Technology & AI

The Science Behind Cambridge Analytica: Does Psychological Profiling Work?

The researchers who warned about abuses of Facebook data show how psychological profiling gets results.

April 12, 2018

| by Edmund L. Andrews

 

Image
A dazed man with an overlay of symbols of eyes, shopping carts, computers and other marketing symbols | Photo Illustration by Tricia Seibold with iStock/izusek and iStock/axel2001

Online ads are measurably more persuasive when they target a user’s psychological traits. | Photo Illustration by Tricia Seibold with iStock/izusek and iStock/axel2001

Silicon Valley and Washington are both in an uproar about revelations that Cambridge Analytica, a pro-Trump “psychographic” consulting firm, got ahold of detailed personal data on 87 million Facebook users.

But while much of the furor has been over privacy and ethics, a practical question remains: Is psychological targeting an effective tool of digital propaganda?

The answer, according to a Stanford researcher who pioneered many of the original techniques, is “yes.”

“I’ve been warning about these risks for years,’’ says Michal Kosinski, a psychologist and assistant professor of organizational behavior at Stanford Graduate School of Business. “Our latest research confirms that this kind of psychological targeting is not only possible but effective as a tool of digital mass persuasion.”

Kosinski never worked for Cambridge Analytica and never acquired Facebook data without users’ permission.

Taking Advantage of Facebook “Likes”

As a doctoral student and deputy director at Cambridge University Psychometrics Center from 2008 to 2014, Kosinski worked with a colleague to investigate whether it was possible to identify people’s psychological traits from their Facebook “likes.”

People who “liked” Battlestar Galactica were likely to be introverts, for example, while people who “liked” Lady Gaga were likely to be extroverts. Kosinski and his Cambridge colleague, David Stillwell, were able to correlate “likes” with other basic personality traits: openness, conscientiousness, agreeableness, and neuroticism. Armed with only 10 “likes,” they could evaluate a person’s traits more accurately than that person’s coworkers. With 70 “likes,” they could do better than a person’s close friends.

And now, in a new study, Kosinski and his colleagues — including Stillwell, Sandra Matz of Columbia Business School, and Gideon Nave of Wharton School of Business — confirm the next logical step: Ads are indeed more persuasive when they are tailored to those psychological traits.

Research Meant to Warn

Kosinski isn’t boasting about this.

“Most of my studies have been intended as warnings,” he says. “You can imagine applications that are for the good, but it’s much easier to think of applications that manipulate people into decisions that are against their own interests.”

He and his colleagues created a Facebook app that allowed people to fill out a personality questionnaire that measures five basic personality traits. They then asked users for access to their “likes,” and eventually amassed a database with 3 million profiles.

By correlating people’s “likes” with their scores on the personality questionnaire, Kosinski and Stillwell developed algorithms to accurately infer a host of personality traits from a person’s Facebook activity.

The founders of Cambridge Analytica adopted similar techniques and applied them to politics. They also went a big step further, using their own app to secretly collect the Facebook activity on tens of millions of users who had simply been friends of people who had taken the app’s quiz.

In their new study, Kosinski and his colleagues wanted to see if psychological targeting actually delivered better results in advertising. The researchers ran three experimental ad campaigns over Facebook.

Measuring the Effect of Targeted Ads

In promoting a line of cosmetics, for example, they ran dueling ads aimed at introverts and extroverts. All told, the ads reached 3 million people.

 

Quote
Most of my studies have been intended as warnings. You can imagine applications that are for the good, but it’s much easier to think of applications that manipulate people into decisions that are against their own interests.
Attribution
Michal Kosinski

The ad for extroverts featured a woman dancing and the slogan “Dance like no one’s watching (but they totally are).” By contrast, the ad for introverts featured a woman contemplating herself in a mirror and a quiet slogan: “Beauty doesn’t have to shout.”

Sure enough, people were 50% more likely to buy the cosmetics if they saw the ad aimed at their particular type.

The results were similar when the researchers promoted a crossword puzzle app for smartphones with ads that targeted users based on their openness to new things.

People who had been identified as very open were urged to “unleash your creativity” on “an unlimited number” of puzzles. People identified as likely to cling to the familiar were told to “settle in with an all-time favorite.”

Those who saw the ad aimed at their particular level of openness were 30% more likely to download the game than those who didn’t.

In a third test, Kosinski and his colleagues tested rival ads for a video game that they already knew appealed heavily to introverts. The first ad featured a standard action-packed pitch: “Ready? Fire!…” The second ad was tailored to introverts: “Phew! Hard day? How about a puzzle to wind down with?” Here, the ads for introverts generated 30% more clicks and 20% more downloads.

Kosinski says it’s probably impossible to prohibit psychological targeting as a tool of political propaganda, but he says people can defend themselves by becoming aware of how it works. They may also be able to enact policies that prevent abuses.

“It’s a bit like fire,” he says. “You can use fire to both warm your house and burn it down. You can’t ban fire, and you can’t stop some people from committing arson. What you need are firefighters and fire-safety equipment.”

For media inquiries, visit the Newsroom.

Explore More