Exploring the Influence of Open-Mindedness and Social Identity on Opinion Polarization in a Modified Hegselmann-Krause Model

In this article we further develop agent-based simulations on an extended Hegselmann-Krause (HK) model with social identity dynamics. In Kalvas et al. (2023) we introduced a model of opinion dynamics where agents were bounded in communication based on differences in opinion and identity, the latter of which was modeled as dynamic group assignments dependent on the state of the opinion space at any given time and the observer's sensitivity to opinion clustering. We had also previously presented regression results to demonstrate the effectiveness of identity as a polarizing factor. In this article we present an in-depth analysis of the macro-behavior of our model, specifically the effect on polarization of key variables such as open-mindedness (traditionally known as ε or uncertainty in the HK literature), and the granularity of social identity groups. We also examine the effect of the parameters of naturalistic distributions of these agent variables. We expand upon some known results in opinion dynamics such as the consensus-driving tendency of heterogeneity in ε, and also present results on how identity influences the relationship between openness and polarization.