Abstract for: The Impact of Aggregation Assumptions and Social Network Structure on Diffusion Dynamics

Diffusion problems in general, and innovation diffusion problems in specific, are one of the most frequently revisited issues in system dynamics domain. Although the models used for analyzing specific diffusion problems differ in details, in most cases a set of assumptions is recognized to be common. In this study, we aim to conduct a set of experiments in order to question the validity and potential impact of fundamental assumptions regarding the aggregation and social network structure. First, a generic model focuses on the impact of information dynamics that accompany the diffusion process of an innovation is introduced. The experiments conducted on the aggregate and individual-level versions of the model reveal that the behavior of the system converges to the aggregate model assuming perfect mixing as the network gets denser. Secondly, the change in diffusion levels as a consequence of changing network densities was monotonic. However, direction of change was different for different groups of scenarios tested. In other words, in some cases diffusion level increases as the network gets denser, while in some other cases the opposite is observed.