Using Snowball-Based Methods in Hidden Populations to Generate a Randomized Community Sample of Gang-Affiliated Adolescents

• NCJ Number: 209299
• Journal: Youth Violence and Juvenile Justice Volume: 3 Issue: 2 Dated: April 2005 Pages: 151-167
• Author(s): Rebecca D. Petersen; Avelardo Valdez
• Date Published: April 2005
• Length: 17 pages
• Grant Number(s): R49/CCR615627-01
• Sponsoring Agency: Ctr's for Disease Control and Prevention (CDC)
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

This article presents a research methodology for generating a randomized community-based sample of Mexican-American adolescent females involved in gangs.

Abstract

Hidden populations, such as gang members, refer to subsets of a population whose membership is not readily distinguished or specified based on existing knowledge or for sampling capabilities. "Snowball" recruitment and sampling can be effective with such groups. Snowballing is based on social network logic whereby people are linked by a set of social relationships and contacts. Asking respondents to name or mention other people who qualify for a study population takes advantage of such a social network. The research methodology featured in this article involved in-depth interviews with 150 gang-affiliated Mexican-American adolescent females ages 14 to 18. The interviews were conducted from the fall of 1998 through the summer of 2001 in San Antonio, TX. A sampling frame of 519 names of girls associated with 27 gangs was generated based on sustained contact with young men and the girls associated with these gangs. They were involved in a previous federally funded study on violence among male gangs in San Antonio. Based on rapport and trust established with many gang members involved in the previous study, a community researcher had "privileged access" to the target population. After estimating the number of affiliated girls for each gang through various mechanisms described in this article, a list of 519 names was composed of potential study participants. This article describes stratified random sampling for the study, the locating of potential participants, and the structure for the interviews. The multidimensional design presented provides a practical way to obtain a randomized community-based sample of gang-affiliated girls from various types of gangs in order to obtain a greater representation of this hidden population. 2 tables and 83 references