The Kruskal–Wallis tests are Cochran–Mantel–Haenszel mean score tests

Abstract

The Kruskal–Wallis tests are appropriate tests for the completely randomised design, both for when the data are untied ranks, and, with adjustment, for when there are ties and mid-ranks are used. Both these tests are shown to be Cochran–Mantel–Haenszel mean score tests. The relationship between the Kruskal–Wallis test statistic and the ANOVA F test statistic when there are no ties generalises to the same relationship between the Cochran–Mantel–Haenszel mean score test statistic and the ANOVA F test statistic. It thus also relates both Kruskal–Wallis test statistics to the ANOVA F test statistic. A small simulation study finds that p-values may be more accurately found using the F test.

This is a preview of subscription content, access via your institution.

References

  1. [1]

    Agresti, A.: Categorical Data Analysis, 3rd edn. Wiley, New York (2013)

    Google Scholar 

  2. [2]

    Best, D.J., Rayner, J.C.W.: Conover’s F test as an Alternative to Durbin’s test. J. Modern Appl. Stat. Methods 13, 76–83 (2014)

    Article  Google Scholar 

  3. [3]

    Lachin, J.M.: Power and sample size evaluation for the Cochran–Mantel–Haenszel mean score (Wilcoxon rank sum) test and the Cochran-Armitage test for trend. Stat. Med. 30, 3057–3066 (2011)

    MathSciNet  Article  Google Scholar 

  4. [4]

    Rayner, J.C.W.: Introductory Nonparametrics. Bookboon, Copenhagen (2016)

    Google Scholar 

  5. [5]

    Rayner, J.C.W.: Cochran–Mantel–Haenszel Tests for the completely randomised design. J. Korean Stat. Soc. (2020). https://doi.org/10.1007/s42952-020-00068-3

    Article  Google Scholar 

  6. [6]

    Spurrier, J.D.: On the null distribution of the Kruskal-Wallis statistic. J. Nonparametr. Stat. 15(6), 685–691 (2003). https://doi.org/10.1080/10485250310001634719

    MathSciNet  Article  MATH  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to J. C. W. Rayner.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Rayner, J.C.W., Livingston, G. The Kruskal–Wallis tests are Cochran–Mantel–Haenszel mean score tests. METRON 78, 353–360 (2020). https://doi.org/10.1007/s40300-020-00192-4

Download citation

Keywords

  • Completely randomised design test
  • Homosexual marriage example
  • Mid-ranks
  • Scores
  • Whiskey example

Mathematics Subject Classification

  • 62G
  • 62F