Extension of Mangat Randomized Response Technique Using Alternative Beta Priors

Abstract

In this study, an extension of Mangat Randomized Response Technique using alternative beta priors has been considered and new Bayes estimators of population proportion of respondents possessing stigmatized attribute were developed when data were gathered through administration of survey questionnaire on induced abortion on 300 matured women in the metropolis. Dominance picture of the proposed Bayes estimators has been portrayed for a wide range of values of population proportion assuming alternative Beta distributions as Prior information. It is observed that the proposed Bayes estimators performed better than the Bayes estimator proposed by Hussain et al [15] when a simple Beta prior was used for small, medium as well as large sample sizes respectively. This is evident as our proposed Bayes estimators have least mean squared errors (MSEs) as π approaches one.