Against all predictions, Donald Trump won the election. Was it a collective desire for reform from the American people, or a company preying on the fears of the undecided? Two researchers from Zurich-based Das Magazin went digging, and their findings are eye-opening.
Patty is 25 years old, and according to Facebook and her search history, she likes horses, mountain biking, Glee, Quinao, and works retail at a clothing store. From this information, I insinuate that she is liberal and extroverted.
What I just did is called a prediction. An interpretation of information, or data, can result in a deeper understanding of the subject. When there is enough data to put into a system to build universal profiles, ie when we know what kind of people like what Patty likes, we can reliably extrapolate information about Patty. This is called psychographics.
Psychographics and Big Data
In 2008, a man by the name of Michael Kosinski was accepted into Cambridge University to complete his PhD, in the esoteric field of psychographics. He began to study how personality, traits and characteristics can be found through data.
Before I get too into the psychological aspects of big data, let me explain big data. I’m sure you’ve heard of it and probably know what it is, but just to clarify, it is everything that has ever been done on the internet. Everything you have ever done on the internet leaves a trace, and that is your data. Big data is everyone’s data merged together, and it is analyzed to reveal trends, patterns and behaviors.
Back in the days before (rodeo) the internet, in the 1980s, two teams of psychologists developed a scale for measuring personality. They created 5 basic categories to evaluate. Openness (How open are you to doing new things?), Conscientiousness (How perfect do you want things to be?), Extroversion (How outgoing are you?), Agreeableness (Are you hard-headed or go with the flow?), and Neuroticism (How easily irritated are you?). They named this the OCEAN model.
Unfortunately, in the 80s, they had minimal data to work with, because it required people to fill out lengthy, detailed questionnaires. Then came their long-awaited, digital savior; the internet.
Tracking Your Every Move
Back to Kosinki and his PhD. He took the OCEAN model and applied it to the internet. In the first decade of the 2000s, he developed a method of organizing people into categories based on their Facebook likes. They began to find reliable information from the aggregation of millions of people’s data, creating a master outline for who is what type of person. They found predictable results like men who “like” MAC cosmetics are likely to be gay, and people who “like” philosophy are likely to be introverted. One data point is undescriptive, but millions of data points allow the researchers to connect the dots.
By 2012, with 68 Facebook “likes”, he and his team could predict the skin color of the liker with 95% accuracy, and political party affiliation with 85% accuracy. From only 10 “likes”, he knows more about you than your colleagues, and with 300 “likes”, he knows more about you than your partner! He concluded that our smartphones are vast psychological questionnaires that we constantly (consciously and unconsciously) fill out.
Predicting Your Next Move
As he began to unravel the inner workings of the internet, he realized how dangerous this game was. He was making headway in psychological profiling, but figured that people could easily be manipulated if companies found out who they are.
A key idea from his research is that data can be used backwards. He found that data can create psychological profiles, and inversely, data can be used to predict certain types of people. With enough reserach, one could create a “like” schema for undecided democrats, for example. In the United States, you have to “opt-out” of public data. It’s public until made private. All of those “OKs” you click for Facebook apps, or Terms of Services you agree to, or Google searches you make are building your public, online psychological profile, and contributing to the mass of data that is used to identify people.
This sounds like a marketer’s dream. They can find exactly who their target customer is, not only hypothetically, but literally have their name, and create a personalized ad for them.
So came along the company Strategic Communications Laboratories (SCL), a marketing company based on psychological modeling. One of their employees, and fellow psychologist at Cambridge University, approached Kosinski, with interest in using his method of data collection and interpretation. Kosinski smelled trouble and cut off connection with his colleague. Meanwhile, SCL was hired by the leading force behind the Brexit movement, leave.eu…
Then, SCL created a subsidiary, Cambridge Analytica, to focus on United States elections.
Cambridge Analytica and Donald Trump
On September 19, 2016, Cambridge Analytica CEO Alexander James Ashburner Nix gave a speech at The Concordia Summit, a type of seminar for leading economists and businesses. He explained Cambridge Analytica’s methods, and claimed their marketing is based on a combination of three elements: behavioral science using the OCEAN Model, Big Data analysis, and ad targeting.
He states that campaigners have been focused on demographics. “A really ridiculous idea,” he says. “The idea that all women should receive the same message because of their gender—or all African Americans because of their race.” Through his detailed investigations and new approach, “we were able to form a model to predict the personality of every single adult in the United States of America.”
Cambridge Analytica first got involved with Republicans Ted Cruz and Ben Carson. With their growth in popularity, Donald Trump took notice. He hired the company, and in mid-August, tweeted “Soon you’ll call me the new Mr. Brexit!” Now we know why.
“Pretty much every message that Trump put out was data-driven,” says Alexander Nix. They participated in what is known as dark posts, which are advertisements that can only be seen by their intended viewers.
Throughout the election, it seemed like Trump was ambivalent, uncertain of his message. Wavering between opinions and contradicting himself, from the outside it looked like he was lost. But really, he was appealing to everyone. Say one thing to get one vote, something else to get someone else’s vote. And you have both people’s votes.
Everyone’s different, a computer can’t decide who you are, unless you’re a sheep. Stay true to yourself, be aware of who and what may be (unconsciously) influencing you.
Earlier versions of story appeared in Das Magazin and Vice Motherboard
Photos courtesy of National Review,
The post Psychograhpics: The Science Behind the Election appeared first on Verge Campus.