Estimating the correlation in network disturbance models. (18th September 2021)

• Record Type: Journal Article
• Authors: Barbour, A D; Reinert, Gesine
• Editors: Gleeson, James
• Is Part Of: Journal of complex networks. Volume 9:Number 5(2021)
• Publication Date: 2021-09-18
• Subjects: network disturbance -- network autocorrelation -- maximum likelihood; Numerical analysis -- Periodicals; Computer networks -- Periodicals; Social networks -- Periodicals; 518.05
• DOI: 10.1093/comnet/cnab028
• Languages: English
• ISSNs: 2051-1310

Abstract

Abstract: The network disturbance model of P. Doreian (1989), expresses the dependency between observations taken at the vertices of a network by modelling the correlation between neighbouring vertices, using a single correlation parameter $\rho$ . It has been observed that estimation of $\rho$ in dense graphs, using the method of maximum likelihood, leads to results that can be both biased and very unstable. In this article, we sketch why this is the case, showing that the variability cannot be avoided, no matter how large the network. We also propose a more intuitive estimator of $\rho$, which shows little bias. The related network effects model is briefly discussed.