Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


12810.pdf (3.82 MB)

ETD Abstract Container

Abstract Header

The Threshold Prior in Bayesian Hypothesis Testing

Glore, Mary Lee
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416570546

Abstract Details

2014, PhD, University of Cincinnati, Arts and Sciences: Mathematics (Statistics).
Bayesian hypothesis testing provides an attractive alternative that overcomes the difficulties with interpreting the p-value or calibrating it. Bayesian hypothesis testing also provides measures of evidence for not only the alternate but the null hypothesis as well. Bayesian hypothesis testing has its own challenges including choice of prior distributions for the model parameters. Researchers have suggested various choices of prior distributions. In Dynamic Factor Volatility Modeling: A Bayesian Latent Threshold Approach, Nakajima and West used a threshold based prior for model selection and prediction. [NW13b] In this dissertation, we investigate the properties of this “Threshold prior method” as applied to testing a point null hypothesis. We show how the answers using this prior compare to the traditional g-prior method. We also investigate the convergence rates of the Bayes factors in favor of the null and alternate hypotheses as sample size goes to infinity when the null and alternate hypotheses are true. In addition we compare the Threshold prior and g-prior methods for a multivariate regression model. We also investigate the alternative approach of using a loss function along with traditional priors, instead of the threshold approach.
Siva Sivaganesan, Ph.D. (Committee Chair)
Magda Peligrad, Ph.D. (Committee Member)
Xia Wang, Ph.D. (Committee Member)
104 p.
Statistics
Bayes; regression; Mathematica; nonlocal; g-prior; time series

Recommended Citations

Citations

  • Glore, M. L. (2014). The Threshold Prior in Bayesian Hypothesis Testing [Doctoral dissertation, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416570546

    APA Style (7th edition)

  • Glore, Mary Lee. The Threshold Prior in Bayesian Hypothesis Testing. 2014. University of Cincinnati, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416570546.

    MLA Style (8th edition)

  • Glore, Mary Lee. "The Threshold Prior in Bayesian Hypothesis Testing." Doctoral dissertation, University of Cincinnati, 2014. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416570546

    Chicago Manual of Style (17th edition)

ucin1416570546
569
© 2014, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.