Social Network Analyses for Longitudinal Data Estimating Network by forrests


									Social Network Analyses for Longitudinal Data: Estimating Network Dynamics and Individual Behavioral Trajectories (2 ECTS)
Teacher: William Burk, Behavioural Science Institute, Radboud University Nijmegen, The Netherlands ( Place: Jyväskylä, University of Jyväskylä, Agora (Mattilanniemi 2) Time: 21-23rd October 2009 (from Wednesday to Friday) Credits: 2 ECTS for those who attending all the three course days

Course description: This intensive three-day course is designed to provide a general overview of longitudinal social network analysis, with a particular focus on actor-based modeling approach implemented with the software program, Simulation Investigation for Empirical Network Analysis (SIENA: Snijders, Steglich, & Schweinberger, 2007). The course includes both lectures and computer demonstrations. Lectures will quickly cover basic concepts and terminology, primarily focusing on conceptual issues of actor-based models and practical issues, ranging from model specification and interpretation of parameter estimates to data formatting and troubleshooting. Multi-level and multiple group extensions of actor-based models, and new developments in these techniques will also be described in the lectures. Computer demonstrations will provide students with an opportunity to perform SIENA analyses using data that will be provided. Students are not expected to have prior familiarity with these statistical techniques, but students are expected to complete assigned reading before the start of the course. Required reading: Burk, W.J., Steglich, C.E.G., and Snijders, T.A.B. (2007). Beyond dyadic interdependence: Actor-oriented models for co-evolving social networks and individual behaviors. International Journal of Behavioral Development, 31, 397404. Snijders, T.A.B., Steglich, C.E.G., and van de Bunt, G.G. (2009). Introduction to actorbased models for network dynamics. Social Networks, in press. All reading material, data examples, as well as the SIENA software (and R) are available online:

TIMETABLE OF THE COURSE October 21 Afternoon session I at 12.15-15.00 pm (ROOM: Agora’s Aquarium): What is SNA and how can it be incorporated into developmental research?  Introduction to longitudinal SNA: Pioneers, concepts, and terms Coffee break at 15.00-15.30 pm Afternoon session II at 15.30-18.00 pm (ROOM: Agora’s Aquarium): Actor-based models of network-behavioral co-evolution  Pre-cursors and advantages  Data requirements and model specifications October 22 Morning at 9.15-12.00 am (COMPUTER CLASSROOM: Ag B113.1): Touring SIENA in R: A walkthrough  Step-by-step procedure for model identification and specification  An example: Deviance in Swedish adolescent friendship networks Lunch break at 12.00-13.15 pm Afternoon at 13.15-16.00 pm (COMPUTER CLASSROOM: Ag B113.1): Strategies for examining multiple networks  Structural zeroes, multiple groups, meta-analyses October 23 Morning at 9.15-12.00 am (ROOM: Agora’s Aquarium): Lessons learned: Avoiding pitfalls and troubleshooting problems  Categories, missingness, and model convergence  Network delineation and heterogeneity (over time and within networks) Lunch break at 12.00-13.15 pm Afternoon at 13.15-16.00 pm (COMPUTER CLASSROOM: Ag B113.1): Final thoughts  A sneak peek in the future…  Q&A

To top