Social networking sites display multiple facets of life: social, professional, and romantic. Self-regulated presentations of self on Facebook, Twitter, dating sites, and other newer forms of communication have led to an unprecedented amount of individual data which can be used by corporations, governments, and other interested parties.

Recent research has repeatedly demonstrated how seemingly innocuous public information such as number of Twitter followers or Facebook “likes” can accurately predict very personal information the user thinks is private, such as personality traits, sexual preference, or health status.

While research on this topic is typically used for marketing purposes, such as targeted ads and recommender systems, it has implications for healthcare providers as well. Kosinski et al. were able to use general Facebook “likes” of photos, friends, and products to identify substance abuse and relationship status with accuracy of around 65 to 73 percent, race and sexual preference with between 75 to 95 percent accuracy, and also cited the example of a U.S. retailer mining shopping records to predict pregnancies among female customers. They point out that “an unexpected flood of vouchers for prenatal vitamins and maternity clothing may be welcome, but it could lead to a tragic outcome, e.g. by revealing (or incorrectly suggesting) a pregnancy of an unmarried woman to her family in a culture where this is unacceptable.”

How does the real person compare to the profile?

Social networking sites provide a wealth of data for research, since a clear log of behavior and interactions is available for mining. Ivcevic and Ambady’s Face to (Face)Book: the Two Faces of Social Behavior? examined whether there were significant differences in Facebook users’ perception of their online behavior compared to their actual online behavior, and how their Facebook behavior compared to their offline behavior, as reported by informants. The researchers found “substantial similarity between online and everyday traits and social behavior,” and that “impressions of personality based on Facebook pages seem to reflect the actual and not the ideal self.” Similarly, they found that Facebook users’ perceptions of their own behavior on Facebook were consistent with their actual behavior. Public social networking sites are a more meaningful source of data for making predictions about individuals when the virtual and actual self are so similar.

Facebook and personality

Facebook has over 1 billion users worldwide, and is almost ubiquitous among younger age brackets. One of the interesting characteristics of Facebook discussed by Caci et al. in their 2014 study Personality Variables as Predictors of Facebook Users is “nonymity,” the opposite of anonymity, or the tendency of Facebook users to post under their real names and use their actual photos rather than anonymous handles and avatars.

Caci et al. looked at personality variables which included openness, conscientiousness, extraversion, agreeableness, and neuroticism, finding that high neuroticism was associated with frequent, longer daily use. They speculate that a “surveillance function tendency” may explain why “nervous and emotionally unstable individuals might try to control what is going on online as often as they can,” also suggesting that impulsivity or boredom may play a role. Conscientiousness was found to correspond to fewer, shorter sessions, with the researchers theorizing that “discipline in daily activities and devotion to work or family” could account for the more minimal involvement in social networking sites. Openness was associated with early adoption and larger groups of friends, echoing previous research with similar findings.

Three basic Twitter counts predict personality

Twitter is fundamentally different from Facebook in that it tends to be more public and open, while Facebook users typically “friend” people that they already know. A study by Quercia et al. classified four types of Twitter users:

• “Listeners” who follow many users
• “Popular” users who are followed by many
• “Influencers”, and the
• “Highly read” who are listed in multiple tweets

Popular users and Influencers tended to be extroverted and emotionally stable, while Influencers scored high on “conscientious” traits like organization. Three counts of basic public information were all that was needed to predict personality with a surprising degree of accuracy: the number the user is following, the number of followers, and listed counts (number of times the user has been listed in others’ reading lists).

Privacy implications

The authors of the Twitter study liken the privacy implications of their discovery to PleaseRobMe.com, a site devoted to “raising awareness about over-sharing” which demonstrates the dark side of seemingly harmless information: By posting Foursquare and Twitter updates with one’s whereabouts, one (unintentionally) implies that one is away from home, a piece of information which could be advantageous to potential burglars.

Similarly, regarding the predictive power of Facebook “likes”, Kosinski et al. reported that the Colbert Report and “science” were among the best predictors of high intelligence, while liking Sephora, “I love being a mom”, Harley Davidson, and Lady Antebellum were predictive of low intelligence. Likes of Wicked the musical and MAC cosmetics were associated with male homosexuality, whereas Wu-Tang Clan, Shaq, and “Being Confused After Waking Up from Naps” were predictive of male heterosexuality. The Twitter or Facebook user may think they are guarding more personal information when they like a product or “check in” to a place, yet the systematic collection and use of this data can often accurately predict user intelligence, sexual orientation, race, addictions, and other personal information.

Narcissism and social networking sites

It has been observed that social media has certain features which may be particularly appealing to narcissists. For example, Facebook “likes” and other features can serve as a frequent source of the affirmation narcissists seek. Sanja Kapidzic points out that “Narcissism, especially, is linked to prominent aspects of self-presentation such as the frequency of status updates, or the amount of self-promoting content displayed.” Social networking sites “provide narcissists with both an audience and a stage for highly-controlled self-presentation.”

A recent rash of Kim Kardashian Instagram “selfies” of her navel framed by beaches in Mexico appears to suggest to us anecdotally what Kapidzic’s research on Narcissism as a Predictor of Motivations Behind Facebook Profile Picture Selection concluded: that narcissism is a predictor of the motivations behind choices of profile pictures intended to emphasize the attractiveness and personality of the user.

Kapidzic reiterates that “studies analyzing self-presentation in various online environments have found that users strategically manipulate visual cues to reflect an ideal, rather than their actual, self, also citing the widely acknowledged tendency of many online dating service subscribers to choose more favorable representations of themselves to appear younger, thinner, or more attractive. Other research found that content with an overt mission to persuade readers of the user’s positive traits are also associated with narcissism.

Some of the takeaways from these studies are that seemingly nonspecific or impersonal data can often be used to draw fairly accurate conclusions of a very personal nature. Also, healthcare providers, who are increasingly interacting with patients via nontraditional channels, should be aware of the insight that can be gleaned from social media use, which is no longer just for marketing professionals. While certainly no substitute for face to face interaction, physicians and adolescent counselors should be aware that social media can serve as a window for information regarding social support systems, substance abuse, and other individual behavior.

References

Caci B, Cardaci M, Tabacchi ME, & Scrima F (2014). Personality variables as predictors of Facebook usage. Psychological reports, 114 (2), 528-39 PMID: 24897905

Ivcevic Z, & Ambady N (2013). Face to (face)book: the two faces of social behavior? Journal of personality, 81 (3), 290-301 PMID: 22812602

Kapidzic S (2013). Narcissism as a predictor of motivations behind Facebook profile picture selection. Cyberpsychology, behavior and social networking, 16 (1), 14-9 PMID: 23249240

Kosinski, M., Stillwell, D., & Graepel, T. (2013). Private traits and attributes are predictable from digital records of human behavior Proceedings of the National Academy of Sciences, 110 (15), 5802-5805 DOI: 10.1073/pnas.1218772110

Quercia D, Kosinski M, Stillwell DJ, Crowcroft J (2011.) Our Twitter profiles, our selves: Predicting personality with Twitter. IEEE International Conference on Social Computing. Proceedings of the 2011 IEEE International Conference on Privacy, Security, Risk, and Trust, and IEEE International Conference on Social Computing, pp 180–185.

Image via Nopporn / Shutterstock.