|
Activity Number:
|
448
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #302129 |
|
Title:
|
Bayesian Variable Selection for Logistic Regression with Misclassification
|
|
Author(s):
|
Stephanie Powers*+ and James Stamey
|
|
Companies:
|
Baylor University and Baylor University
|
|
Address:
|
9821 Chapel Rd. #621, Waco, TX, 76712,
|
|
Keywords:
|
Bayesian Variable Selection ; Non-Differential Misclassification ; Misclassification ; Logistic regression
|
|
Abstract:
|
Response misclassification is a problem in many studies with binary outcomes. In a logistic regression where the binary response is subject to misclassification, we investigate the impact ignoring misclassification has on determining a subset of significant covariates. We then develop a Bayesian approach to select significant covariates while accounting for the misclassification. The procedure is applied to a real data example of a smoking cessation study and extensions to potential pharmaceutical problems are discussed.
|
