Abstract
This paper develops a new class of joint modeling of mixed correlated ordinal and continuous responses with elliptically contoured errors. This joint model includes the latent variable approach of using an elliptically contoured distribution for mixed ordinal and continuous responses. A Markov Chain Monte Carlo sampling algorithm is described for estimating the posterior distribution of the parameters. For sensitivity analysis to investigate the perturbation from associate responses, it is demonstrated how one can use some elements of covariance structure. Influence of small perturbation of these elements on the posterior normal curvature is also studied. To illustrate the application of such modeling the data (medical) is analyzed.
Recommended Citation
Samani, Ehsan B.
(2013).
Local Influence in Bayesian Elliptically Contoured-Ordinal Model for Mixed Data,
Applications and Applied Mathematics: An International Journal (AAM), Vol. 8,
Iss.
2, Article 4.
Available at:
https://digitalcommons.pvamu.edu/aam/vol8/iss2/4