Elsevier

Social Networks

Volume 9, Issue 2, June 1987, Pages 171-186
Social Networks

QAP partialling as a test of spuriousness☆

https://doi.org/10.1016/0378-8733(87)90012-8Get rights and content
Under a Creative Commons license
open archive

Abstract

A test of spuriousness for structural data is proposed. Partials are calculated using OLS estimates. The test of significance is based on Hubert's QAP, a nonparametric permutation test. Results of Monte Carlo simulations indicate that statistical bias and efficiency characteristics of this procedure are very reasonable

Cited by (0)

☆

I wish to thank Ronald Breiger for his helpful comments on earlier drafts of this manuscript. Requests for reprints should be sent to the author at the Johnson Graduate School of Management, Cornell University, Ithaca, NY 14853, U.S.A.

Copyright © 1987 Published by Elsevier B.V.