Nonparametric multiple comparisons with more than one control

Distribution-free methods are considered for comparing the medians of k1 treatments to the medians of k2 controls in the one-way layout under the assumption of equal sample sizes within the treatments and equal sample sizes within the controls. The problem is motivated by a biomedical example. For M...

Full description

Saved in:
Bibliographic Details
Published inJournal of nonparametric statistics Vol. 13; no. 3; pp. 397 - 412
Main Authors Solorzano, Eleanne, Spurrier, John D.
Format Journal Article
LanguageEnglish
Published Gordon and Breach Science Publishers 01.01.2001
Subjects
A-optimal
Distribution-free
Linear-rank statistic
Many-one procedure
Optimal design
Treatment versus control
Online AccessGet full text

Cover

More Information
Summary:Distribution-free methods are considered for comparing the medians of k1 treatments to the medians of k2 controls in the one-way layout under the assumption of equal sample sizes within the treatments and equal sample sizes within the controls. The problem is motivated by a biomedical example. For Mann-Whitney statistics, exact one-sided tests and simultaneous confidence intervals are developed for the case of k1 = k2 = 2. Upper bounds for the Type 1 error probability are established for general k\ and k2. Asymptotic probability points and asymptotically optimal designs are developed for use with Chernoff- Savage statistics.
ISSN:1048-5252
1029-0311
DOI:10.1080/10485250108832858