Goodness of fit of social network models

David R. Hunter, Steven M. Goodreau, Mark S. Handcock

Research output: Contribution to journalArticle

292 Citations (Scopus)

Abstract

We present a systematic examination of a real network data set using maximum likelihood estimation for exponential random graph models as well as new procedures to evaluate how well the models fit the observed networks. These procedures compare structural statistics of the observed network with the corresponding statistics on networks simulated from the fitted model. We apply this approach to the study of friendship relations among high school students from the National Longitudinal Study of Adolescent Health (AddHealth). We focus primarily on one particular network of 205 nodes, although we also demonstrate that this method may be applied to the largest network in the AddHealth study, with 2,209 nodes. We argue that several well-studied models in the networks literature do not fit these data well and demonstrate that the fit improves dramatically when the models include the recently developed geometrically weighted edgewise shared partner, geometrically weighted dyadic shared partner, and geometrically weighted degree network statistics. We conclude that these models capture aspects of the social structure of adolescent friendship relations not represented by previous models.

Original languageEnglish (US)
Pages (from-to)248-258
Number of pages11
JournalJournal of the American Statistical Association
Volume103
Issue number481
DOIs
StatePublished - Mar 1 2008

Fingerprint

Goodness of fit
Social Networks
Network Model
Statistics
Model
Social Structure
Network model
Social networks
Longitudinal Study
Graph Model
Vertex of a graph
Random Graphs
Maximum Likelihood Estimation
Demonstrate
Health
Evaluate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Hunter, David R. ; Goodreau, Steven M. ; Handcock, Mark S. / Goodness of fit of social network models. In: Journal of the American Statistical Association. 2008 ; Vol. 103, No. 481. pp. 248-258.
@article{ea257e658d314c2c94c74f2552c38e42,
title = "Goodness of fit of social network models",
abstract = "We present a systematic examination of a real network data set using maximum likelihood estimation for exponential random graph models as well as new procedures to evaluate how well the models fit the observed networks. These procedures compare structural statistics of the observed network with the corresponding statistics on networks simulated from the fitted model. We apply this approach to the study of friendship relations among high school students from the National Longitudinal Study of Adolescent Health (AddHealth). We focus primarily on one particular network of 205 nodes, although we also demonstrate that this method may be applied to the largest network in the AddHealth study, with 2,209 nodes. We argue that several well-studied models in the networks literature do not fit these data well and demonstrate that the fit improves dramatically when the models include the recently developed geometrically weighted edgewise shared partner, geometrically weighted dyadic shared partner, and geometrically weighted degree network statistics. We conclude that these models capture aspects of the social structure of adolescent friendship relations not represented by previous models.",
author = "Hunter, {David R.} and Goodreau, {Steven M.} and Handcock, {Mark S.}",
year = "2008",
month = "3",
day = "1",
doi = "10.1198/016214507000000446",
language = "English (US)",
volume = "103",
pages = "248--258",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "481",

}

Goodness of fit of social network models. / Hunter, David R.; Goodreau, Steven M.; Handcock, Mark S.

In: Journal of the American Statistical Association, Vol. 103, No. 481, 01.03.2008, p. 248-258.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Goodness of fit of social network models

AU - Hunter, David R.

AU - Goodreau, Steven M.

AU - Handcock, Mark S.

PY - 2008/3/1

Y1 - 2008/3/1

N2 - We present a systematic examination of a real network data set using maximum likelihood estimation for exponential random graph models as well as new procedures to evaluate how well the models fit the observed networks. These procedures compare structural statistics of the observed network with the corresponding statistics on networks simulated from the fitted model. We apply this approach to the study of friendship relations among high school students from the National Longitudinal Study of Adolescent Health (AddHealth). We focus primarily on one particular network of 205 nodes, although we also demonstrate that this method may be applied to the largest network in the AddHealth study, with 2,209 nodes. We argue that several well-studied models in the networks literature do not fit these data well and demonstrate that the fit improves dramatically when the models include the recently developed geometrically weighted edgewise shared partner, geometrically weighted dyadic shared partner, and geometrically weighted degree network statistics. We conclude that these models capture aspects of the social structure of adolescent friendship relations not represented by previous models.

AB - We present a systematic examination of a real network data set using maximum likelihood estimation for exponential random graph models as well as new procedures to evaluate how well the models fit the observed networks. These procedures compare structural statistics of the observed network with the corresponding statistics on networks simulated from the fitted model. We apply this approach to the study of friendship relations among high school students from the National Longitudinal Study of Adolescent Health (AddHealth). We focus primarily on one particular network of 205 nodes, although we also demonstrate that this method may be applied to the largest network in the AddHealth study, with 2,209 nodes. We argue that several well-studied models in the networks literature do not fit these data well and demonstrate that the fit improves dramatically when the models include the recently developed geometrically weighted edgewise shared partner, geometrically weighted dyadic shared partner, and geometrically weighted degree network statistics. We conclude that these models capture aspects of the social structure of adolescent friendship relations not represented by previous models.

UR - http://www.scopus.com/inward/record.url?scp=41249094522&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=41249094522&partnerID=8YFLogxK

U2 - 10.1198/016214507000000446

DO - 10.1198/016214507000000446

M3 - Article

AN - SCOPUS:41249094522

VL - 103

SP - 248

EP - 258

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 481

ER -