Journal of the Korean Statistical Society
- Volume 16 Issue 1
- /
- Pages.26-36
- /
- 1987
- /
- 1226-3192(pISSN)
- /
- 2005-2863(eISSN)
On a Subset Selection Procedure Based on Hodges-Lehmann Estimators
- Song, Moon-Sup (Department of Computer Science and Statistics, Seoul National University, Seoul 151) ;
- Kim, Soon-Ock (Department of Computer Science and Statistics, Seoul National University, Seoul 151)
- Published : 1987.06.01
Abstract
In this paper, we study on a subset selection procedure based on Hodges-Lehmann estimators derived from the Wilcoxon test. To estimate the standard error of the Hodges-Lehmann estimators, the biweight A-estimator of scale is used. The Pitman efficiency of the proposed rule is compared with the Gupta's rule and the trimmed-means rule through a small-sample Monte Carlo study. The results show that the proposed rule satisfies the $P^*$-condition and is very efficient in various heavy-tailed distributions.
Keywords