Recent Developments in Respondent-Driven Sampling

Explore recent advancements in respondent-driven sampling techniques through this 32-minute colloquium talk by Mamadou Yauck from Université du Québec à Montréal. Delve into the challenges of surveying hidden populations and learn about the Engage study's application of RDS. Examine the tree representation of samples, post-recruitment adjustments, and finite population inference in RDS. Investigate regression methods for RDS, including model setup, data-generating processes, and fitting models to observed data. Consider the debate on weighting in RDS regression and discover the neighborhood bootstrap approach. Gain insights into the latest publications and engage in a discussion on the future of respondent-driven sampling in public health research.

Intro
Surveying hidden populations
Respondent-Driven Sampling
The Engage study
Tree representation of the sample
Main challenges
RDS: sampling over a network
Post-recruitment adjustments
Finite population inference for RDS
Regression methods for RDS: challenges
Regression methods for RDS: review
Model setup
The underlying, data-generating model
Fitting the model to the observed data
To weight or NOT to weight?
Latest publication on RDS regression
Overview & review
The neighborhood bootstrap
Discussion
References