Respondent-Driven Sampling (RDS)

Respondent-driven sampling (RDS) is a method for drawing probability samples of "hidden," or alternatively, hard-to-reach, populations. Populations such as these are difficult to sample using standard survey research methods for two reasons: First, they lack a sampling frame, that is, an exhaustive list of population members from which the sample can be drawn. Second, constructing a sampling frame is not feasible because one or more of the following are true: (a) The population is such a small part of the general population that locating them through a general population survey would be prohibitively costly; (b) because the population has social networks that are difficult for outsiders to penetrate, access to the population requires personal contacts; and (c) membership in the population is stigmatized, so gaining ...