American Statistical Association
Certain interactions, referred to as removable interactions, arise in statistical models due to the scale of the outcome's measurement. All the removable interactions in a model can be written parsimoniously using a single scalar parameter - equivalently, when interactions are removable under some scale of the outcome, we can fit an additive model to the data under an invertible transformation of the outcome by estimating a single scalar parameter to identify the specific transformation. These concepts, their uses in epidemiology and clinical studies, and some of the related statistical methods are illustrated using data from three endometrial cancer case-control studies.
|Date:||Wednesday, October 8, 2014|
|Time:||4:00 - 5:00 P.M.|
Memorial Sloan-Kettering Cancer Center
Department of Epidemiology and Biostatistics
307 East 63rd Street
(between First and Second Avenues)
3rd Floor Conference Room
New York, New York
Note: To gain access to the building, please follow the directions by the telephone in the foyer.