Washington Statistical Society on Meetup

U.S. Census Bureau
DSMD Distinguished Seminar Series

Title: Regression Toward the Mean and Multi-level Modeling

  • Presenter: Dr. Carl Morris, Department of Statistics, Harvard University
  • Discussant: Dr. Thomas Louis, US Census Bureau
  • Chair: Ruth Ann Killion, Chief, Demographic Statistical Methods Division, US Census Bureau
  • Date: Thursday, March 6, 2014 (rescheduled from Tuesday, March 4, 2014)
  • Time: 3:30 pm
  • Where: Conference Rooms 3&4, US Census Bureau, 4600 Silver Hill Road, Suitland, Maryland
  • Contact: Cynthia Wellons-Hazer, 301-763-4277, Cynthia.L.Wellons.Hazer@census.gov

Abstract:

When one expects forecasts and estimates to be affected by regression toward the mean (RTTM), a two-level hierarchical description of the data offers an excellent model. We start by reviewing how RTTM arises, and misunderstandings about it. As time allows, there will be examples of how RTTM affects rating sports players and teams, ranking of hospitals, determining adequate sample sizes, and accounting for RTTM selection bias. We finish with implications for statistical theory, including hierarchical modeling and the Bayes/frequency interface.

About Speakers:

Dr. Carl N. Morris (Ph.D. Statistics, Stanford, 1966; B.S. Statistics, California Institute of Technology, 1960) is Professor at the Harvard Faculty of Arts and Sciences, Department of Statistics. Dr. Morris joined Harvard in 1990 with Professorships evenly split between the Statistics Department (Arts and Sciences), and the Department of Health Care Policy (Harvard Medical School), but that changed in 1995, when Chairing the Department of Statistics required his full-time attention. Dr. Morris' career includes Editorships of two leading statistics journals, Editor of Journal of the American Statistical Association (1983-1985), and Executive Editor of Statistical Science (1989-1991). He is a Fellow of the ASA, IMS, and Royal Statistical Society; an elected member of ISI; and a member of the Biometric Society.

Dr. Thomas Louis (PhD in Mathematical Statistics, Columbia University) is Professor of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Associate Director for Research and Methodology and Chief Scientist of US Census Bureau. His research includes Bayesian methods; clinical and field trials; health services research, environmental risk assessment, and genomics. Current applications include health services studies, clinical trials on the treatment of Uveitis and behavioral interventions to reduce obesity, and studies of malaria epidemiology, vector biology and parasite genomics in Southern Africa. He has published over 290 articles, books/chapters, monographs and discussions. Professor Louis is an elected member of the International Statistical Institute, a Fellow of the American Statistical Association and of the American Association for the Advancement of Science.