Confidence Regions Corresponding to Multiple Comparisons Procedures

• Activity Number: 60
• Type: Topic Contributed
• Date/Time: Sunday, August 9, 2015 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Brian Wiens*
• Companies: Portola Pharmaceuticals
• Keywords: Fallback ; Bonferroni ; closed testing

Abstract

We consider confidence regions that correspond to multiple comparisons procedures including the Bonferroni, Holm, fallback and 4A procedures. Two-sided confidence regions are preferred to allow for a better understanding of the maximum as well as minimum benefit that an investigational treatment provides. A straightforward method of developing confidence intervals is to include all points such that the global null hypothesis that the point is the true value is not rejected in favor of a two-sided alternative. Test-based confidence intervals (or inverted hypothesis tests) are not a new idea. In spite of some appealing properties, they may be overshadowed by more recently proposed confidence regions. We discuss advantages and disadvantages, and present an example.