Stochastic Models

Overview

This course is taught by Robert Krause and runs from Monday June 12 through Friday June 16. Each day runs 10am-12 and 12:45-2:45 ET (New York time). In addition, there is an optional session each day from 3:30-5pm for questions, problems, etc. The workshop offers a practical introduction to cross-sectional ERGM (p* models) and longitudinal SIENA models (SAOM) using R. 

The course starts with a discussion and overview of statistical inference for complete network analysis, and some simple statistical tests are run in class. We then discuss more complex models, with a specific focus on ERGM (p* models) for cross-sectional social network data and SIENA models (SAOM) for longitudinal social network data. The principles of relational events and related models will also be touched upon.

The course aims to be interactive, using breakout sessions for the exercises. Participants will receive homework; which includes running further analyses and interpreting results in small groups. The results will then be discussed again in the next meeting. Please read the software information below about the programs that we will use. The course's primary objective is to ensure participants are able to run and interpret statistical models. However, to allow participants to understand the logic of both ERGM and SIENA models (SAOM), the first day will focus on the foundations and logic of statistical models and will therefore be more “theoretical”.

Prior familiarity with R is helpful but not required.

Software

• R for running "statnet" and “RSiena” (http://cran.r-project.org/). Download R for Windows or other platform as needed.

• Specific packages (RSiena and statnet/sna) will best be downloaded from the internet at the start of the course (to ensure everyone has the same version!). Detailed instructions will follow by email a week before the start of class.

• For your own convenience, it might be helpful if you have two monitors (or two pcs) available, so you are simultaneously able to see the programs and “attend” class). 

• See our software page for any additional details.

Readings

• Agneessens, F. (2020). Dyadic, nodal and group-level approaches to study the antecedents and consequences of networks: Which social network models to use and when. In The Oxford Handbook of Social Networks. Oxford University Press.

• Borgatti, S. P., Everett, M. G., Johnson, J. C., & Agneessens, F. (2022). Analyzing Social Networks Using R. SAGE. 

• Robins, G., P. Pattison, Y. Kalish, and D. Lusher (2007). On exponential random graph models for cross-sectional analysis of complete networks: An introduction to exponential random graph (p*) models for social networks. Social Networks, 29(2): 173-191 [pdf]

• Lusher, D., J. Koskinen, and G. Robins (eds.) (2013) Exponential Random Graph Models for Social Networks. Structural Analysis in the Social Sciences. New York: Cambridge University Press.

• Snijders, T.A.B., G. van de Bunt, G., and Ch. Steglich (2010). Introduction to stochastic actor-based models for network dynamics. Social Networks, 32: 44-60.

• Krause, R. W., Huisman, M., Steglich, C., & Snijders, T. (2020). Missing data in cross-sectional networks–An extensive comparison of missing data treatment methods. Social Networks, 62, 99-112.

• Krause, R. W., Huisman, M., & Snijders, T. A. (2018). Multiple imputation for longitudinal network data. Statistica Applicata-Italian Journal of Applied Statistics, (1), 33-57.

Instructor contact information

Robert Krause  <Robert.w.Krause@uky.edu>