Facebook, Google, and Twitter, among others, are enabling psychologists to mine giant data sets that allow mega-scale naturalistic observations of human behavior. The recent Society of Personality and Social Psychology convention offered several such “big data” findings, including these (some also recently published):
- “Computer-based personality judgments are more accurate than those of friends, spouses, or family.” That’s how Michal Kosinski, Youyou Wu, and David Stillwell summed up their research on the digital trail left by 86,220 people’s Facebook “likes.” As a predictor of “Big Five” personality test scores, the computer data were more significantly accurate than friends’ and family members’ judgments. (Such research is enabled by the millions of people who have responded to tests via Stillwell’s myPersonality app, and who have also donated their Facebook information, with guarantees of anonymity.)
- Another study, using millions of posts from almost 69,792 Facebook users, found that people who score high on neuroticism tests use more words like “sad,” “fear,” and “pain.” This hints at the possibility of using social media language analysis to identify people at risk for disorder or even suicide.
- Researchers are also exploring Smartphones as data-gathering devices. Jason Rentfrow (University of Cambridge) offers an app for monitoring emotions (illustrated here), and proposes devices that can sense human behavior and deliver interventions. In such ways, it is becoming possible to gather massive data, to sample people’s experiences moment-to-moment in particular contexts, and to offer them helpful feedback and guidance.
Amid the excitement over today’s big data, psychologist Gary Marcus offers a word of caution: “Big Data is brilliant at detecting correlation. . . . But correlation never was causation and never will be. . . . If we have good hypotheses, we can test them with Big Data, but Big Data shouldn’t be our first port of call; it should be where we go once we know what we’re looking for.”