Facebook's psychographic profiling isn't going away

2025-04-28 05:35:08admin

When Cambridge Analytica got its hands on the interests, likes, and photos of 50 million Facebook users, they must've known they struck gold.

Psychometrics -- an over 100-year-old science dedicated to measuring emotions and other matters of the mind -- once suffered from a paucity of data. But the social network has opened pandora’s box.

SEE ALSO: Forget data. Free labor is Facebook's lifeblood

Historically, analysts were stuck with limited information when it came to psychometric information -- a lot of it came from surveys, which of course had to be consciously filled out by a subject. We now have a deluge of social media information that can give analysts a colossal archive of our "likes," language, friend networks, and history to play with.

"Psychometrics by itself is not a new field," Sharath Guntuku, a computer science researcher at the University of Pennsylvania, told Mashable. "What's new is using social media as a psychometric tool."

Guntuku uses social media information — which he gets direct consent from hospital patients to use — as part of a "pioneering mission" to measure psychological and physical well-being. But, as the Cambridge Analytica controversy shows, modern social media data can twist the technique in at best unethical, and at worst, deceitful and manipulative ways.

"There’s a huge amount of data that’s ripe for abuse," said Luke Stark, a postdoctoral fellow at Dartmouth College's Sociology Department, in an interview.

How valuable is social media data in psychographics?

"Social media can be very effective at predicting emotions," said Guntuku.

Psychometric analyses of people's social media language have proven to be quite accurate. An analysis of language alone can give researchers an accurate identification of gender 80 percent of the time, Guntuku said. Computer programs can also correctly identify someone's age within less than six years 94 percent of the time.

Language alone lets researchers predict a subject's gender 80 percent of the time

When combined with rich data about someone's feelings about provocative societal issues — like immigration, war, and gun-control — an electioneering company like Cambridge Analytica can determine whom to target with social media ads, and how. As reported by theGuardian, Politico published "sponsor-generated content" directly paid for by the Trump Campaign entitled "Clinton Corruption: Ten Inconvenient Truths about the Clinton Foundation." This interactive infographic reportedly served as one of Cambridge Analytica's most successful 2016 election ads.

Still, social media data, while plentiful, has some limitations that can make psychographic analyses challenging. That's why psychographics are predictive, and not completely certain. Analyzing social media behavior (language in posts or "likes") might easily show if someone is happy or sad, "but if you're depressed or not — it’s a more difficult question to answer," Guntuku said.

"Diagnoses aren't always accurate, even in clinical settings."

Mashable Top Stories Stay connected with the hottest stories of the day and the latest entertainment news. Sign up for Mashable's Top Stories newsletter By clicking Sign Me Up, you confirm you are 16+ and agree to our Terms of Use and Privacy Policy. Thanks for signing up!

Did Facebook's psychographic data really influence a presidential election?

Some say this Facebook data, collected in 2014, wasn't that useful in predicting the personality traits of American voters, and thus gave no real advantage to the Trump campaign in swaying voters in Donald Trump's favor. But there's little reason to accept this conclusion at this point. Revelations are still coming to light, and how exactly Cambridge Analytica violated its data-sharing agreement with Facebook isn't fully known.

"Whether or not it’s effective in changing the outcome of an election — the jury is still out," Stark said.

Christopher Wylie, one of the founders of Cambridge Analytica, exposed how the marketing company used data (without consent) from millions of Facebook users. Credit: Getty Images

It may never be concretely known if the Facebook data ultimately gave the Trump campaign its razor-thin edge in the Electoral College, but at minimum, it benefited them in some way.

"You would find a way to know something," said James Roberts, who heads the Psychometric Research and Development Lab at Georgia Tech. "Whatever [Cambridge Analytica was] attempting to do, it was done better with the data than without."

What is known, however, is that psychographic profiles devised from Facebook's data can be extremely valuable. Follow the money: Facebook brought in over $40 billion dollars in revenue in 2017, over $13 billion more than the previous year. Facebook generates nearly all its income from digital ad revenue, over 85 percent of which come from mobile ads. Mark Zuckerberg's company has a keen understanding of how to sell advertising space -- and to whom.

The problem arises when this data is used outside these companies, specifically for manipulation.

For instance, Cambridge Analytica might have used Facebook data to identify sentiments of a certain voter demographic, and then modeled advertisements to play into their ideologies, or to shape their perception of reality, as the company claims it could do.

People are aware that their online behavior will be used for targeted advertisements, but no one signs up for targeted manipulation. Things get even worse when people are manipulated with propaganda or "fake news."

And once published for public consumption, those falsehoods spread like wildfire. MIT researchers recently found that false news stories are likely to be retweeted nearly twice as muchas accurate news reports.

After the Cambridge Analytica scandal, will social media data be better protected?

Due to the intense backlash over Cambridge Analytica's violations, U.S. regulators like the Federal Trade Commission (FTC) are investigating Facebook's privacy policies, and might enforce stricter data protection rules for social media companies.

"Incidents like this one will likely serve to drive adoption of additional regulation here in the United States, like the General Data Protection Regulation (GDPR) going into effect in E.U. nations two months from now," said Scott Nestler, who teaches ethics in data science and analytics at the University of Notre Dame's Mendoza College of Business, via email.

SEE ALSO: FTC confirms it’s investigating Facebook — a huge threat to the social media giant

The GDPR requires that people be given clearer ideas of what they're consenting to and that their data is anonymized in case it gets out -- as data in the digital age is wont to do.

Whatever comes of the Cambridge Analytica scandal, the results certainly won't mean an end to psychographic data being used to model our psyche -- at a minimum for sales purposes. Long before Mark Zuckerberg was born, companies employed metrics to determine what Americans want. But now, this data is more detailed and abundant than ever, and its applications are growing. It's not going away, so at a minimum, it should be better protected.

"Facebook data around personality needs to be considered as sensitive as other medical data," Stark said.

Featured Video For You