No Detail Found!

Sensitivity of ranked set sampling to the shape characteristics of the underlying population: A g and k distribution approach.

Author(s)
Syed S. Hossain Institute of Statistical Research and Training (ISRT), University of Dhaka, Dhaka, Bangladesh
A. Khan Institute of Statistical Research and Training (ISRT), University of Dhaka, Dhaka, Bangladesh
Abstract
Ranked set sampling (RSS) is seen to perform better than the usual simple random sampling (SRS) method in terms of efficiency. But it is also observed that the relative performance of the RSS method over the SRS method depends highly on the underlying population distribution. This paper attempts to investigate the nature and extent of the change in relative precision of the RSS estimator for different distributional pattern. This has been done by measuring the relative precision for distributions with varying degrees of skewness and kurtosis. A simulation is conducted assuming that the population follows the g and k quantile function distribution in which the shape parameters can be measured by the magnitude and direction of the departure from normality.
Publication Details
Page(s) 207-216
DOI DOI not available
Published Journal: Pakistan Journal of Statistics, Volume: 24, Issue: 3, Year: 2008

Request the full text from PASTIC

0

Citations

0

Downloads

20

Views