ABSTRACT

We investigate the problem of deriving the optimal sample size in a medical setup when misclassification is present. The methodology adopts the average coverage criterion for the binomial parameter utilizing efficient numerical techniques. The approach follows the Bayesian point of view.