Date of Original Version



Working Paper

Abstract or Table of Contents

Cohesion and structural equivalence are two competing social network models to explain diffusion of innovation. While considerable work has been done on these models, the question of which network model explains diff usion has not been resolved. This paper examines di ffusion of Caller Ring Back Tones in a cellular telephone network. Since these societal scale networks are very large (e.g., our call detail record data set has more than one million customers and one billion calls over a three months period from a large cellular service provider in India), the study of di ffusion in these settings require the development of methods to extract connected subpopulations from the network. We develop a novel technique to detect densely connected and self-contained components of the network and demonstrate through random restarts that the technique can enumerate distinct connected components in the network. Using a standard network auto-correlation model, we study the competing e ffects of cohesion and role equivalence on each of the distinct connected components detected using our sampling technique. The comparison of the results from the two models shows cohesion is more statistically signfi cant than role equivalence. The results have consistent pattern across di fferent sizes of subpopulation. We also conducted meta-analysis to summarize the size for both network eff ects. We find a signfi cant summarized eff ect and its size changes as E=I index changes. 2