Community structured network model based on age differences

Abstract This paper presents a statistical network model that extends the Block Two-Level Erdős–Rényi (BTER) framework by incorporating node-level attributes, specifically age, as a basis for community formation. To capture broader degree variability, the model integrates an adapted Chung-Lu mechanism for inter-community links. The resulting structure is particularly suited for analyzing small-sized networks where connections are shaped by age differences between individuals. We calibrate the model using empirical data from a Dutch women’s friendship survey, which captures age-based homophily in social relationships. Through simulation, we generate networks whose degree distributions and clustering coefficients closely resemble those observed in real-world community networks. This approach offers a flexible and interpretable framework for studying how node-level characteristics influence network topology, and provides a useful tool for agent-based modeling applications.