|
Activity Number:
|
340
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #305845 |
|
Title:
|
Effects of Misclassification on Exposure-Disease Association with Odds Ratio
|
|
Author(s):
|
Jun-mo Nam*+
|
|
Companies:
|
National Cancer Institute
|
|
Address:
|
EPS, Room 8028, Rockville, MD, 20892-7240,
|
|
Keywords:
|
bias ; efficiency ; interval estimation ; misclassification ; odds ratio
|
|
Abstract:
|
Effects of the exposure misclassification in inference on an exposure-disease association with the odds ratio are a major concern in case-control studies. When the misclassification is independent of case or control (non-differential misclassification), we expect the estimated odds ratio is biased towards to the null value. However, the estimate for a single study may overestimate the true odds ratio by chance. We investigate statistical methods for identifying non-differential misclassification. Regardless of differential or non-differential misclassification, it causes bias and a loss of efficiency in interval estimation of the odds ratio. If misclassification is present, the actual coverage probability of a confidence interval for the odds ratio is smaller than a nominal one and the expected width of the confidence interval is broader.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2006 program |
