Network Structure and Diffusion

The diffusion of information, diseases, or other generic “goods” through a social network is determined by a joint function of the dyadic transmission probability and the overall structure of the network. But what network features matter? While recent theoretical work has highlighted a number of potential features, few have examined diffusion potential in empirical networks. This paper grounds theory on network diffusion by simulating diffusion over observed friendship networks. In addition to the well-known effects of average distance and local clustering, results show that the number of nodeindependent paths between pairs significantly moderates the effect of average distance.