Assessing logistic regression applied to respondent-driven sampling studies: a simulation study with an application to empirical data
Sandro Sperandei, Leonardo Soares Bastos, Marcelo Ribeiro‐Alves, Arianne Reis, Francisco Inácio Bastos
Abstract
The aim of this study is to investigate the impact of different logistic regression estimators applied to RDS studies via simulation and the analysis of empirical data. Four simulated populations were created with different connectivity characteristics. Each simulated individual received two attributes, one of them associated to the infection process. RDS samples with different sizes were obtained. The observed coverage of three logistic regression estimators were applied to assess the association between the attributes and the infection status. In simulated datasets, unweighted logistic regression estimators emerged as the best option, although all estimators showed a fairly good performance. In the empirical dataset, the performance of weighted estimators presented an unexpected behavior, making them a risky option. The unweighted logistic regression estimator is a reliable option to be applied to RDS samples, with a performance roughly similar to random samples and, therefore, should be the preferred option.