Activity Number:
|
338
- SPEED: Biostatistical Methods, Application, and Education, Part 1
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, July 30, 2019 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistics in Epidemiology
|
Abstract #306842
|
Presentation
|
Title:
|
Joint Valid Moments Bayesian Marginal Logistic Regression Model with Time Dependent Covariates
|
Author(s):
|
Maria Vazquez* and Jeffrey Wilson
|
Companies:
|
and W. P. Carey School of Business, ASU
|
Keywords:
|
Binary outcomes;
longitudinal data;
MCMC;
applications in medicine
|
Abstract:
|
Longitudinal studies on binary outcomes in biomedical and health related applications though also found elsewhere usually consist of covariates that changes over time. These time-dependent covariates affect outcomes differently from one-time point to another. Thus, researchers are often interested in a multiple of responses and as such prefer joint modeling of the outcomes to accommodate the interdependence between them in assessing the risk factors rather than separate modeling. Recent studies in addressing time-dependent covariates assume that the effect of the covariates on these outcomes is the same across time. We propose a method based on valid moments (MVM) Bayesian marginal model that relies on the partitioned data matrix which addresses the time-dependency. Examples from the National Alzheimer’s Coordinating Center and Ad Health Study are presented.
|
Authors who are presenting talks have a * after their name.