No Detail Found!

Stratified extreme-cum-median ranked set sampling

Author(s)
Asad Ali Department of Economics, School of Business and Economics, University of Management and Technology, Lahore, Pakistan
Muhammad Moeen Butt Department of Economics, School of Business and Economics, University of Management and Technology, Lahore, Pakistan
Mueen ud Din Azad Department of Economics, School of Business and Economics, University of Management and Technology, Lahore, Pakistan
Zahoor Ahmed University of Lahore, Sargodha Campus, Sargodha, Pakistan
Muhammad Hanif National College of Business Administration and Economics, Lahore, Pakistan
Abstract
Ranked set sampling is one of the traditional sampling procedures used for cost reduction and obtaining more representative sample selection in surveys. When the problem of heterogeneity occurs in population, use of stratification is recommended in literature for increasing efficiency of estimators. In the presence of heterogeneity and outliers (extreme values), Stratified Median Ranked Set Sampling (StMRSS) and Stratified Extreme Ranked Set Sampling (StERSS) have been suggested by researchers. In this article, we propose a new ranked set sampling procedure to estimate the population mean by combining the aforementioned techniques. The new procedure is named as Stratified Extreme-cum-Median Ranked Set Sampling (StEMRSS). Efficiency comparison of this procedure with its competitor procedures has been made theoretically and conditions of efficiency have been derived. Simulation study comparisons for symmetric as well as asymmetric probability distributions have also been conducted. The results for small and medium sample size show that StEMRSS outperforms its competitor procedures for both perfect as well as imperfect rankings. For large sample size the newly proposed sampling procedure and StMRSS are found to be almost equally efficient. A real-life data set (Andersen, 1993) is also considered to study the empirical significance of the newly proposed sampling procedure. It is observed that the results of real-life application are compatible and conformable to aforementioned Monte Carlo simulation outcomes. The newly proposed technique is highly recommended for heterogeneous and Non-normal populations providing option of flexible rankings either on study or auxiliary variable.
Publication Details
Page(s) 215-235
DOI DOI not available
Published Journal: Pakistan Journal of Statistics, Volume: 37, Issue: 3, Year: 2021

Request the full text from PASTIC

Keywords
estimation, Relative efficiency, Ranked set sampling, Stratification, Asymmetric distribution, Unbiased, Monte Carlo Simulation., Symmetric distribution
0

Citations

0

Downloads

42

Views