|
Activity Number:
|
92
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 4, 2003 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #301321 |
|
Title:
|
Multiple Outputation: Inference for Complex Clustered Data by Averaging Analyses from Independent Data
|
|
Author(s):
|
Dean A. Follmann*+ and Michael A. Proschan
|
|
Companies:
|
National Institutes of Allergy and Infectious Diseases and National Heart, Lung & Blood Institute
|
|
Address:
|
6700 A Rockledge Drive, Bethesda, MD, 20892-0001,
|
|
Keywords:
|
bootstrap ; GEE ; generalized linear mixed models ; multiple imputation ; resampling ; within-cluster resampling
|
|
Abstract:
|
This talk applies a simple method for settings where one has clustered data, but statistical methods are only available for independent data. We assume the statistical method provides us with a normally distributed parameter estimate E and an estimate of its variance V. We randomly select a data point from each cluster and apply our statistical method to this independent data. We repeat this multiple times, and use the average of the associated E as our estimate. An estimate of the variance is given by the average of the Vsn minus the sample variance of the E. We call this procedure multiple outputation, as all "excess'' data within each cluster is thrown out multiple times. Hoffman, Sen, and Weinberg (2001) introduced this approach for generalized linear models when the cluster size is related to outcome. In this paper we demonstrate the broad applicability of the approach. In addition, asymptotic normality of estimates based on all possible outputations as well as a finite number of outputations is proven given weak conditions. Multiple outputation provides a simple and broadly applicable method for analyzing clustered data.
|
