NCJ Number
225481
Journal
Canadian Journal of Criminology and Criminal Justice Volume: 50 Issue: 5 Dated: October 2008 Pages: 559-587
Date Published
October 2008
Length
29 pages
Annotation
Using longitudinal data on students’ friendship networks and their delinquent behavior in Dutch high schools, this study investigated whether similarity in delinquent behavior among friends was the results of a selection or of an influence process.
Abstract
Results indicate influence to be a general process without variation over the schools, while selection played a role in only 4 of the 16 schools. The study, shows that disentangling selection and influence processes leads to a more precise specification of when each of the two processes occurs. In view of the results, future research should inquire more deeply into the characteristics of contexts such as schools, as well as neighborhoods, where selection and influence processes take place and study their role in the evolution of networks and delinquent behavior. Most criminologists agree that delinquents prefer relationships with other delinquents. However, while there is certainly more similarity among friends than among non-friends regarding delinquent behavior, the nature of the underlying mechanism for this similarity remains an open question. Some researchers argue that the similarity is primarily the result of selection processes. Others argue that it is mainly the result of influence processes. This article contributes to the understanding of selection and influence processes. The study on which it is based (Dutch Social Behavior) used direct measurements of friendship networks and of delinquency and a newly developed technique for longitudinal network analysis in 16 Dutch high schools. Tables, figure, and references