ScholarLens Back to search
2016•Communication in Statistics- Theory and MethodsRequires access

The two-step approach—a significant ANOVA F-test before Dunnett's comparisons against a control—is not recommended

Ludwig A. Hothorn

Open publisher page 20 citations

Abstract

Decision-tree approaches are commonly used in the analysis of toxicological assays. This paper considers the widely-used and recommended approach of first applying the ANOVA F-test and, if it is significant, using the many-to-one comparison procedure of Dunnett (1955 Dunnett, C.W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Amer. Statist. Assoc. 50(272):1096–1121.[Taylor & Francis Online], [Web of Science ®] , [Google Scholar]). Data configurations are provided for which the F-test is not significant, but the Dunnett-test is significant. Because a conditional F-test before the Dunnett-test may increase the false negative rate, it is recommended that the ANOVA F-test not be used at all as a global pre-test. In addition, related non-parametric tests are investigated.

About this research paper

What this paper is about

Decision-tree approaches are commonly used in the analysis of toxicological assays. This paper considers the widely-used and recommended approach of first applying the ANOVA F-test and, if it is significant, using the many-to-one comparison procedure of Dunnett (1955 Dunnett, C.W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Amer. Statist. Assoc. 50(272):1096–1121.[Taylor & Francis Online], [Web of Science ®] , [Google Scholar]). Data configurations are provided for which the F-test is not significant, but the Dunnett-test is significant. Because a conditional F-test before the Dunnett-test may increase the false negative rate, it is recommended that the ANOVA F-test not be used at all as a global pre-test. In addition, related non-parametric tests are investigated.

Why it matters

OpenAlex reports 20 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Decision-tree approaches are commonly used in the analysis of toxicological assays. This paper considers the widely-used and recommended approach of first applying the ANOVA F-test and, if it is significant, using the many-to-one comparison procedure of Dunnett (1955 Dunnett, C.W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Amer. Statist. Assoc. 50(272):1096–1121.[Taylor & Francis Online], [Web of Science ®] , [Google Scholar]). Data configurations are provided for which the F-test is not significant, but the Dunnett-test is significant. Because a conditional F-test before the Dunnett-test may increase the false negative rate, it is recommended that the ANOVA F-test not be used at all as a global pre-test. In addition, related non-parametric tests are investigated.

Key concepts: Analysis of variance, Statistics, Test (biology), Mathematics, Computer science, Biology, Paleontology

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
The two-step approach—a significant ANOVA F-test before Dunnett's comparisons against a control—is not recommended — Research Paper | ScholarLens