Kosinski’s Facebook research. The personality data models allowed the Trump campaign to tailor its message in ways, not even Facebook intended – and its users had no idea was happening.
So the next time you log into Facebook and click like, what will you be telling big data companies about you? Kosinski’s website offers users a chance to see right now what they already know about you. Interesting for sure – and a little frightening.