Comparing three groups
arXiv:2005.04930 · doi:10.1080/00031305.2021.2002188
Abstract
We revisit simple and powerful methods for multiple pairwise comparisons that can be used in designs with three groups. We argue that the proper choice of method should be determined by the assessment which of the comparisons are considered primary and which are secondary, as determined by subject-matter considerations. We review four different methods that are simple to use with any standard software, but are substantially more powerful than frequently-used methods such as an ANOVA test followed by Tukey's method.
References in corpus (4)
- The sequential rejection principle of familywise error control
- Exact testing with random permutations
- Only Closed Testing Procedures are Admissible for Controlling False Discovery Proportions
- Comparisons of multiple treatment groups with a negative control or placebo group: Dunnett test vs. closed test procedur