András Vörös
Department of Social Statistics, University of Manchester
SOST71032 Social Network AnalysisWritten assignment
Introduction to Exponential Random Graph Models (ERGMs)
An example
(Thanks to Bálint Néray and Zsófia Boda for a lot of materials on these slides!)
Context: a company (entertainment industry)
Nodes: 38 executives
Network: communication between nodes
Tie: another executive with whom it was important to communicate to get work
completed effectively (binary, directed ties)
Individual attributes:
experience (number of projects actors have been involved with)
level of seniority
office membership
Dyadic attributes: advice (from whom actors received advice)
2
An example: communication in a company
3
An example: communication in a company
Lusher, Koskinen & Robins 2013, p. 38
Important communication ties
4
An example: communication in a company
Lusher, Koskinen & Robins 2013, p. 38
Mutual communication ties
5
An example: communication in a company
Lusher, Koskinen & Robins 2013, p. 38
Important communication ties
Node size by experience
6
An example: communication in a company
Lusher, Koskinen & Robins 2013, p. 38
Important communication ties
Node color by seniority
7
An example: communication in a company
Lusher, Koskinen & Robins 2013, p. 38
Important communication ties
Node shape by office membership
8
An example: communication in a company
Lusher, Koskinen & Robins 2013, p. 38
Communication ties multiplex
with advice
9
An example: communication in a company – ERGM results
Lusher, Koskinen & Robins 2013, p. 43
10
An example: communication in a company – ERGM results
Lusher, Koskinen & Robins 2013, p. 43
11
An example: communication in a company – convergence
Lusher, Koskinen & Robins 2013
With covariates only
12
An example: communication in a company – convergence
Lusher, Koskinen & Robins 2013
With Markov effects only
13
An example: communication in a company – convergence
Lusher, Koskinen & Robins 2013
With higher-order effects
(and covariates)
14
Please continue with the next topic