E-mail updates
Follow on Twitter
Subscribe to RSSNew research analyzing the "likes" of nearly 60,000 Facebook users found that a person's race, gender, political views, religion and even sexual orientations could be identified with a high degree of accuracy. Among the findings: if you "like" curly fries, you're probably more intelligent than average, and if you "like" cuddling, you're probably a bit more politically liberal.
When you click a "like" button on Facebook, you could be telling the world whether you're gay or straight, liberal or conservative, intelligent or not so much — even if you don't intend to. That's what researchers found when they ran tens of thousands of Facebook profiles and questionnaires through a computer algorithm to find the obvious as well as not-so-obvious connections.
The results were published on Monday in the Proceedings of the National Academy of Sciences, and you can sample the method for yourself at a website called YouAreWhatYouLike.com.
"The main message of the paper is that whether they like it or not, people do communicate their individual traits in their online behavior," said lead author Michal Kosinski, operations director at the University of Cambridge's Psychometrics Center.
Some of the correlations are obvious: For example, If you're a fan of the "I'm Proud to Be a Christian" Facebook page, it's a pretty safe bet that you're a Christian. But others are hard to explain: Why is it that liking the "Curly Fries" page is associated with having a high IQ? Why does the computer model put "Sometimes I Just Lay in Bed and Think About Life" in the category for homosexual females, while "Thinking of Something and Laughing Alone" is linked to heterosexual females?
"These little patterns are really not perceptible to humans," Kosinski said. Sometimes, it takes a computer.
Kosinski and his colleagues conducted their experiment over the course of several years, through their MyPersonality website and Facebook app. More than 8 million people took the MyPersonality survey, which asked participants about their personal details and also had them answer questions about personality traits. About half of the test-takers gave their OK for the researchers to match up their survey results with Facebook likes, on an anonymous basis. More than 58,000 of the volunteered profiles from U.S. respondents were selected for matching.
The results were analyzed to produce correlations in more than a dozen categories, including five widely accepted personality attributes (openness, conscientiousness, extraversion, agreeableness and emotional stability). Those are the attributes analyzed on the "You Are What You Like" website. The other categories included IQ, religion, politics, sexual orientation, age, gender, race, relationship status, alcohol and drug use, tobacco use, life satisfaction, number of friends — and even whether a Facebook user's parents had separated by the time the user was 21.
This PDF file shows you which Facebook pages are the best fit for each of the categories.
YouAreWhatYouLike.com
Researchers set up a website that assesses your personality based on Facebook "likes."
The researchers' computer model did the best at predicting black-vs.-white and male-vs.-female (95 and 93 percent accuracy, respectively). It could distinguish correctly between Republicans and Democrats 85 percent of the time, and between Christians and Muslims 82 percent of the time.
The accuracy rates for predicting sexual orientation were 88 percent for males and 75 percent for females. But don't think reaching that result was as easy as seeing who clicked the "like" button for "Gay Marriage." Less than 5 percent of the gay users were fans of such obvious pages, Kosinski and his colleagues said. The predictions were based instead on inferences from likes for less obvious pages. For example, the computer model associated the fan pages for Kathy Griffin and "Wicked, The Musical" with homosexual males, while heterosexual males were associated with the pages for Bruce Lee and WWE wrestling.
OK, maybe the pages weren't all that much less obvious.
The model wasn't as accurate (60 percent) when it came to predicting whether a user's parents stayed together or separated before the user turned 21. But even that level of predictive power could be "worthwhile for advertisers," the researchers said. "For instance, digital systems and devices (such as online stores or cars) could be designed to adjust their behavior to best fit each user's preferred profile," they wrote.
"I know the paper might sound like we're criticizing Facebook, but not at all," Kosinski told NBC News. "I'm a fan of Facebook."
Kosinski pointed out that an analysis of your credit card purchases, online music preferences, video rentals and Web browsing habits could come up with personal profiles at least as detailed as the ones that he and his colleagues produced. It just so happens that the Facebook likes were accessible enough to yield a vivid illustration of how such analyses work.
"It's possible this will lead some people to say, 'Maybe I shouldn't be using Facebook, or I shouldn't be using Google.' And that could be bad," he said. That kind of technophobia could hamper technological and economic progress, he said. Instead, the research should lead people to think twice about what they share online.
"We hope this information will help users start a discussion with organizations like Facebook, Google, Microsoft, or even policymakers about the rules of the game online," Kosinski said.
Update for 3:55 p.m. ET March 11: Kosinski's two co-authors, David Stillwell of Cambridge and Thore Graepel of Microsoft Research, passed along their comments in a news release from Cambridge.
"Consumers rightly expect strong privacy protection to be built into the products and services they use, and this research may well serve as a reminder for consumers to take a careful approach to sharing information online, utilizing privacy controls and never sharing content with unfamiliar parties," Graepel said.
"I have used Facebook since 2005, and I will continue to do so," Stillwell said. "But I might be more careful to use the privacy settings that Facebook provides."
More about Facebook research:
- Facebook posts are more memorable than faces
- Facebook's roots go way, way back
- Scientists map 'Facebook for birds'
The PNAS paper, titled "Private Traits and Attributes Are Predictable From Digital Records of Human Behavior," includes a conflict-of-interest statement: Stillwell received revenue as owner of the MyPersonality Facebook app. Kosinski received funding from the Boeing Co. and Microsoft Research.
Alan Boyle is NBCNews.com's science editor. Connect with the Cosmic Log community by "liking" the log's Facebook page, following @b0yle on Twitter and adding the Cosmic Log page to your Google+ presence. To keep up with Cosmic Log as well as NBCNews.com's other stories about science and space, sign up for the Tech & Science newsletter, delivered to your email in-box every weekday. You can also check out "The Case for Pluto," my book about the controversial dwarf planet and the search for new worlds.
This story was originally published on Mon Mar 11, 2013 3:02 PM EDT
Leave a Comment:
Here's how well the computer model did at predicting various traits and attributes using Facebook likes:
Caucasian vs. African-American - 95 percent
Male vs. female - 93 percent
Gay vs. straight (male) - 88 percent
Democrat vs. Republican - 85 percent
Christianity vs. Islam - 82 percent
Lesbian vs. straight (female) - 75 percent
Smoker vs. non-smoker - 73 percent
Alcohol drinker vs. non-drinker - 70 percent
Drug user vs. non-user - 65 percent
Single vs. in relationship - 67 percent
Parents together vs. separated when child is 21 - 60 percent
Prediction accuracy of regression for numeric attributes and traits is more difficult to describe. For you statisticians, here are the Pearson product-moment correlation coefficients between the actual and predicted values: Age (0.75) ... Number of Facebook friends (0.47) ... Density of friendship network (0.52) ... Openness (0.43, test-retest reliability 0.55) ... Conscientiousness (0.29, test-retest reliability 0.7) ... Extraversion (0.4, test-retest reliability 0.75) ... Agreeableness (0.3, test-retest reliability 0.62) ... Emotional stability (0.3, test-retest reliability 0.68) ... Intelligence (0.39, test-retest reliability 0.78) ... Satisfaction with life (0.17, test-retest reliability 0.44).
The satisfaction-with-life showed relatively low prediction accuracy, and the researchers addressed that in the paper. They say the low accuracy "may be attributable to the difficulty of separating long-term happiness from mood swings, which vary over time. Thus, although the SWL score includes variability attributable to mood, users' Likes accrue over a longer period and, so, may be suitable only for predicting long-term happiness."
It seems likely these kinds of databases are much more common then a few research projects and a facebook profile. Credit card purchases don't even begin to cover how we are tracked.
I don't worry so much about academics or the government, but when Walmart has its own satellite network to track us all, I start to be a bit more concerned. Targeted marketing may help you find the car you want at just the right time, but it can also manipulate you into making unhealthy purchases you wouldn't otherwise make. Just for starters.
Edit: Interesting article.
Alan Boyle, Science Editor, NBC News
It seems like the non-random sampling could lead to skewing of the data. People who are more cautious with their online presence would likely not have volunteered for this study.
DingleB, that was the first question I asked Kosinski. He acknowledged that "there is a chance that it's just that we had a biased sample." But he was particularly interested in the correlations that involved "things that even if they would like to, they would not know how to communicate." He pointed out that the sample was drawn from millions of surveys. "If you have a larger sample, then you would be able to make predictions that are even more accurate," he said. The researchers are now looking at non-U.S. correlations from their sample. By the way, Kosinski said he and his colleagues are also working with JustGiving.org to see whether there are correlations between likes and the willingness to make contributions. He said the detailed results are not ready to be shared yet, but "we're able to predict very nicely who would be willing to donate money."
Most statistics are compiled from willing participants. One of the things I find interesting about this study is this: I lie about around half of the questions when I take a poll or survey, I almost never like or upvote something I don't actually like or agree with. So, while the sample may be skewed in that only those willing to be counted got counted, the sampling method of using facebook likes likely gave a very accurate look at someones preferences, probably more accurate then the control test they gave to compare the numbers to.
big government (bastard) is watching you, believe it or not !
Who really cares ???? How are you going to categorize me now??
Governments aren't really in a position to leverage this kind of information. Governments move slowly in trends lasting decades or centuries. Corporations and the sociopaths that run them internationally are the parties with an immediate interest in this kind of data.
Sex may sell, but violence brings the audience to the table. Not really the kind of thing Big Brother is interested in.
This is just an honest question. How did they come to their conclusion? Did someone from from Facebook or someone researching on their behalf actually go out and investigate the actual people that were used to facilitate this study?
For example, you claim that they can predict with 73% accuracy who is a smoker or non-smoker. Did they actually look at their predictions and then go out and personally investigate which Facebook users actually smoked?
Is their some other method they used?
They used self reporting on the survey and then followed the facebook likes.
The facebook likes are almost certainly a more accurate insight into a personality then self reporting. Taking a survey is not taking the MBTI, a diagnostic tool is different then a research tool.
Self reporting in surveys and polls have a margin of error based on how subtle the information they are gathering is collected. Straight forward surveys are less common then push polling, it's all in the degree of subtlety to get straight information about someone as compared to what someone wants you to think about them.
There are three kinds of lies...lies, damn lies, and statistics. The more you know about statistics and how they are compiled and used the more clearly you can make decisions in an information saturated society. Remember....you are the product when you watch TV.
One would hope one could figure most of that stuff out from a profile picture or that it could figure out something that people are proud of about themselves. What in that list, aside from drug use or smoking, would your typical person want to hide? Facebook isn't supposed to be annonymous. I would hope most people would be pretty forthright or self confident enough to tell almost anyone their race, sex, gayness, and religious belief. I wouldn't consider alcohol use or parents' marital state to be particularly personal information (if you even consider the 60% guess rate that acceptable).
How well does it do at revealing information someone doesn't want the outside world to know?
Noting that I make a pretty active effort to like off the wall things, and ignore things I actually give a hoot about... so unless it's also tracking the time spent in places....
I could make a pretty good case that someone who wastes their time going around clicking that stupid "like" button cannot possibly be any version of "intelligent".
This shows profiling works. This is why terrorist should be profiled at airports. Taking a load off TSA will take pressure off law abiding citizens.
I have always been very skeptical of survey/poll data. I must agree with DingleB, those who are surveyed or polled are those who are willing to take said poll or survey. Not to mention some people just "like" things on facebook to support a friend regardless of whether they truly agree with the issue.
I "liked" a few artists on facebook before, not because I agree with everything they symbolize, believe or do, but because I may have liked one song they sang. Now to take that one "like" and try and draw a correlation into my religious, political, or moral belief system is going a bit too far.
But that's JMO.
Do they correlate having a high IQ with not using Facebook at all? :P
What the study does not show is that the people with the Highest I.Q. are the ones that never click "Like" buttons, even if they have a Facebook account, because they already knew how their information is used (they actually read and understood the Privacy Agreement) and chose not to participate.
(Click the Up Arrow to Like this comment. ;-) )
I up voted your comment, but only to inject some static into my news vine profile. ;-)
High IQ, or paranoid?
My IQ is pretty good, but I agree with JRS. It is not paranoia for me, just independence. Whenever anyone asks me to Like or Not Like something, I feel they are pressuring me to give an answer. I talk a great deal and will tell strangers my life story, but I will NOT be pressured into answering dumb stuff. Once in a blue moon I will Like something on a comment board, but that is when something is said that I really feel good or bad about and no one is saying I should comment. They just leave the choice to me.
I find this discussion good for all to learn how data mining is accomplished. I also find paranoia comments on this subject used mostly as a derogatory personal attack. Some should not blindly believe in the "goodness" of government. Look closely at what is happening in our schools with respect to record keeping and you will find glaring omissions to personal privacy law. Perhaps if readers googled "Privacy in Education for Parents and Adult-Age Students" the article could point out a different perspective.
JRS: How does clicking the like button betray a lack of intelligence? I would venture to say that many of MSNBC's posters read these comments to help them form an opinion on a subject on which they had not yet formed a clear position. There is no pressure but the internal one that drives curiosity, the desire to be well-informed. Seeing how many 'like' a certain comment also provides me with a valuable reality test, i.e., just how much does my opinion/stance deviate from the majority of respondents.
This is not to say, of course, that intelligent people derive their opinions solely from adherence to a 'majority rules' theory of democracy or economics, only that it can be useful to see where others stand before making up one's own mind.
JRS-619990
LOL I "liked" your comment because it made me laugh... but don't think I would vote for you to be POTUS! LOL
anyway, I have been legit tested and my I.Q. is 145.. they say that's on the higher end of the I.Q. scale, but believe me if you see my facebook at times, you would think my I.Q. is lower than Forrest Gump. Facebook is about having fun and keeping in touch with people you otherwise wouldn't hang out with on the day-to-day of your everyday.. and believe me, a lot of what you see me "liking" and 'doing" on "online", I wouldn't normally do in "offline"
Anyway, I am still giggling at your comment!
Aaaaaand...this is surprising exactly how? Of course the things you like--or "like"--reveal traits! Someone needed to conduct a study to discover this?...
Conjuring Cat
Why, yes!.. and they thank you for your tax dollars! lol
No, on a serious note, I have my B.B.A. in Marketing and my favorite classes were Marketing Research and Consumer Behavior... you'd be amazed at the things these companies track, how they track and what "information" they pull and deduce from said data.
It's a load of horse manure. I don't need a computer to tell me the links. If you are a fan of gay pages, you probably are gay. If you like pages with makeup, and fashions, and shoes, you are probably a female. I don't know any guy who would like a page for feminine products. i fI like the Obama and Biden pages, I'm probably a democrat. If I like all the Jesus pages, I'm probably a Christian. No brainers here.
The point is that there are a lot of less obvious correlations, and the researchers are trying to figure out the explanations behind those linkages. By the way, the linkages weren't made just on the basis of some super-brain "liking" Curly Fries. There had to be at least 100 likes to make the correlation. But I hear you on the fact that a lot of the correlations could have been predicted.
So how can they tell if someone is a supporter of a cause, such as gay rights vs that person actually being gay? Also some things I click on such as a pic someone posted up, are just things I find funny, it doesn't really mean anything.
XD Shooter;
I have seen many, many instances of people "liking" something they strongly disagree with just so they keep up-to-date on it.
Such as Right-wing Republicans "liking" a Left-wing Democrat page/article/etc -- or the other way 'round -- just so they can further nit-pick the hell out of it.
This is why you should not invite people to be your friends on Facebook if you do not trust them. Don't always give in to your co-worker's or your boss's request to be Facebook friends. Also, Facebook likes should be considered protected speech under the 1st Amendment.
Because they'll find out your sexual preference, or that you like curly fries? I agree, I sure as hell don't want anyone to find out how I like to eat my potatos.
Willow Sunstar;
And if someone has been bad-mouthing you on Faceypages? What recourse do you have then, if what they "like" is protected under the 1st Amendment?
Here's a big friggin' clue for you, Willow: Slander and libel are NOT protected 1st Amendment "free speech" -- nor should bullying, threats and harassment be protected.
The site classified me as "liberal" when I'm politically to the right of Archie Bunker and as "cooperative rather than aside and competitive" when in fact I'm a gamer and tenacious in debate. I'm very much not impressed.
Cheer up. At least they got your gayness right.
"...I don't know how someone controlled you,
They bought and sold you...."
George Harrison (From: While My Guitar Gently Weeps)
Facebook... yuck.
More momentous ‘research’ stating the bleedin’ obvious.
If you 'like' Christian stuff then it's a pretty sure bet that you're a Christian and not a Muslim or an atheist. Do we really need people stuck in their ivory tower universities to tell us this?
If you looked at the profile I use to keep in touch with family you would think I was a conservative Christian. I am not. If I had used that profile and participated in this study I would have been an outlier that was tossed off one end or the other of the data.
I'm a real mess since I "like" everything!
It's addicting isn't it!
LOL now I have decided to NOT "like" anything I do like, and "like" everything I don't like... or maybe i'll just mix it up... or maybe I won't...
I'll leave it to FACEBOOK to figure it all out! lol
Screw FaceBook !, It's a High PRICE to Pay(Your Privacy) for making virtual "Friends" ! I started a facebook website 2 years ago--------I haven't added to it,Contacted anyone, or touched it in anyway since. It's a playground for hackers & Players. NO Thank You !!!
What are the "hackers" going to gain from your facebook account? Did you list your credit card and/or social security number on your profile? Anything you post online, on any site, can be "hacked." Facebook is definitely not the least secure.
Also, re: virtual friends. If you don't know someone, don't accept the friend request. It's actually a pretty useful tool for keeping up with people you do know but don't see on a regular basis.
I am an old lady - I actually have listed about 45 friends on facebook. They are relatives from around the country and friends in my small town. I comment once in awhile, especially to family. I also like that a lot of thoughtful or funny stuff is posted that I would never search the internet for. I can't imagine how anyone has the time to have a hundred or more friends. More like fly specks on the wall for as much attention as you pay to them.
Or those "likes" could just reveal that you enjoyed reading/looking at whatever the other person posted. Why does everything have to have a "hidden" meaning?
And sometimes, a cigar is just a cigar.
I love this idea that our actions speak louder than anything else, and that big data can tell us something. However, my own use of the app labeled me as "shy" and that is pretty far off the mark. It also said I was well-organized.
I've had a Facebook account since 2008. I go on it nearly every single day. But I have never liked any pages (or posted any statuses, and rarely post to friends' walls or like comments/statuses) pretty much for this reason...I feel like it just says too much about me all at once and just leaves it open for people to assume they know what I'm like.
Although, the prospect of actually going ahead and liking all the things I like seems fun if only to be able to look at a bunch of things I enjoy on a single page.
While I agree, for example, most gay folks would "like" gay posts, I am a perfect candidate for skewing the results, totally heterosexual but supporting the LGBT community 100%......just one of several instances where an "outsider" can alter the numbers....
Sounds like you're right in line with the results. Though, I think they're only taking results from "liked" pages, not comments your friends post.
there are many gays who do not support gay marriage as well, so that could go either way.
Crap. lol. Now all my friends will know I'm gay. :D Oh, wait. They already do.
right? like the photos of 2 men or woman kissing on their fb page wasn't enough of a clue.
the tiara and boas didn't set off any red flags either. LOL.
seriously? just because you hit "like" on a gay-subject post/comment does not make you gay. you might be for or support it but it does not make you gay.
this is just stupid and just another need to classify, label, or group people for the sake of isolating, separation, or inclusion/exclusion.
some folks have waaaaaaaaayyyyyyyyyyyy too much time to waste on crap like this.
Good thing they did this study because I had no idea people could figure out I was a liberal when I liked "Being Liberal"
I like mashed potatoes and gravy, french fries and vinegar, home fries with scrambled eggs and potato and leek soup. Damn I'm screwed.
And the "takeaway is"??????
Nothing worse then a d1k sucking conservitard,,,losers like all gay losers
So if you like Jesus Christ, Sarah Palin and the Bible you're satisfied with life, but if you like science you're dissatisfied with life? Whatever, Not!
This is completely off... I know because I asked my wife. When it says I'm well organized and cooperative, calm and relaxed... These things couldn't be further from the truth....
I cant stand facebook. I don't "like" anything.
And now im really glad I never "liked" anything...
Oh well, I was never a follower. I guess that says something about me too. :(
why do people have to nitpick when a majority of the post are innocent and were shared IF you had a sense of humor. Stop criticizing. The people that don't like FB is fine however stop worrying and sticking your nose where it doesn't belong