Empirical Investigation of Negative Influence
Using Stochastic Actor-Oriented Models for an empirically grounded agent-based model
Negative influence refers to the phenomenon in which people’s opinions become more different after interaction. This contrasts with the classical assumption of assimilation, or positive influence, where interaction leads individuals to become more similar. Negative influence is one of the fundamental mechanisms that may explain polarization, although empirical evidence for its existence remains mixed.
In this project, we aim to find empirical evidence for negative influence. We apply Stochastic Actor-Oriented Models (SAOMs) to the longitudinal network data from The Arnhem School Study (TASS), which records adolescents’ social networks and opinion dynamics. We find that negative influence can occur when the opinion distance between individuals is sufficiently large. For more information, see our recent publication.
This project is supported by NWO under the 2018 ORA grant ToRealSim. I am very fortunate to work with Prof. Andreas Flache (RUG) and Prof. Tom Snijders (RUG/Oxford) on this project. I am currently further investigating the sources of negative influence with Dr. Mike Lees (UvA) and Prof. Andreas Flache.