Developing a field-intensive methodology for generating a randomized sample for gang research

Abstract

This paper was stimulated by concerns for methodological issues in gang research. The paper describes the strategies in developing a design for drawing a stratified proportional random sample of gang members and goes beyond limited non-probability quota samples, to compare gang members across communities and gang variations. First, this paper proposes procedures and techniques for drawing a random sample of gang subjects based on probability proportional to size quota. Second, it describes the role of ethnographicfield work and social mapping in defining sampling frames. Third, it presents the use ofMaplnfo in developing and mapping catchment areas. Fourth, it describes the procedures utilized in generating gang rosters and their role in generating interview sample quotas. And finally, it discusses the processes for developing sampling quotas and drawing of the sample.