A Simulation Study: Robust Ratio Double Sampling Estimator of Nite Population Mean in the Presence of Outliers

Abstract

In this study, we suggest a family of ratio estimators for the population mean parameter using various robust regression techniques. These robust regressions techniques are Huber MM, Last Trimmed Square (LTS), and Least Median Square (LMS) estimates.We evaluate the performance of estimators in terms of the Mean Square Error (MSE),and we compare the efficiency of our proposed robust-regression-ratio-type estimators with existing estimators under the optimal conditions. These comparisons show that our robustratio-type estimators give more excient results than the existing estimators under double sampling. In addition, the simulation and the empirical studies based on a data set that includes unusual observations show that our proposed estimators have a lower MSE than the existing estimators.