STERGM – Separable Temporal ERGMs for modeling relational dynamics over time with statnet

Statnet Development Team
U of Washington, UC Irvine, Penn State, UCLA
morrism@uw.edu

LIMITED to 30 Seats (This Workshop is already fully booked)

This workshop will provide an introduction to the estimation and simulation of dynamic ERGMs in statnet. This workshop will cover the statistical theory and methods for separable temporal ERGM modeling (STERGM) , and the associated diagnostics and visualization tools (NDtv) available for dynamic networks in statnet. STERGM can be used for both estimation and simulation of dynamic network data. An example of the type of “network movie” these tools can produce can be found at statnet.org/movies.

STERGMs can be estimated from both network panel data and cross-sectional, egocentrically sampled network data. Egocentrically sampled network data are often collected in settings when other network sampling approaches are not practical. When the alter-specific questions include retrospective information on the start and end of relationships, this can be treated as egocentric event history data. Such data can support a basic but surprisingly rich class of models for dynamic network analysis, relying on the available marginal information on relationship duration and cross-sectional network structure.

Prerequisites: This workshop will assume familiarity with R, and the network, SNA and ergm packages in statnet. The “Exponential-family Random Graph (ERG or p*) Modeling with statnet” workshop is recommended as preparation.

statnet is a collection of packages for the R statistical computing environment that supports the representation, manipulation, visualization, modeling, simulation, and analysis of network data. statnet packages are contributed by a team of volunteer developers, and are made freely available under the GNU Public License. statnet packages can be used with any computing platform that supports R (including Windows, Linux, and Mac), and they support statistical analysis of large networks, longitudinal network dynamics, and missing data.