REGRESSION-IN-RATIO ESTIMATORS IN THE PRESENCE OF OUTLIERS BASED ON REDESCENDINGM-ESTIMATOR

Abstract

In this paper, a robust redescending M-estimator is used to construct the regression-inratio estimators to estimate population when data contain outliers. The expression of mean square error of proposed estimators is derived using Taylor series approximation up to order one. Extensive simulation study is conducted for the comparison between the proposed and existing class of ratio estimators. It is revealed form the results that proposed regression-in-ratio estimators have high relative efficiency (R.E) as compared to previously developed estimators. Practical examples are also cited to validate the performance of proposed estimators.