Keywords: network analysis, egocentric network data, exponential-family random graph models, HIV
Principled statistical estimation of network models from egocentrically sampled networks now makes it possible to parameterize complex stochastic epidemic simulations using data from traditional sample surveys. We present a new framework for estimating ERGMs (and their temporal counterparts, TERGMs) from egocentrically sampled networks. Such data are routinely collected by the International Demographic and Health Surveys and the US National Survey of Family Growth. We demonstrate their potential to inform local policy in an application to HIV transmission among heterosexuals in King County, WA.