TERGM: Exponential Random Graph Models for Dynamic Network Data
Exponential random graph models (ERGMs) are flexible statistical models for relational data that are capable of representing and identifying an extensive range of interdependencies common in political networks. Conventional ERGMs are limited to the analysis of cross-sectional network data. The temporal ERGM (TERGM) is a recently developed extension of ERGMs that is designed for longitudinal network data.
How long does it take for a tie to be reciprocated? Will a friend of a friend become a friend? Do new actors in a network prefer to form ties with actors who already have many ties? Fundamental questions of network dynamics such as these can be directly addressed with the TERGM.
This workshop will introduce the TERGM and demonstrate its application in the free and open source R statistical software. You will be provided with real-world longitudinal network data as well as R code to apply TERGMs to that data.
What you will need
- A laptop with Internet access
- An existing installation of R
- A recent version of the ERGM package installed
ERGM and Application in R
- Skyler J. Cranmer and Bruce A. Desmarais. Inferential Network Analysis with Exponential Random Graph Models. Political Analysis, 19(1):66—86, 2011.
- David R. Hunter, Mark S. Handcock, Carter T. Butts, Steven M. Goodreau, Martina Morris. ergm: A Package to Fit, Simulate and Diagnose Exponential-Family Models for Networks. Journal of Statistical Software, Vol. 24, Issue 3, May 2008.
- TERGM Method Derivation and Description
- Skyler J. Cranmer, Bruce A. Desmarais, and Justin H. Kirkland. Toward a Network Theory of Alliance Formation. International Interactions, 38(3):295—324, 2012.
- Bruce A. Desmarais and Skyler J. Cranmer. Micro-level interpretation of exponential random graph models with application to estuary networks. Policy Studies Journal, 40(3):402—434, 2012.
About the instructor
Bruce Desmarais received his PhD from UNC Chapel Hill in 2010 and joined UMass Amherst that year as an assistant professor in the Department of Political Science and a core faculty member in the Computational Social Science Initiative. Bruce's research focuses on the development and application of methods for the analysis of social, organizational and political networks. Applications in his work include international security, collaboration among legislators, organizational communication networks, and the intersection of scientific and policymaking expertise networks. Bruce regularly teaches interdisciplinary courses in network analysis at UMass Amherst, in research training institutes and at professional conferences