General Regression Methods for Respondent-Driven Sampling Data

TL;DR

This paper introduces a new regression methodology tailored for Respondent-Driven Sampling data, addressing network dependence and homophily, and applies it to study HIV treatment optimism among MSM in Montreal.

Contribution

It develops a general regression framework for RDS data that accounts for network structure and homophily, filling a gap in multivariate modeling strategies.

Findings

Identifies socio-demographic predictors of HIV treatment optimism.

Provides a principled approach for regression analysis with RDS data.

Demonstrates the methodology on real-world RDS data from Montreal.

Abstract

Respondent-Driven Sampling (RDS) is a variant of link-tracing sampling techniques that aim to recruit hard-to-reach populations by leveraging individuals' social relationships. As such, an RDS sample has a graphical component which represents a partially observed network of unknown structure. Moreover, it is common to observe homophily, or the tendency to form connections with individuals who share similar traits. Currently, there is a lack of principled guidance on multivariate modeling strategies for RDS to address homophilic covariates and the dependence between observations within the network. In this work, we propose a methodology for general regression techniques using RDS data. This is used to study the socio-demographic predictors of HIV treatment optimism (about the value of antiretroviral therapy) among gay, bisexual and other men who have sex with men, recruited into an RDS study in Montreal, Canada.