|
Activity Number:
|
134
- Bayesian Modeling
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 9, 2021 : 1:30 PM to 3:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
|
Abstract #318050
|
|
|
Title:
|
Numerical Approximation of the Marginal Likelihood of Random Effect Model for Clustered Bivariate Binary Outcomes
|
|
Author(s):
|
Edmund Essah Ameyaw* and John Essah Kwagyan
|
|
Companies:
|
Howard University and Howard University
|
|
Keywords:
|
Clustered bivariate binary outcome;
Simulaton;
random effect;
Gauss Hermite Quadrature;
Laplace approximation;
Marginal Likelihood
|
|
Abstract:
|
We developed a random effect (logistic -Gaussian) model for clustered bivariate binary outcomes. We consider Gauss Hermite Quadrature and the Laplacian approach to the approximation of the marginal likelihood function. We performed simulation and compare the computational and statistical properties. We showed that though the Gauss-Hermite Quadrature approximation has an increased computation time, due to the complex nature of the approach, it produces a more accurate results compared to the Laplacian approach.
|
Authors who are presenting talks have a * after their name.
