Network Models of the Diffusion of Innovations

Thomas W. Valente
Institute for Prevention Research, Department of Preventive Medicine
Keck School of Medicine, University of Southern California, USA

Diffusion of innovations theory posits that new ideas and practices spread within and between communities via interpersonal influences. Network models of the diffusion of innovations have been developed in order to measure these interpersonal influences and debates exist regarding appropriate ways to estimate influence effects. In this talk I describe the many ways interpersonal influence has been measured and modeled. In addition to exposure/contagion effects, other interpersonal influences include structural equivalence, indirect ties, thresholds, Simmelian ties, homophilous ties and so on. The interaction between these individual level influences and macro network properties are discussed. The talk concludes with observations on the intersection between theory and interpersonal influence which leads to implications for using social network data for program implementation and intervention. Strategies and algorithms used to accelerate the diffusion of innovations and results from current studies are presented.