Uncovering the Structural Underpinnings of Political Discussion Networks: Evidence From an Exponential Random Graph Model

Abstract

How do individuals structure their political discussion networks, and what factors systematically shape such patterns? While much research has focused on the effect of personality traits and one's motivations, abilities, and opportunities, network-structural factors present different principles of tie formations. Evidence from an Exponential Family Random Graph Model and the meta-analysis of an Exponential Random Graph Model (ERGM) from 20 different groups' networks indicated that the pattern of political discussions was shaped by general discussion and network-endogenous structural processes. Results suggest that informal discussions of individuals serve as the important foundations of political behavior, and the processes of citizens' everyday political interactions emerge through complex interactions that cannot be regressed to mere individuals' predispositions or pure social selection processes.