Why do we smile and say hello when we are introduced to someone, while the French kiss each other on the cheek? Why do we refer to SPAM as ‘SPAM’ instead of ‘junk email’? And why do we go to the very last position in a movie queue in Boston, but all crowd together to board the subway in Tokyo? The simple answer is that we do these things because everyone else does.
But Damon Centola, Ph.D., Associate Professor of Communication at the Annenberg School for Communications, is looking for a more nuanced explanation. By studying people’s interactions in online social networks, he hopes to begin to unravel the mysteries of how and when new norms will emerge in everyday life.
Spearheaded by Dr. Centola and Andrea Baronchelli, Ph.D., Assistant Professor of Mathematics at City University London, the “Name Game” study seeks to understand how a group of strangers can collectively produce coordinated patterns of behavior.
“In small groups, it’s relatively easy for people to understand and manage a set of shared expectations,” Dr. Centola says. “The process is more complex in larger groups, where arriving at a consensus or convention about basic behaviors—such as how to greet one another, which side of the road to drive on, and what constitutes a ‘fair’ bargaining outcome—can lead to massive social dilemmas.”
We experience social dilemmas more often than we realize, suggests Dr. Centola. Every time we question whether we should take a local road or the highway to avoid a traffic jam, we are making a decision that is interdependent with thousands of other people. Most days, we fail to solve the dilemma and wind up sitting in traffic. If we could coordinate our routes, however, we could find the best options for all drivers, and reduce overall traffic. Computer models can show that this coordination is possible, but real-world applications are more challenging.
To better understand how this process of solving a social dilemma works in the real world, Dr. Centola and colleagues developed the online Name Game. Participants are paired and are shown a picture of someone’s face that they are asked to give a name. If both parties agree on one name for the picture, they win a small amount of money. If they fail, they do not win anything, but can then see their partner’s choice. The game continues with new partners until all participants in the network have played a dozen or more times.
While participants keep whatever money they win, thus far Dr. Centola and his colleagues have found that the patterns of interaction of participants in the social network play a much larger role in the dynamics of collective success than the amount of money that they are paid.
“Remarkably, simple changes to the interaction patterns among online players can transform a group facing a social dilemma of ‘all against all’ into a population with universal coordination,” says Dr. Centola.
The Name Game is just one of several studies that Dr. Centola is conducting to understand the complex dynamics of social behavior. Since arriving at the Annenberg School this fall from the Massachusetts Institute of Technology, he has established the Network Dynamics Group to form a bridge between theoretical research on complex social systems and policy-relevant applications of online technologies.
While many researchers interested in social networks use Facebook or Twitter as data sources for studying human behavior, Dr. Centola’s strategy is to start from a virtual “state of nature” and build large-scale online experiments. This controlled approach permits him to demonstrate the cause-and-effect of how networks impact behavior, rather than just identifying correlations in existing networks. In a study published in the journal Science in 2010, for instance, he created a website for people to rate and share online health resources with other “network members.” He found that changing the network structure among the participants significantly improved the spread of a new online dieting tool across the population.
Looking ahead, Dr. Centola’s research will expand the use of online experiments to understand how we can use the Internet to improve public policy. Future work will address such topics as global warming and energy consumption practices, the impact of online social networks on health outcomes (such as diabetes management or medication compliance) and the effects of communications technologies on changing norms in developing nations.
“So many of society’s problems—from disease epidemics, to global warming, to corporate malfeasance—have their root in the beliefs and norms that support people’s decisions to, for example, get vaccinated, buy a hybrid car, or defraud shareholders,” says Dr. Centola.
“By anticipating the spread of these beliefs, we might allow policymakers to turn these complex social processes from a source of unpredictability into strategies for more effective public policies.”