Social networks change behavior through clusters

While Luddites such as Dr. Deborah Peel keep telling patients to avoid online contacts, new research is revealing that social networks may be key to changing behavior, lowering the disease rate, and cutting health costs.

MIT management professor Damon Centola says it's the density of a social network that is the key to success in changing behavior. His paper on this, "The Spread of Behavior in an Online Social Network Experiment," is in the September 3 issue of Science. (Picture from Dr. Centola's page at MIT.)

The conclusion contradicts conventional wisdom about social networks, which is that long ties will spread information faster than dense ones.

Information may spread faster through long ties, Centola learned, but it will be internalized and acted upon if many people near to you are all saying the same thing.

To prove his point Centola created a social network of about 1,500 people, organized around health. Half the participants were organized as strangers along areas of interest, the other half were clustered with people close to them.

The rate of adopting to a specific behavior change, in this case signing up for a health forum, was four times greater in those organized as a cluster, he found.

"Social reinforcement from multiple health buddies made participants much more willing to adopt the behavior," he found. While disease may spread through a single contact, concerted action against it requires multiple contacts.

There are important implications for social networks aimed at health. Connecting people who are physically close may be more important than connecting people whose conditions are similar, for instance.

Change, in other words, requires an intervention. Even when you're online.