EMPIRICALLY VALIDATING THE THRESHOLD DISTRIBUTION ASSUMPTION IN COMPLEX CONTAGION

Abstract

Diffusion research based on the threshold model made assumptions about the threshold distribution. Yet the assumptions were rarely tested. This study estimated the threshold distribution of people in networks using four empirical datasets. The results showed that the threshold distributions in these datasets were not as usually assumed. The research suggests that many of the conclusions of simulation work done with the threshold model could be based on untrue assumptions.