Date of Award

2019

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy in Mathematical Sciences (PhD)

Administrative Home Department

Department of Mathematical Sciences

Advisor 1

Min Wang

Committee Member 1

Jianping Dong

Committee Member 2

Renfang Jiang

Committee Member 3

Yu Cai

Abstract

This dissertation consists of three distinct but related research projects. The first two projects focus on objective Bayesian hypothesis testing and estimation for the intraclass correlation coefficient in linear models. The third project deals with Bayesian quantile inference for the semiparametric mixed-effects double regression models. In the first project, we derive the Bayes factors based on the divergence-based priors for testing the intraclass correlation coefficient (ICC). The hypothesis testing of the ICC is used to test the uncorrelatedness in multilevel modeling, and it has not well been studied from an objective Bayesian perspective. Simulation results show that the two sorts of Bayes factors have good performance in the hypothesis testing. Moreover, the Bayes factors can be easily implemented due to their unidimensional integral expressions. In the second project, we consider objective Bayesian analysis for the ICC in the context of normal linear regression model. We first derive two objective priors for the unknown parameters and show that both result in proper posterior distributions. Within a Bayesian decision-theoretic framework, we then propose an objective Bayesian solution to the problems of hypothesis testing and point estimation of the ICC based on a combined use of the intrinsic discrepancy loss function and objective priors. The proposed solution has an appealing invariance property under one-to-one reparameterization of the quantity of interest. Simulation studies are conducted to investigate the performance the proposed solution. Finally, a real data application is provided for illustrative purposes. In the third project, we study Bayesian quantile regression for semiparametric mixed effects model, which includes both linear and nonlinear parts. We adopt the popular cubic spline functions for the nonlinear part and model the variance of the random effect as a function of the explanatory variables. An efficient Gibbs sampler with the Metropolis-Hastings algorithm is proposed to generate posterior samples of the unknown parameters from their posterior distributions. Simulation studies and a real data example are used to illustrate the performance of the proposed methodology.

Share

COinS