Lower bounds on bayes factors for a linear regression model |
| |
Authors: | Shigeru Iwata |
| |
Institution: | University of Kansas , Lawrence, KS, 660455-2113 |
| |
Abstract: | This paper develops a Bayesian analysis of testing linear parameter restictions in a normal regression model. Instead of searching for a suitable "objective" prior for such a test, several interesting classes of informative priors are postulated, and the lower bounds on the Bayes factors are derived. |
| |
Keywords: | Bayes factor hypothesis testing F-test posterior odds |
|
|