Respondent-Driven Sampling: Risks and Benefits Of A Novel Sampling Strategy

Respondent-Driven Sampling (RDS) is an innovative sampling technique that has recently gained considerable popularity as a method for studying ?hidden? and ?hard-to-reach? populations with particular attention to populations at high risk of HIV infection. RDS can be very successful as a survey recruitment strategy for populations that are otherwise intractable to study, and has been used in at least 128 studies around the world. Yet despite the widespread use of RDS, there remain serious questions about the statistical validity of the methodology. In particular it is unclear whether or not samples collected using RDS can be used to infer population-level characteristics. This talk will give an introduction to the RDS methodology and the scientific issues that the method raises. In particular we will examine whether or not the RDS methodology can produce unbiased estimates of population characteristics. This talk will be of particular interest to researchers considering using RDS to recruit study participants. No special background in survey sampling will be required.