New research suggests that social media sites can work as predicators of personality.
According to the theory of the Five Factor model, human personality can be divided into ﬁve diﬀerent core dimensions: openness (having an active imagination, aesthetic sensitivity, and intellectual curiosity), conscientiousness (being organized and achievement-oriented), extraversion (being social, assertive, and talkative), agreeableness (being helpful, cooperative, and sympathetic), and neuroticism (being anxious, envious, guilty, or depressive). Those factors, the theory goes, describe someone's overall personality and offer important clues when it comes to that someone's role in social interactions. That makes them valuable not only to social science researchers, but also to educators and daters and content providers and marketers.
Traditionally, personality factors, the Big Five and otherwise, have been measured according to self-reported info -- through questionnaires and the like. But people's own view of themselves is often notoriously at odds with the way other people view them, so simply asking people about themselves can lead to discrepancies and, occasionally, misleading results.
A team of researchers in Asia have been experimenting with another method of personality prediction: one that uses people's behavior on social media sites as a proxy for personality. Shuotian Bai and Tingshao Zhu, of Beijing's Graduate University of the Chinese Academy of Sciences, along with Li Cheng of Singapore's Bioinformatics Institute, have proposed "an automatic and objective personality prediction system based on user's behaviors on Social Network Sites." They asked over 200 Chinese students with accounts on Renren, a Chinese competitor to Facebook, to take the Big Five Inventory, a test of the five core personality traits developed at Berkeley in the 1990s.
Armed with the self-reported personality data, the researchers also analyzed the students' personal Renren pages, tracking things like age and gender, as well as the frequency of the students' updates, the most common pronouns (I, me, you, they) the students used in those updates, and the emotional content (what they call the "majority emotion count") of the students' posts. Cross-referencing the data gleaned from the self-reported personality tests and the Renren content analysis, the researchers found strong correlations. Personality, their findings suggest, can be accurately reflected on social media sites.
Though it's an early finding, and more work needs to be done to figure out how, exactly, personality and presentation interact with each other on sites like Renren and, here in the States, Facebook ... the suggestion of correlation is, in itself, fascinating. And a tad alarming. Because, if personality can be measured according to our social media presences, we can be measured, as well. And without our explicit permission.
The implications for us, as consumers and psychologies and interacters and human beings, can be huge. Marketers can tailor their offerings to us, based not just on our purchase histories and friend networks, but also on our personality types. News organizations can tailor their content to our un-expressed -- and, yet, omni-expressed -- preferences. Employers can analyze us. And we will have opted into it all -- not through checking a box, but simply through expressing ourselves on the Web.
We want to hear what you think about this article. Submit a letter to the editor or write to email@example.com.