Generating networks of illegal drug users using large samples of partial ego-network data

Abstract

Use of drugs, including illegal ones such as marijuana and cocaine, is primarily a socially learned behavior. Friendship relationships often serve as the means by which youths are given the opportunity to try and continue using a drug. Counter-forces against use can also include friends, but often consist of close ties to parents. Since current research into youth drug use and abuse tends to focus on localized social networks of youths, we have yet to understand general characteristics of drug networks and how they might vary across subpopulations, such as youths of different cities. In this paper, I explore a method of sampling large networks of youths using the National Household Survey on Drug Abuse (NHSDA). This method enables me to obtain population level network measures and assess risk of contact between non-users, users, and sellers and how that risk varies across different metropolitan statistical areas (MSAs). © Springer-Verlag Berlin Heidelberg 2004.