Notes on a cautionary (tall) tale about respondent-driven sampling: A critique of Scott's ethnography
摘要
Serious methodological and ethical flaws are detailed in an ethnographic study of a respondent-driven sampling (RDS) project for drug users in Chicago. The study is also disconnected from the larger social context within which the project operated, and from the existing literature on human-subject problems the author claims he “discovered” about RDS – problems common to traditional outreach projects that researchers have known about and managed successfully for years. Due to an admitted bias in the author's sampling, and an eagerness to accept respondents’ claims uncritically, the author's results are not generalizeable to RDS projects operating in other cities, in Chicago itself, or even to the specific project studied.