|
Activity Number:
|
467
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Social Statistics Section
|
| Abstract - #306433 |
|
Title:
|
Statistical Methodology for Longitudinal Social Network Data
|
|
Author(s):
|
Anton Westveld*+ and Peter Hoff
|
|
Companies:
|
University of Washington and University of Washington
|
|
Address:
|
Box 354322, Seattle, WA, 98102,
|
|
Keywords:
|
social networks ; latent variables ; structured covariances ; longitudinal data ; block Toeplitz
|
|
Abstract:
|
Social network data consist of measured relations occurring from interactions within a set of actors. This type of data allows for the empirical investigation of the interconnectivity of the actors, which is a cornerstone of social science theory. This paper will focus on data generated from the repeated interaction of pairs of actors (temporal dyadic data) resulting in an outcome for each actor at each time point. The network and temporal dependencies are modeled through a random effects approach resulting in a stochastic process defined by two weakly stationary covariance matrices. We will present several parameterizations of these matrices from a simple Markov structure to a very general block Toeplitz structure. A new latent variable approach to the estimation of general structured covariance matrices will also be discussed with the block Toeplitz structure presented as an example.
|
