SpatialNets

Simulation of Large-Scale Interpersonal Networks

This project focuses on the simulation of large, spatially embedded networks, combining empirical findings regarding interaction in space with a stochastic modeling perspective. A large body of social science literature has established a strong, inverse relationship between distance in socio-physical space and the extent of interpersonal interaction. In this project, I have built upon this research by fitting hierarchical Bayesian models to various distance/tie frequency data sets, and then by sampling from the posterior predictive distributions to estimate structural properties conditional on population geometry. The resulting simulations suggest that ``small-world'' properties arise naturally from the distance/tie probability relationship, and that the form of this relationship is a power law. Extrapolation of residual model uncertainty to larger populations appears to confirm the prediction that physical distance accounts for the majority of uncertainty at large scales, and demonstrates that this effect does not depend on population density (though it does depend on geometry).