Actions

Difference between revisions of "Handcock abs"

From Santa Fe Institute Events Wiki

m
m (fixed date)
Line 6: Line 6:
 
'''Organizers:''' [http://www.santafe.edu/~girvan/ Michelle Girvan] (University of Maryland) and [http://www.santafe.edu/~aaronc/ Aaron Clauset] (Santa Fe Institute)
 
'''Organizers:''' [http://www.santafe.edu/~girvan/ Michelle Girvan] (University of Maryland) and [http://www.santafe.edu/~aaronc/ Aaron Clauset] (Santa Fe Institute)
  
===Saturday, January 11, 2008===
+
===Saturday, January 12, 2008===
  
 
11:00 - 11:40 '''Mark Handcock''' ([http://www.stat.washington.edu/handcock/ homepage])
 
11:00 - 11:40 '''Mark Handcock''' ([http://www.stat.washington.edu/handcock/ homepage])

Revision as of 22:35, 3 January 2008

Workshop Navigation


Is There a Physics of Society? January 10-12, 2008, Santa Fe NM

Organizers: Michelle Girvan (University of Maryland) and Aaron Clauset (Santa Fe Institute)

Saturday, January 12, 2008

11:00 - 11:40 Mark Handcock (homepage)

A Longitudinal Model of Network Formation: Heider's theory of Balance vs Simmel's triadic formation

Network models are widely used to represent relational information among interacting units and the implications of these relations. In studies of social networks recent emphasis has been placed on random graph models where the nodes usually represent individual social actors and the edges represent a specified relationship between the actors. Much progress has been made on Exponential family models for cross-sectional networks, and some has been made on related models for networks observed longitudinally.

A fundamental goal of social network theory is to represent the processes of network formation over time. The theory of balance developed by Heider posits that network evolve towards structural balance. We introduce a model that represents the level and dynamics of balance in a network. We extend the model to represent an alternative theory of network evolution based on triadic forms developed by Simmel.

We develop Bayesian inference for the model and present a Markov Chain Monte Carlo (MCMC) algorithm to implement it. We apply the model to longitudinal data and assess the empirical support for the two theories.

This is joint work with David Krackhardt and Martina Morris.