RESPONDENT-DRIVEN SAMPLING MODEL FOR SAMPLING AND ESTIMATION OF HIDDEN POPULATION

Abstract

The Respondent-Driven Sampling (RDS) has emerged as an effective method for recruiting hidden populations, contributing to a notable increase in its global usage. Its ability to provide reliable population estimates, supported by established statistical models like the Naïve estimator, Salganik and Heckathorn RDS estimator (SH-RDS), Volz and Heckathorn RDS estimator (VH-RDS), Gile Sucessive Sampling estimator (G-SS), enhances its robustness. A simulation study involving networked populations was conducted, comparing the performance of a proposed RDS estimator with that of existing ones across various sample sizes and degree distributions. The results revealed that a simulation from a population of 500 with a samplze of 150 showed the following gender proportions. The