Respondent-driven sampling as Markov chain Monte Carlo

Sharad Goel, Matthew J. Salganik

Research output: Contribution to journalArticlepeer-review

125 Scopus citations


Respondent-driven sampling (RDS) is a recently introduced, and now widely used, technique for estimating disease prevalence in hidden populations. RDS data are collected through a snowball mechanism, in which current sample members recruit future sample members. In this paper we present RDS as Markov chain Monte Carlo importance sampling, and we examine the effects of community structure and the recruitment procedure on the variance of RDS estimates. Past work has assumed that the variance of RDS estimates is primarily affected by segregation between healthy and infected individuals. We examine an illustrative model to show that this is not necessarily the case, and that bottlenecks anywhere in the networks can substantially affect estimates. We also show that variance is inflated by a common design feature in which the sample members are encouraged to recruit multiple future sample members. The paper concludes with suggestions for implementing and evaluating RDS studies.

Original languageEnglish (US)
Pages (from-to)2202-2229
Number of pages28
JournalStatistics in Medicine
Issue number17
StatePublished - Jul 30 2009

All Science Journal Classification (ASJC) codes

  • Epidemiology
  • Statistics and Probability


  • HIV surveillance
  • Hard-to-reach populations
  • Hidden populations
  • Importance sampling
  • Markov chain Monte Carlo
  • Respondent-driven sampling
  • Social networks
  • Spectral gap


Dive into the research topics of 'Respondent-driven sampling as Markov chain Monte Carlo'. Together they form a unique fingerprint.

Cite this