Types and Gender Composition of Social Networks: Their Influence on Adolescent Substance Use



Journal Title

Journal ISSN

Volume Title



This dissertation presents three separate studies designed to examine how the different factors and determinants known to influence adolescent alcohol and tobacco use and the gender compositions of different adolescent network types are associated with alcohol and tobacco use among adolescents. Additionally, the similarities and/or differences in networks of adolescent substance users and non-users are also examined.

First, a systematic review of empirical studies (n=48) employing social network analysis to examine adolescents? alcohol, tobacco, and other drugs (ATOD) use behavior will be presented discussing: (a) how this body of literature examined gender differences in ATOD use, (b) whether these network studies examine the gender composition of these adolescents? networks, and (c) what network affiliation types are used to characterize adolescent networks.

Secondly, descriptive characteristics and network analysis of the social networks of 10th grade substance using and non-using adolescents will be presented. Employing a bounded whole network approach, data was collected from 1,707 10th graders in a school district in Los Angeles, California. The students were asked different network questions in order to generate different network types. The network types elicited from these network questions are: Friendship, Admiration, Succeed, Popularity, and Romantic networks. Attributes and network characteristics of users and non-users across these five different networks are presented and the commonalities and/or differences are described also.

Lastly, using data from the same students, a quantitative analysis of the associations between interpersonal (e.g. age, gender) and interpersonal (e.g. parent and sibling substance use) factors, network measures, and gender composition of the networks and their alcohol and tobacco use will be discussed. These associations are then further examined across the five different types of networks mentioned above.

Prior to this study, research studies employing social network analysis did not attempt to examine the gender composition of the networks in which adolescents are embedded; and only a few other studies used networks other than friendship networks to characterize adolescent social networks. Thus, this study represents the first step towards addressing these limitations associated with examining how adolescents? social networks facilitate or constrain their substance use behavior and filling these apparent conceptual gaps.