Researchers at Cambridge University’s Psychometrics Centre have shown that it’s possible to uncover surprisingly accurate estimate of peoples’ personality traits such as race, age, IQ, sexuality, political views and even drug use from automated analysis of what they ‘Like’ on Facebook.
The study, administered in collaboration with Microsoft Research Cambridge, analyzed 58,000 US Facebook users. Volunteers gave access to their Facebook ‘Likes’ along with profile information they had entered Facebook’s myPersonality application. Results extracted from seemingly innocuous Facebook Likes were eerily accurate in determining a test subject’s traits.
The study proved 88 percent accurate for determining male sexuality, 95 percent accurate for distinguishing African-American from Caucasian, and 85 percent accurate differentiating Republican from Democrat. Christians and Muslims were correctly classified in 82 percent of cases. For relationship status and substance abuse, results were correct between 65-73 percent of the time.
The researchers were also able to accurately identify difficult to define attributes such as intelligence, emotional stability, openness and extraversion. One striking conclusion from the study was that when it comes to recognizing openness, Facebook data alone is nearly as informative as using an individual’s actual personality test scores.
The study also had a set of amusing findings. If users ‘Like’ curly fries they often had high IQs. If male users were fans of the TV show “Glee,” they were probably gay. If users ‘Like’ pages titled “Never Apologize For What You Feel It’s Like Saying Sorry For Being Real” or “I’m The Type Of Girl Who Can Be So Hurt But Still Look At You & Smile,” the users parents had probably split up before they left home.
Studies like this show why social networks have become so valuable to advertisers. Major retailers such as Target have been mining their customers data on Facebook for years, looking for patterns that predict future buying behavior. Forbes highlighted the practice in an article amusingly titled, “Target knows you’re pregnant before your dad does.”
Additionally, police in several California cities are using data mining for what they label “predictive policing,” keeping tabs on what people share on social networks to identify potential criminal behavior.
As helpful and telling as these studies are, there still remains debate whether this information should be public. Michal Kosinski is Operations Director at the Psychometric Centre. He conducted the Cambridge University study with colleague David Stillwell and Microsoft Research Center’s Thore Graepel.
“I am a great fan and active user of new amazing technologies, including Facebook. 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,” admitted Kosinski.
Does information like this worry you about social network use Let us know your thoughts in the comments below!
Source University of Cambridge
