An ameliorated two-stage randomized response model for estimating a rare stigmatized characteristic using Poisson distribution

Abstract EN

The present work sheds light on the estimation procedure of a mean number of persons in the population bearing a rare sensitive characteristic using the Poisson distribution.

A modified two-stage randomized response model for the rare sensitive characteristic is used to acquire the truthful response.

Subsequently unbiased estimators are proposed for two situations when the information on another supplementary rare non-sensitive characteristic is known as well as unknown.

The variances of the proposed estimators and their estimates are derived.

Empirical studies are executed to show the dominance of the proposed estimators over some contemporary estimators.