Efficient Searls predictive estimators for the computation of mean under ranked set sampling: an application to COVID-19 data

This study proposes an effective method for estimating the population mean of a primary variable through known subsidiary variable parameters in ranked set sampling (RSS) for predictive estimation. Utilizing a modified Searls technique, it achieves efficient estimation while analyzing sampling properties like bias and mean squared errors (MSE) up to an order-one approximation. Optimal Searls constants are determined, pinpointing the minimum MSE for the proposed estimator based on these optimized scalars. Theoretical evaluations compare the efficiencies of this estimator with competitors based on predictive estimation and MSE. Conditions for its superior efficiency over competing estimators are outlined. Additionally, numerical comparisons via Monte-Carlo simulations on synthetic data using R Studio, and application to COVID-19 data, validate the reliability of the proposed Searls ranked set predictive estimators.