Who are you?

It’s a question so broad as to be ridiculous, but if you’ve posted a Facebook status update recently, there’s a chance you’ve already answered it.

Researchers from the University of Pennsylvania released findings suggesting it’s possible to predict our personality traits based on the language we use on FacebookPersonality, gender, and age in the language of social media: the open-vocabulary approach. Schwartz, H.A., Eichstaedt, J.C., Kern, M.L., et al. Positive Psychology Center, University of Pennsylvania, Philadelphia, Pennsylvania, USA. PLoS One 2013 Sep 25;8(9): e73791.. The study adds to a growing body of research suggesting that analyzing our activity on social media may be one of the best ways to learn about human psychology.

At the same time, all this research raises bigger questions about what it means to have a “personality” at all. Are people just datasets waiting to be plugged into an algorithm? Or, alternatively, are we more complex and unpredictable than any theory could possibly make room for? It’s possible the answer to both questions is “yes.”

What’s the Deal? Social Media, Language, and Personality

The UPenn research marks the largest study of personality and language use to date. The data came from 75,000 Facebook users, who completed a personality questionnaire and voluntarily made their Facebook status updates available to researchers. To measure personality, the researchers relied on the “Big Five,” a collection of traits including extraversion, agreeableness, conscientiousness, neuroticism, and openness. Each trait was represented by a “word cloud,” which included all the words and phrases that statistically predict that particular characteristic.

The findings range from “duh” to pretty revelatory. Extraverts were more likely to use words such as “party” and “weekend,” while introverts mentioned the Internet, anime, and manga. The word cloud for agreeable people included words such as “church” and “excited,” while the cloud for less agreeable people featured lots of profanity. Users deemed more emotionally stable referred more to social activities and athletic events. Older users were less likely to use the word “I” and more likely to mention family members. Researchers were even able to accurately judge users’ gender 92 percent of the time, based strictly on the language used in status updates.

This research comes on the heels of a similar study, which looked at the Facebook “likes” of over 58,000 participants who made their Facebook walls available to researchers and took a range of personality questionnaires and IQ testsPrivate traits and attributes are predictable from digital records of human behavior. Kosinski, M., Stillwell, D., Graepel, T. Free School Lane, The Psychometrics Centre, University of Cambridge, Cambridge, U.K. Proceedings of the National Academy of Science USA 2013 Apr 9;110(15):5802-5.. In this case, the results were even more striking. Based exclusively on the items people had “liked” on Facebook, researchers could discriminate between homosexual and heterosexual men 88 percent of the time, African Americans and Caucasians 95 percent of the time, and Democrats and Republicans 85 percent of the time. Researchers could even predict with 60 percent accuracy whether a person’s parents had split up before that person turned 21. One finding (which generated a lot of media attention) was the fact that people with high IQs tended to like curly fries.

These results seem to suggest that people are pretty predictable, if we know the right signs to look for. But this research also proves that, the more data psychologists accrue, the less certain they can be about their ability to understand the human mind.

Why It Matters: What Personality Research Says About Individuality

The researchers behind the UPenn study say their work has significant implications for the future of personality psychology. For years, psychologists have asked people to fill out lengthy surveys and questionnaires that supposedly measure the Big Five traits. But it seems like a more practical — and possibly more accurate — way to gauge personality is simply to have participants submit their social media profiles for analysis.

For the general public, this kind of work has even greater potential significance. My initial reaction upon hearing about the UPenn study was a combination of horror and disappointment. All this time, I’ve been thinking of myself as a special snowflake, a mysterious woman — and yet all someone has to do is simply glance at my Facebook profile to know if I’m emotionally stable, introverted, or intelligent? (Note: Nowhere on my profile do I mention curly fries.) That’s possibly more than some of my acquaintances might be able to discern. What happened to cultivating individuality, to the idea that it takes a lifetime just to get to know yourself?

On the other hand, this research suggests human personality is more complicated than we ever could have imagined. I spoke with Dr. Hansen Andrew Schwartz, who led the UPenn study, and he told me via email that even when he presented the research to a bunch of psychologists, no one could predict absolutely every word that was in every cloud. One important function of this research is to demonstrate how much of human thought and behavior still eludes us.

While some of the language in the word clouds is obvious (for example, the idea that an extravert would talk about a “party”), other connections are less straightforward (Who knew that an extravert would also be likely to use the word “tanning” or “thinkin?”). In a way, this research has allowed us to take a step beyond simplistic stereotypes (think: extraverts are outgoing and introverts are shy) that have shaped the past near-century of personality psychology.

The findings also generate hypotheses that create tremendous opportunities for further research. If emotionally stable people tend to talk more about athletics, is being part of a sports team a key component of a healthy lifestyle? If older people are more inclined to mention their family members, do relationships become more important as we age?

Perhaps most significantly, the insights we gain from the UPenn study and similar research may bridge some gaps between people from different social and cultural backgrounds. Dr. Martin Seligman, a researcher known as the “father” of positive psychology who helped lead the UPenn study, put it most succinctly in a university press release: “When I ask myself, ‘What’s it like to be an extrovert [sic]?’ ‘What’s it like to be a teenage girl?’ ‘What’s it like to be schizophrenic or neurotic?’ or ‘What’s it like to be 70 years old?’ these word clouds come much closer to the heart of the matter than do all the questionnaires in existence.”

In other words, instead of taking educated guesses at the kind of lifestyle that an introverted female or a homosexual male might lead, we now have evidence provided by those people themselves.

The Takeaway

As the number of people using social media increases and psychologists develop better tools for analyzing the data those users share, our understanding of personality as a complex phenomenon will only deepen. But if all this research suggests human personality is frighteningly quantifiable, it also suggests that a true understanding of what makes every person unique is perpetually just out of our reach. A quick glance at your Facebook wall might yield some good insight into who you are, but the more exciting part is how many questions it raises about who you might be, and ultimately who you might become.

Special thanks to Dr. Martin Seligman and Dr. Hansen Andrew Schwartz for their help with this article.

Do you think your Facebook wall is a good representation of your personality? Let us know in the comments below or tweet the author at @ShanaDLebowitz.