Facebook 'likes' good indicators of users' personal traits
(AP Photo/Paul Sakuma, file)
PHOENIX (CBS5/CNN) -
People click the thumbs up or the "like" button on Facebook pictures, pages and posts.
It's a seemingly meaningless action that can ultimately speak volumes.
A new study out of Cambridge University says what people like on Facebook says a lot about them.
In fact, it can often reveal quite accurately race, political leanings, even how smart users are.
Bailey Carlin is the social community manager for Santy Integrated, an advertising and marketing agency in Phoenix, and says those tiny thumbs up are a pretty big deal.
"I love a like. We have clients that will base success off of a like or a comment or share," Carlin explained.
Carlin says a simple click of the mouse isn't so simple.
"In reality, it's something that you're revealing without really knowing it," she says.
The study takes a look at all of the things a user likes and predicts something very specific about them to advertisers.
"The more likes that you have, the more assumptions we can make," Carlin said.
Those assumptions can be made with great accuracy about race, religion, sexual orientation and even IQ.
The study used more than 58,000 volunteers and showed those who liked Stephen Colbert, curly fries, and Morgan Freeman's voice are more intelligent than those who like Sephora, Lady Antebellum and Bret Michaels.
Dr. Michael Yasinski tells CBS 5, "Once you break it down with these algorithms, it looks like, based on this article, they're able to tell you whether or not you use drugs, for example, if you're gay or straight, black or white."
And those predictions, Yasinski says, come down to stereotypes.
"Feels like a lot of steps in the wrong direction. That's what we're trying to get away from as a society, and we're going to start labeling people again," he said.
Yasinski also worries that the information generated may soon become a screening tool for potential employers.
"They will take a data report that takes a look at all of your online behavior, sites you've visited, the likes on Facebook, the posts you do. It's going to equate into things like, ‘You're a hard worker' or ‘You're not a hard worker.' ‘You're creative' or ‘not creative,'" he says.
These types of predictions only work with a lot of "likes" to allow more of the bigger picture.
Those in the study averaged 170 page likes, which is why the accuracy was so impressive in many of categories that describe users' personal traits.
Copyright 2013 CBS 5 (KPHO Broadcasting Corporation). All rights reserved.