|
|||||
|
|
Intrinsic Priors for Testing Two Normal Means with Intrinsic Bayes Factors |
|
| Authors: | Dal HO Kim Sang Gil Kang Woo Dong Lee |
| Affiliation: | 1. Department of Statistics , Kyungpook National University , Taegu, Korea dalkim@knu.ac.kr;3. Department of Applied Statistics , Sangji University , Wonju, Korea;4. Department of Asset Management , Daegu Haany University , Kyungsan, Korea |
| Abstract: | ABSTRACT In this article we consider the problem of comparing two normal means with unknown common variance using a Bayesian approach. Conventional Bayes factors with improper non informative priors are not well defined. The intrinsic Bayes factors are used to overcome such a difficulty. We derive intrinsic priors whose Bayes factors are asymptotically equivalent to the corresponding intrinsic Bayes factors. We illustrate our results with numerical examples. |
| Keywords: | Bayes factor Intrinsic priors Non informative prior Two-sample normal problem |