Prince Allotey
Department of Statistics
Western Michigan University
In this study, we develop random effects regression models for bivariate spatially correlated censored data. In our model, the large scale variations are characterized by a linear function of explanatory variables and small scale variations are characterized by bivariate spatial processes defined over the lattice of interest. Our model has the flexibility to incorporate information from various spatial scales.
Our analyses have been performed in Bayesian paradigm and inferences are drawn using Markov Chain Monte Carlo algorithms. Several competing models are fitted and compared in the process. Important properties of our models are studied through extensive simulation studies. We illustrate the usefulness of our proposed model using data from the Michigan Department of Health and Human Services.
---
All statistics students are expected to attend.