Multilevel Longitudinal Analysis of Social Networks

TL;DR

This paper introduces a Bayesian multilevel extension of stochastic actor-oriented models for analyzing multiple network panels, demonstrated through a study of friendship and delinquency networks in schools.

Contribution

It develops a novel multilevel Bayesian SAOM framework for analyzing multiple network panels, addressing co-evolution of networks and behaviors.

Findings

Model successfully captures dynamic interdependence between friendship and delinquency networks.

Application to school data reveals significant network-behavior co-evolution.

Method enhances understanding of social network dynamics in longitudinal studies.

Abstract

Stochastic actor-oriented models (SAOM) are a broadly applied modelling framework for analysing network dynamics using network panel data. They have been extended to address co-evolution of multiple networks as well as networks and behaviour. This paper extends the SAOM to the analysis of multiple network panels through a random coefficient multilevel model, estimated with a Bayesian approach. This is illustrated by a study of the dynamic interdependence of friendship and minor delinquency, represented by the combination of a one-mode and a two-mode network, using a sample of 81 school classes in the first year of secondary school.