A stratified two-stage unrelated randomized response model for estimating a rare sensitive attribute based on the Poisson distribution. (2nd April 2016)

• Record Type: Journal Article
• Authors: Lee, Gi-Sung; Hong, Ki-Hak; Son, Chang-Kyoon
• Is Part Of: Journal of statistical theory and practice. Volume 10:Number 2(2016)
• Page Start: 239
• Page End: 262
• Publication Date: 2016-04-02
• Subjects: Poisson distribution -- rare sensitive attribute -- rare unrelated attribute -- stratified double sampling -- stratified sampling -- two-stage unrelated randomized response model; 62D05; 519.505
• DOI: 10.1080/15598608.2015.1126696
• Languages: English
• ISSNs: 1559-8608

Abstract

ABSTRACT: This article estimates the mean number of individuals with a rare sensitive attribute by using the Poisson distribution and stratified two-stage sampling and extends the Land et al. model to a stratified population. A rare sensitive parameter is estimated for the case in which stratum size is known, and proportional and optimal allocation methods are taken into account. We extended the Land et al. model to the case of an unknown stratum size; a rare sensitive parameter is estimated by applying stratified double sampling to the Land et al. model, and these two allocation methods are checked. Finally, the efficiency of the proposed model is compared with that of Land et al. in terms of the estimator variance.