Want to predict postpartum depression? Or to learn about how people cope with traumas? How about testing how often people think twice before sharing something potentially embarrassing? Social media and new technology are providing new ways to explore these many diverse areas of research.

Pairing together psychologists who want to understand human behavior with computer scientists who are natural problem-solvers is the perfect partnership, said Jamie Pennebaker at last night’s Presidential Symposium in Austin. “We need to start thinking about alternative models for research,” he said.

Eric Horvitz of Microsoft called big data the “future of social psychology.” He’s been working on ways to harness the power of Twitter, Facebook, and other social media to create public health tools. In recent work, he and colleagues used Twitter posts and patterns of activity to predict the onset of mood shifts in new moms that put them at risk of postpartum depression. “Postpartum depression is typically a surprise to people at a joyous time,” Horvitz said. Using their analysis, the hope is to design a private service for identifying moms at risk. (Read more about this work here.) His team will be unveiling new data next week that shows the predictive power of a one-time Facebook crawl of posts by new moms who were also surveyed about their postpartum time. It’s important to note here that these new moms opted in to let the researchers search their Facebook timelines.

Adam Kramer of Facebook also presented a slew of studies about how the words we choose and our patterns of network activity can reveal who we are and how we behave. His team has studied emotional contagion — whether our emotions spread to others — on Facebook, as it allows them to studied emotional expression without the researchers directing an interaction (and thereby, perhaps unintentionally, influencing it).

His team has also used Facebook to look at the question of how often people think  twice before sharing something. We’ve all had moments on email, Faecbook, Twitter, where we think: do I really want to send/post this? And we often talk ourselves down… but how often? Kramer and colleagues tracked when people on Facebook would type 5 characters or more and then waited 10 minutes or longer without posting anything. That was their way of identifying when someone Facebook thought twice before posting.

Turns out it happens often: Over a 17-day period, 71% of people drafted at least one post that they never posted. That amounted to 33% of posts and 13% of comments that never were. Not surprisingly, people were more likely to think twice if they were sharing the post with a larger audience. People do a “good job of policing” their own comments, Horvitz said.

There were lots more big ideas at this kick-off session. Kramer talked about studying familial communications “in the wild” (Interesting tid-bit: Parents and children become Facebook friends after a mean 371 days, he said) and Roxane Cohen Silver of UC-Irvine talked about how online surveys has revolutionized the way scientists study collective trauma. 

Welcome to the big data revolution!