Efficiency study of a robust regression-type estimator for population mean under different ranked set sampling methods with outlier handling

Abstract

Classical estimators often suffer from inefficiency when data is contaminated with outliers, leading to skewed results. To overcome this issue, we propose robust estimators for estimating the population mean in various ranked set sampling scenarios. Our research introduces an innovative technique that enhances Zaman estimators under simple random sampling (SRS) and extends these advancements to ranked set sampling (RSS) and median-ranked set sampling (MRSS) designs. Real and simulated datasets, incorporating outliers, are employed to compare the performance of the proposed robust estimator using robust regression methods, that is, LAD, LMS, LTS, Huber-M, Hampel-M, Turkey-M, and Huber-MM, against classical and Zaman estimators. The results demonstrate the superior performance of our robust regression-type estimator in effectively handling outliers.