Application of Bayes Factor for One Way Analysis of Variance Using Random Effect Models
Bayesian analysis of variance (ANOVA) is gaining acceptance as an alternative to the hypothesis testing based on p-values in many areas of scientific research. Its Bayes factors are constantly developed and modified to suit prevailing situations in hypothesis testing. We transformed the Bayes factor...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | Bulgarian |
| Published: |
University of Sofia
2018-12-01
|
| Series: | Bulgarian Journal of Science and Education Policy |
| Subjects: | |
| Online Access: | http://bjsep.org/getfile.php?id=268 |
| Summary: | Bayesian analysis of variance (ANOVA) is gaining acceptance as an alternative to the hypothesis testing based on p-values in many areas of scientific research. Its Bayes factors are constantly developed and modified to suit prevailing situations in hypothesis testing. We transformed the Bayes factor proposed by Wang & Sun (2013) for one random effects model using Natural Log of Gamma Function approximation. The result shows that our modified Bayes factor proved more efficient than the Wang & Sun (2013) Bayes factor in testing the hypothesis of zero between factor variability when the sample size becomes large. |
|---|---|
| ISSN: | 1313-1958 1313-9118 |