A Class of Logarithmic Type Estimator Under Non-Response and Measurement Error Using ORRT Models

To analyse the finite population mean of the sensitive variable(s) more efficiently, a novel extended logarithmic type estimator is suggested in the joint influence of non-response and measurement error by utilizing optional randomized response models. For evaluating the properties of the proposed estimator, the bias and mean squared error (MSE) are retained to the first order of approximation. To illustrate the superiority of the suggested estimator over the class of existing ones, an empirical inquest and percent relative efficiency (PRE) are computed by employing the simulated data which has been carried out by using R software. Under specific real conditions, the proposed technique is observed to be more simple and effective than the existing class of estimators for estimating the sensitive trait.