Article ID Journal Published Year Pages File Type
1149079 Journal of Statistical Planning and Inference 2010 9 Pages PDF
Abstract

This paper considers the one-sample sign test for population quantiles in general ranked set sampling, and proposes a weighted sign test because observations with different ranks are not identically distributed. It is shown analytically that optimal weight always improves the Pitman efficiency for all distributions. For each quantile, the sampling allocation that maximizes the sign test efficacy is identified and shown to not depend on the population distribution. Moreover, distribution-free confidence intervals for quantiles based on ordered values of optimal ranked set samples are discussed.

Related Topics
Physical Sciences and Engineering Mathematics Applied Mathematics
Authors
, ,