Nicholas Christakis, Sol Goldman family professor of social and natural science, internal medicine and biomedical engineering at Yale University, discussed his work in this area at the recent Nudgestock conference on behavioural science.
“We work mostly in developing-world villages, but the technology and the ideas that we invent and test are broadly applicable to many situations, including commercial applications,” he said. (For more, read WARC’s report: Three degrees of separation: the mathematics of (real) social networks.)
The typical public health intervention in such places, he explained, is delivered to, say, six random individuals with researchers returning months later to measure what proportion of those people responded.
But Christakis isn’t concerned about that so much as the knock-on effects. “I want to know what does everyone else in the village do when you give those six people intervention.” And why choose them randomly if you’re able to select “structurally influential individuals” at the centre of networks?
Christakis reported that using the last of these led to 75% of people taking up the health intervention within two weeks.
And in a new, bigger experiment, he is exploring whether it’s possible to create artificial tipping points. “Are there ways in which by thoughtfully choosing who’s structurally influential, we can drive adoption and change public health practices for the better in these communities?”
In marketing terms, he is attempting to shift the S-shaped infusion of the innovation curve to the left. “Can we find the right 20% of the people to select such that if you give them the intervention, it’s as effective as if you gave 100% of the people intervention?”
Sourced from WARC