Faculty of Science , Japan Women’s University , 2-8-1 Mejirodai Bunkyoku, Tokyo, 112-8681, Japan
Abstract:
Two methods for testing the equality of variances in straight lines regression with a change point are considered. One is likelihood ratio test and the other is Bayesian confidence interval, based on the highest posterior density for the ratio of variances, using non-informative priors. Methods are applied to the renal transplant data analyzed by Smith and Cook(1980) and Stephens(1994).