Multiple Comparisons for a Bivariate Exponential Populations under Random Censorship

We consider a two-component system which their lifetimes have Freund’s bivariate exponential model. In this case, we suggest a multiple comparisons procedure using fractional Bayes factor. Here we assume that the components enter the study at random over time and the analysis is carried out at random censored time. We derive fractional Bayes factor for all possible hypotheses of multiple comparisons under noninformative priors for the parameters, and calculate the posterior probabilities for all hypotheses and select a hypothesis which has the highest posterior probability as the best model. Finally, we give a numerical example to illustrate our procedure. From the results of the numerical example, we know that the data lends greatest support to the best hypothesis.