|
Activity Number:
|
257
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #304087 |
|
Title:
|
Weighting Methods for Population-Based Case-Control Study with Complex Sampling
|
|
Author(s):
|
Yan Li*+ and Barry I. Graubard and Ralph DiGaetano
|
|
Companies:
|
National Cancer Institute and National Cancer Institute and Westat, Inc.
|
|
Address:
|
6120 executive blvd , Rockville, MD, 20852,
|
|
Keywords:
|
case-control study ; design-based ; frequency matching ; population weight ; poststratification
|
|
Abstract:
|
The use of complex survey sampling designs in population-based case-control studies is becoming more common, particularly for sampling controls. Researchers have proposed various methods for analyzing data from case-control studies with stratified SRS designs. There are, however, limited methods for case-control studies under general complex sampling designs. The standard design-based approach to inference is not suitable for the analysis of population-based case-control studies because of the extreme variation of the sample weights and the mismatched-frequencies of the matching variables between cases and controls. In this paper, we proposed a weighting method to reduce the variation of the sample weights and also to recapture the efficiency due to frequency matching. The performance of the proposed method was studied using extensive simulation studies and a real data application.
|
