The multiple comparison of ranked parameters
- 1 January 1989
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Theory and Methods
- Vol. 18 (4) , 1217-1237
- https://doi.org/10.1080/03610928908829963
Abstract
Suppose we have k normal populations with means μ1, …, μk and common variance σ2 (known or unknown). The ordered values of μi are denoted by .Lam(1986) obtained confidence intervals for all pairwise comparison of the ranked parameters for i ≠ j. This paper deals with the multiple comparison of specific pairs of ranked parameters. In particular, upper as well as lower confidence bounds for are derived. The results are applied to obtain a confidence interval for the true achieved probability of correct selection under the indifference-zone approach of Bechhofer (1954) and the subset selection approach of Gupta (1965).Keywords
This publication has 8 references indexed in Scilit:
- A Lower Confidence Bound on the Probability of a Correct SelectionJournal of the American Statistical Association, 1986
- A new procedure for selecting good populationsBiometrika, 1986
- Multiple Comparisons with the Best TreatmentJournal of the American Statistical Association, 1983
- Simultaneous Confidence Intervals for all Distances from the "Best"The Annals of Statistics, 1981
- Indifference-zone ranking and selection: confidence intervals for true achieved P(CD)Communications in Statistics - Theory and Methods, 1977
- On Some Multiple Decision (Selection and Ranking) RulesTechnometrics, 1965
- New Tables for Multiple Comparisons with a ControlPublished by JSTOR ,1964
- A Single-Sample Multiple Decision Procedure for Ranking Means of Normal Populations with known VariancesThe Annals of Mathematical Statistics, 1954