Studies in respondent-driven sampling : Directed networks, epidemics, and random walks

Respondent-driven sampling (RDS) is a link-tracing sampling methodology especially suitable for sampling hidden populations. A clever sampling mechanism and inferential procedures that facilitate asymptotically unbiased population estimates has contributed to the rising popularity of the method. The papers in this thesis extend RDS estimation theory to some population structures to which the classical RDS estimation framework is not applicable and analyse the behaviour of the RDS recruitment process. At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: In press. Paper 3: Accepted. Paper 4: Manuscript.