RESPONDENT-DRIVEN SAMPLING - A NEW APPROACH TO THE STUDY OF HIDDEN POPULATIONS

A population is ''hidden'' when no sampling frame exists and public ac knowledgment of membership in the population is potentially threatenin g. Accessing such populations is difficult because standard probabilit y sampling methods produce law response rates and responses that lack candor. Existing procedures for sampling these populations, including snowball and other chain-referral samples, the hey-informant approach, and targeted sampling, introduce well-documented biases into their sa mples. This paper introduces a new variant of chain-referral sampling, respondent-driven sampling, that employs a dual system of structured incentives to overcome some of the deficiencies of such samples. A the oretic analysis, drawing on both Markov-chain theory and the theory of biased networks, shows that this procedure can reduce the biases gene rally associated with chain-referral methods. The analysis includes a proof showing that even though sampling begins with an arbitrarily cho sen set of initial subjects, as do most chain-referral samples, the co mposition of the ultimate sample is wholly independent of those initia l subjects. The analysis also includes a theoretic specification of th e conditions under which the procedure yields unbiased samples. Empiri cal results, based on surveys of 277 active drug injectors in Connecti cut, support these conclusions. Finally, the conclusion discusses how respondent-driven sampling can improve both network sampling and ethno graphic 44investigation.