BAYES PREDICTION INTERVALS FOR THE BURR TYPE XI1 DISTRIBUTION IN THE PRESENCE OF OUTLIERS

Abstract

Using a sample fiom Burr type XU distribution, Bayes prediction intervals are derived for the maximum and minimum of a future sample fromthe same distribution, but in the presence of a single outlier of the type 8,8. The prior of Q is assumed to be the gamma conjugate. A real example is given to illustrate the procedure. Also, the comparison between the values of the prediction bounds for different values of ? and different future sample sizes is given