|
SPEAKER: |
|
|
TITLE: |
"Stepwise Multiple Hypotheses Testing with Applications" |
|
DAY: |
Wednesday, January 21, 1998 |
|
TIME: |
3:30 p.m. [Coffee in BSB-202 at 3:00 p.m.] |
|
PLACE: |
BSB-108 |
We consider the problem of testing a family of hypotheses, subject to the requirement that the family-wise error rate (i.e., the probability of any Type I error) is no greater than a specified level. Stepwise testing procedures of both the step-down and step-up type will be described. Practical applications to comparing treatments in medical trials will be given. My talk is based to a large extent on recent joint work with Dr. Ajit Tamhane of Northwestern University.
Dr. Charles Dunnett is Professor Emeritus in the Departments of Mathematics & Statistics and Clinical Epidemiology & Biostatistics at McMaster University. He has a BA degree from McMaster in mathematics and physics, an MA from University of Toronto in mathematics and a DSc from University of Aberdeen in statistics. He also did graduate study at Columbia University in mathematical statistics. During the Second World War, he served as a radar officer in the Fleet Air Arm and was awarded the MBE. He was employed as a statistician at Health and Welfare Canada in Ottawa and at Lederle Laboratories in Pearl River, New York, where he was head of the Statistical Design and Analysis Department. He joined McMaster as Professor in 1974 and was Chair of the Department of Applied Mathematics in 1977-9. In 1987, he was appointed Professor Emeritus and since then has continued his research interests in medical statistics and multiple testing methods. He has spent research leaves at Cornell University, University of Aberdeen and University of Swansea and has been a Fellow of the American Statistical Association since 1965. He was awarded the Gold Medal of the Statistical Society of Canada in 1986.
The following articles have been provided by Dr. Dunnett to be used as background for his talk. They are on reserve at Thode Library (STATS 770: Statistics Seminar).
[1] Dunnett, C.W. & Tamhane, A.C. (1991). Step-Down Multiple Tests for Comparing Treatments With a Control in Unbalanced One-Way Layouts, STATISTICS IN MEDICINE 10, pp. 939-947.
[2] Dunnett, C.W. & Tamhane, A.C. (1992). A Step-Up Multiple Test Procedure, JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION 87, pp. 162-170.
[3] Dunnett, C.W. & Tamhane, A.C. (1992). Comparisons Between a New Drug and Active and Placebo Controls in an Efficacy Clinical Trial, STATISTICS IN MEDICINE 11, pp. 1057-1063.
[4] Dunnett, C.W. & Tamhane, A.C. (1995). Step-Up Multiple Testing of Parameters With Unequally Correlated Estimates, BIOMETRICS 51, pp. 217-227.
We have also provided a copy of the following article, in which Dunnett talks of his career. The article includes two fine photographs of him taken in 1944.
[5] Anonymous (1988). A Conversation with Charles W. Dunnett, SSC 1986 Gold Medallist, STATISTICAL SOCIETY OF CANADA BULLETIN LIAISON 3(1), pp. 25-33.