An innovative optimal randomized response model using correlated scrambling variables.

This paper introduces an innovative and optimal randomized response model by making use of correlated scrambling variables for estimating the population mean of a sensitive variable. The resultant estimators are found to be unbiased. Variance expressions of the proposed estimators have also been derived. Analytical as well as empirical evidences are provided in favour of the proposed estimators relative to their competitors. A possible extension to a general method is also suggested. [ABSTRACT FROM AUTHOR]