Permutation Test Approach for Ordered Alternatives in Randomized Complete Block Design: A Comparative Study

Randomized complete block design is one of the most used experimental designs in statistical analysis. For testing ordered alternatives in randomized complete block design, parametric tests are used if random sample are drawn from Normal distribution. If normality assumption is not provide, nonparam...

Full description

Saved in:
Bibliographic Details
Published inGazi University Journal of Science Vol. 27; no. 1
Main Authors GOKPINAR, Esra; Gazi University, Faculty of Science, Department of Statistics, Ankara, GUL, Hasan Hüseyin; Gazi University, Faculty of Science, Department of Statistics, Ankara, GOKPINAR, Fikri; Gazi University, Faculty of Science, Department of Statistics, Ankara, BAYRAK, Hülya; Gazi University, Faculty of Science, Department of Statistics, Ankara, OZONUR, Deniz; Gazi University, Faculty of Science, Department of Statistics, Ankara
Format Journal Article
LanguageEnglish
Published GAZI UNIVERSITY JOURNAL OF SCIENCE 01.01.2014
Gazi Üniversitesi Yayınları
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:Randomized complete block design is one of the most used experimental designs in statistical analysis. For testing ordered alternatives in randomized complete block design, parametric tests are used if random sample are drawn from Normal distribution. If normality assumption is not provide, nonparametric methods are used. In this study, we are interested nonparametric tests and we introduce briefly the nonparametric tests, such as Page, Modified Page and Hollander tests. We also give Permutation Version of Page test. We compare the performance of these tests in terms of the empirical type I error rates and powers of tests.Key words: Permutation test, Page test, Hollander test, Modified Page test, Simulation study
Bibliography:http://gujs.gazi.edu.tr/article/view/1060001347
ISSN:2147-1762