What is it about?

There has been a longstanding debate about what method of Type I error control should be used when there are multiple dependent variables: Should MANOVA-protection, the Bonferroni procedure, or a modified Bonferroni-type procedure (e.g., Holm, Hochberg, or Hommel) be used? This study shows that the answer is complex and depends on several factors, which are herein described in detail and illustrated in a series of graphs.

Featured Image

Why is it important?

Previous studies have not thoroughly compared multiple-testing procedures in multivariate contexts, even in the simple two-group case. This paper compares the power, familywise error rates, and per-family error rates of several procedures, using a comprehensive range of effect sizes and correlations among the dependent variables. This paper also corrects several common myths about MANOVA-protected testing.

Read the Original

This page is a summary of: Power and Type I Error Control for Univariate Comparisons in Multivariate Two-Group Designs, Multivariate Behavioral Research, March 2015, Taylor & Francis,
DOI: 10.1080/00273171.2014.968836.
You can read the full text:

Read

Contributors

The following have contributed to this page