Do you "like" Channel 4 News on Facebook?
We are still waiting to see what this reveals about people who do - the Cambridge University Study into online privacy on Facebook did not take Channel 4 News "likes" into account.
But in studying the seemingly random "likes" of 58,000 Facebook users, researchers were able to match preferences in music, food, or films, with more fundamental and personal characteristics, such as sexuality, intellect, race or political leanings.
"Your likes may be saying more about you than you realise," said Cambridge University researcher David Stillwell, one of the study's authors.
At the front end, a Facebook "like" comes across as this very social form of socialising, while at the same time turning them into forms of data not just for Facebook, but also for third parties and profilers. Carolin Gerlitz, University of Amsterdam
In 95 per cent of cases, the algorithm model used by researchers was able to determine African Americans from Caucasian. Gay men were picked out 88 per cent of the time, and the difference between Democrat and Republican was determined in 85 per cent of cases. Drug use was determined from "likes" in 65 per cent of cases and in 60 per cent of cases, researchers could statistically ascertain when users' parents had divorced before they were 21.
Some of best predictors of intelligence were, surprisingly, an expressed "like" for thunderstorms and "curly fries", whereas low intelligence was associated with people who "like" the statement "I love being a good mom" or Harley Davidsons.
In terms of sexuality, likes for Wicked the Musical had a strong correlation with gay men, while the statement "being confused after waking up from naps" was a strong predictor of male heterosexuality.
The price of a 'like'
According to Facebook's 2012 figures, there are around 2.7 billion new likes every day.
But this study shows that our endorsements of everything from a clothing sale, to a pop star, are revealing much more about ourselves than we think - and are in fact, a free gift of huge amounts of personal data to third parties.
"We all need to be more sensitive," social media analyst Ian Maude, from Enders Analysis, told Channel 4 News, adding that we are subconsciously allowing more and more personal information to become public.
"There is clearly a trade-off between having access to these services, which are ostensibly free, and the information we're giving up."
Most predictive likes:
Gay men > Wicked the Musical, Mac Cosmetics and Glee
High intellect > The Daily Show, Lord of the Rings and Curly Fries
Low intellect > I love being a good Mom, Harley Davidson, Sephora
Smoking > Cradle of filth, Rob Zombie
Non-smoking > Honda, Rock and How to make a girl smile
Parents separated pre-21 > "Never apologise for what you feel it's like saying sorry for being real"
Parents not separated > "Watching peoples lives fall apart via status updates and news feeds"
Extraverted > Michael Jordan, Dancing and Beerpong
Shy > Fanfiction.net, anime and programming
Read the full study published in the PNAS Journal here and visit youarwhatyoulike.com where the study participants' information was reviewed.
Facebook launched its like button in 2009, and the small thumbs-up symbol has since become ubiquitous on the social network and across the web, as various brands encourage us to "like" them via share buttons on their own pages.
However the apparently innocent "like" could have far-reaching consequences offline if it can be used to statistically predict something much more sensitive.
"I appreciate automated book recommendations, or Facebook selecting the most relevant stories for my newsfeed. However, I can imagine situations in which the same data and technology is used to predict political views or sexual orientation, posing threats to freedom or even life," said Michal Kosinski, one of the study's authors, from the University of Cambridge Psychometrics Centre.
"Just the possibility of this happening could deter people from using digital technologies and diminish trust between individuals and institutions."
While the Cambridge University study, published in the PNAS journal, assessed active "likes", Facebook still manages to retrieve users' data through Facebook cookies placed on machines via Facebook Connect.
No comments:
Post a Comment