Structure Matters: Network and Spatial Dependence for Community Dynamics

This session highlights statistical methods for dependent data in communities, where behaviors and outcomes are shaped by social ties and spatial adjacency. Papers address negative ties and friendship, longitudinal three-way perception networks, directed multiview ordinal ties, and challenges in estimating peer influence under selection. Complementary talks show how explicitly modeling spatial structure improves inference for place-based health and violence outcomes—illustrating how "structure-aware" models enable better community understanding and action.