Extending the Ranked Exponential Random Graph Modeling Framework to Network Data From Subjective Ratings

• In Event: 2149 - Big Data and New Methods in Sociological Research

Abstract

Applying the statistical techniques used to model complete subjective interpersonal ranked order data in the exponential random graph (ERG) framework to valued rating data, we present a model that can successfully explain the valued data presented by Johnson et al. (2003) in their ethnographic exploration of network evolution in isolated social environments. We introduce new model diagnostics to aid in understanding model fit and aid in model selection. Implications for understanding network evolution are contextualized within Johnson et al.'s (2003) ethnographic findings for organizational behavior.

Authors

• Francis Lee, University of California-Irvine

• Pavel N. Krivitsky, University of Wollongong