|
Activity Number:
|
568
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #305679 |
|
Title:
|
Generalized Causal Mediation Analysis
|
|
Author(s):
|
Jeffrey M. Albert*+
|
|
Companies:
|
Case Western Reserve University
|
|
Address:
|
10900 Euclid Avenue, Cleveland, OH, 44106-4945,
|
|
Keywords:
|
causal inference ; generalized linear models ; direct and indirect effects ; dental data ; path analysis
|
|
Abstract:
|
The goal of mediation analysis is to assess direct and indirect effects of a treatment or exposure on an outcome. Methods for estimating mediated (pathway) effects are available for continuous variables related via linear models. In addition, methods have been developed for categorical variables, but usually limited to a single stage of mediation. We present a method applicable to multiple stages of mediation, and for mixed variable types related via generalized linear models. We define pathway effects via a causal (potential outcomes) approach and present a general formula that provides the effect of exposure through any specified pathway. Regression parameters are estimated in stages via generalized linear models, and variances of estimates are estimated via a bootstrap method. The method is applied to a longitudinal study of dental caries in very low birth weight adolescents.
|
- 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 2009 program |
