A Robust Class of Multiple Testing Procedures Under Dependence

• Activity Number: 78
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Nasrine Bendjilali*+ and Boualem Bendjilali and Wei-Min Huang
• Companies: Rowan University and RVCC and Lehigh University
• Keywords: multiple testing procedures ; p-value ; type I error rates

Abstract

The majority of multiple testing procedures developed to date rely on the assumption of independence between test statistics. In practice, the assumption of independence is violated in most cases. By not taking the underlying dependence structure into account, this can adversely affect the power of any multiple testing procedure. In this paper, we examine the performance of a robust class of multiple testing procedures under various dependence structures. In addition, the study examines the impact of dependence on the stability of the proposed multiple testing procedure and the control of the various type I error rates. Theoretical and simulation results are presented.