Erik Brynjolfsson, a professor of information technology and director of the MIT Center for Digital Business at the Sloan School of Management, says there are “enormous opportunities” to use data more effectively both within the health-care system and within social-networking research. But, he cautions, “it’s fair to say that sorting out causality in social-network research is very difficult to do.” The best way to determine causality is through rigorous experimentation, with control groups examined alongside experimental ones, he says, and such an approach is rarely possible using retrospective information.
A few years ago, Christakis and colleagues examined data from the Framingham Heart Study—a decades-long study of thousands of people in Framingham, Massachusetts—and studied how obesity, happiness, and other phenomena spread among social networks. Christakis showed that weight gain in individuals increased the likelihood of weight gain for others in that person’s social network, but he could only guess about the mechanism at play.
Marketers already understand that online social networks like Facebook offer new opportunities to expand businesses, and that various software products, such as IBM’s Lotus Connections social software, help people analyze their existing relationships to boost efficiency. But the latest ideas may help them get healthier, too. “There’s a new niche,” says Noshir Contractor, a professor of behavioral sciences at Northwestern University, referring to the application of computational social-science insights to the business world. He has cofounded two consulting companies based on computational social science. “Of most of the social challenges we have, I think health is on the top of the areas that could benefit from network analysis.”
Lauren Cox is a reporter for Technology Review.