Print options

March 12, 2013

Personality Can be Predicted via Facebook 'Likes'

by VOA News

Your digital footprint says a lot more about you than you think.

According to a recent study by researchers at Cambridge University, easily accessible online digital records can be distilled to predict some personality traits or behavior we might prefer be kept private. Among those traits and behaviors, according to the study, are “sexual orientation, ethnicity, religious and political views, personality traits, intelligence, happiness, use of addictive substances, parental separation, age and gender.”

The analysis was based on 58,000 volunteers who offered to share their Facebook “likes.” Using the information gleaned from Facebook, the researchers were able to accurately tell a man’s sexual orientation 88 percent of the time, whether they were white or African-American 95 percent of the time, and whether they were a Democrat or Republican 85 percent of the time.

Even religious affiliation, specifically determining if a person was Christian or Muslim, was predicted accurately 82 percent of the time.

Intelligence was another trait researchers were able to accurately measure, and Facebook “likes” associated with intelligence included “thunderstorms,” “The Colbert Report,” a popular comedy show in the United States, “science” and “curly fries.”

Clues to lower intelligence were liking “Sephora,” a chain of perfume stores, “I love being a mom,” “Harley Davidson,” a popular manufacturer of motorcycles and “Lady Antebellum,” a popular country music group.

According to the paper, published in the journal PNAS, information like this could be used by corporations to better target potential consumers.

For example, a U.S. retailer tried to predict pregnancies among its female customers, so it could send them discount offers for prenatal vitamins or maternity clothing.  While the offers might have been welcomed by some, the negative side is that it might lead to sending offers to unwed women from a culture that frowns on pregnancy out of wedlock.

“As this example shows, predicting personal  information to improve products, services and targeting can also lead to dangerous invasions of privacy,” according to the paper.