IMPROVED ESTIMATION PROCEDURES FOR THE MEAN OF SENSITIVE VARIABLE USING RANDOMIZED RESPONSE MODEL

In this article, we have considered the problem of estimation of mean of a rare sensitive quantitative variable, mu(A) while surveying the human population, using the scrambled response technique suggested by Ryu et al. (2005). We have presented a family of estimators mu(A lambda) to estimate the population mean mu(A) and examined its properties. The obtained optimal estimator depends on the optimal value of the constant, say lambda, which involves the different unknown population parameters, so proposed estimator has little practical application. To cover this deficiency, we have suggested two other estimators based on the estimated optimum value of lambda and discussed their properties. We have compared the proposed estimators with Ryu et al. (2005) estimator using simple random sampling and stratified random sampling protocols.