Improved randomized response models in additive scrambling

Abstract

In this investigation, we consider the problem of procuring the reliable information on stigmatizing variable. Two optional randomized response models are proposed for efficient estimation of mean and sensitivity level. A comparison of proposed optional randomized response models (ORRMs) with Huang (2010) and Gupta et al. (2010) models have been made. It is found that the proposed strategies perform better than those of Huang (2010) and Gupta et al. (2010) models whether it is estimation of mean or sensitivity level. A comparison is also made among the Huang (2010), Gupta et al. (2010) and proposed ORRMs to identify the best one. Percentage relative efficiency comparisons are also made to study the performance of the proposed procedures.