|M.Sc Student||Gur Hadas|
|Subject||FDR Controlling Procedure for Discrete Data|
|Department||Department of Industrial Engineering and Management||Supervisor||Ms. Ruth Heller|
|Full Thesis text|
Benjamini and Hochberg (1995) proposed the false discovery rate (FDR) as an alternative to the FWER in multiple testing problems. Since then, researchers have been increasingly interested in developing methodologies for controlling the FDR under different model assumptions. For discrete data commonly used FDR controlling procedures may be highly conservative. Incorporating the discreteness of the tests into the multiplicity adjustments may increase the power dramatically while maintaining the nominal FDR level. We will present alternative, more powerful, procedures that exploit the discreteness of the tests and have FDR levels closer in magnitude to the desired nominal level. Using simulations we observe large power gains of the discrete versions over the original versions. We apply these modified procedures to the pharmacovigilance spontaneous reporting systems in the United Kingdom, which are used for early detection of adverse reactions of marketed drugs.