For nearly 20 years, Michael McDermott has been involved in the design and analysis of clinical trials and other clinical and basic studies in a variety of neurological diseases including Parkinson’s disease, Huntington’s disease, Tourette’s syndrome, epilepsy, periodic paralysis, various muscular dystrophies and other muscle diseases, HIV-associated dementia, multiple sclerosis, and attention deficit-hyperactivity disorder. He was Associate Editor of the journal Neurology from 1997-2003 and has served on the Medical Advisory Committee of the Muscular Dystrophy Association since 1998.

Much of Dr. McDermott’s statistical research has been in the area of order-restricted inference, specifically with regard to developing novel approaches to hypothesis testing problems involving order-constrained parameters (e.g., tests for trend in dose-response studies). More recently he has been working on problems of inference concerning the accuracy of diagnostic tests in two different settings: (1) when all subjects are administered the diagnostic test but the true disease status is verified in only a subset of these subjects (verification bias), and (2) when there are more than two possible diagnostic categories (receiver operating characteristic [ROC] surfaces). He has authored or co-authored more than 140 peer-reviewed articles.