Quantile Estimation Using Ranked Set Samples from a Population with Known Mean

Abstract

Ranked set sampling (RSS) is a cost-efficient technique for data collection when the<br><br>units in a population can be easily judgment ranked by any cheap method other<br><br>than actual measurements. Using auxiliary information in developing statistical<br><br>procedures for inference about different population characteristics is a well-known<br><br>approach. In this work, we deal with quantile estimation from a population with<br><br>known mean when data are obtained according to RSS scheme. Through the<br><br>simple device of mean-correction (subtract off the sample mean and add on the<br><br>known population mean), a modified estimator is constructed from the standard<br><br>quantile estimator. Asymptotic normality of the new estimator and its asymptotic<br><br>efficiency relative to the original estimator are derived. Simulation results for several<br><br>underlying distributions show that the proposed estimator is more efficient than the<br><br>traditional one.