Adapting to One- and Two-Way Classified Structures of Hypotheses While Controlling False Discoveries

• Activity Number: 381 - Recent Advances in Multiple Testing and False Discovery Rate Analysis
• Type: Invited
• Date/Time: Tuesday, July 30, 2019 : 2:00 PM to 3:50 PM
• Sponsor: IMS
• Author(s): Sanat K Sarkar*
• Companies: Temple University
• Keywords: Multiple Testing ; One-Way Classifioed Hypotheses; Two-Way Classified Hypotheses; False Discovery Rate; Local False Discovery Rate

Abstract

Modern scientific investigations, notably in genome-wide association and neuroimaging studies, often give rise to multiple testing problems for hypotheses that are classified according to multiple criteria. This creates a situation where the significance of a hypothesis needs to be carefully assessed by taking into account the underlying classification structure. We refer to a set of hypotheses as one-way classified hypotheses if they are classified according to one criterion, and as two-way classified hypotheses if they are classified according to two different criteria. This talk presents some newer results on controlling false discoveries while testing multiple hypotheses that appear in one-way and two-way classified forms.