Today, the journal Science published the second in a set of groundbreaking papers that will have a profound effect on how social networks can be designed to turbo-charge public health goals.
The publication, by MIT’s Damon Centola, recruited people into one of two identical online social networks dedicated to fitness. Participants in both networks had about a half-dozen connections, but in one group, those connections were random. In the other group, the connections were made to increase the chance you’d be linked to someone like you (age, sex, body-mass index).
The researchers then invited certain non-obese participants to try out a “health diary.” If the participant adopted the diary, his or her six friends would also have the opportunity. The question: would health diaries spread quicker through the random network, or the one where people were grouped by their demographic similarities, a phenomenon called homophily.
The answer wasn’t clear before the study: while we tend to take cues from people like us, changing our immediate community, behaviors have a way of not jumping easily from community to community. Sure, the healthier part of the network might get on board, but would grouping healthier people together make it tougher for the idea to spread to obese participants?
As it turns out, the answer was “no”: obese participants actually benefited the most, even though they were the most distant from where the diary concept was introduced. In the other group, with random connections, fewer obese participants adopted the diary, even though they may have been closer, network-wise, to where the diary was introduced. Being surrounded by similar individuals boosts the odds of changing a behavior.
A year ago, Centola published a related study that found that networks built around small communities, rather than varied and distant connections, adopted health behaviors faster.
Taken together, the studies show that it’s not who is in a network that counts, but how individuals arranged in that network. The obvious implication is for new health networks built around mutual support.
But there are huge consequences for communications, too. Digital communication means that information no longer always flows in a hierarchical way, from media outlet to consumer. Instead, information flows through a network that looks a lot more like Centola’s models. Understanding network relations will be as much a core competency for communications and integrated marketing as media relations was a generation ago. Centola’s work is the first step in teasing apart what nodes and vertexes mean in the real world.
That means the focus is on not only “what you know” and “who you know,” but — increasingly — “where you know”: where people and information fall in the myriads of networks that now define us.