Basing their study on original datasets collected through field-work and data compilation, Johnson, at the University of Miami, and his colleagues in China and elsewhere in the US, focus on how even different social groups follow similar patterns in how they form. Their study, which appears in Physical Review E, presents a comparison between empirical distributions of group sizes with simple theoretical models that describe group formation. They find the data can best be explained using a model that says an individual can join a team if she/he can bring some new, complementary skills to the group. This interpretation goes against the idea that an individual will tend to mainly join—and remain comfortable in—groups of “like-minded” people.
Johnson et al.’s work is a good example of how the recent availability of large datasets about individual behaviors has the potential to bring a new quantitative dimension to social sciences. These data open the exciting possibility of identifying patterns of collective behavior and developing minimal models of human dynamics to explain them. – Marc Barthelemy