|
Activity Number:
|
55
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Sunday, August 2, 2009 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #303514 |
|
Title:
|
The Generalized ANOVA: A Classic Song Sung with Modern Lyrics
|
|
Author(s):
|
Hui Zhang*+ and Hua He and Qin Yu and Rui Chen and Naiji Lu and Xin M. Tu
|
|
Companies:
|
University of Rochester and University of Rochester and University of Rochester and University of Rochester Medical Center and University of Rochester and University of Rochester
|
|
Address:
|
601 Elmwood Avenue Box 630, Rochester, NY, 14642,
|
|
Keywords:
|
Count response ; Inverse probability weighted estimate ; Missing data ; Overdispersion ; U-statistics
|
|
Abstract:
|
The analysis of variance (ANOVA) model suffers from a series of flaws that effect the conclusions and undercut its many potential applications. In this paper, we propose new generalized ANOVA models to concurrently address all these fundamental flaws so that it can be applied to many immediate and potential applications ranging from addressing an age-old technical issue to cutting-edge methodological challenges arising from the emerging effectiveness research paradigm. By integrating the classic theory of U-statistics with state-of-the-art concepts such as the inverse probability weighted estimates, we develop distribution-free inference to address missing data for longitudinal clinical trials and cohort studies. We illustrate the proposed models with both real and simulated study data, with the latter investigating behaviors of model estimates under small and moderate sample sizes.
|
