Abstract
Banking upon on doubly Type-II censored samples, it is important to consider the role of lifetime distribution based on the Bayesian prediction intervals. A two-sample prediction technique was applied using the past and previous parameters for the single statistical significance problem. A combination of approximation approaches identified the major problem of the outlier presence, which affects the prior density estimations when unknown parameters are implemented. The gamma conjugate prior and joint prior cases are used to illustrate the statistical significance of the problem discussed, which are further supported by numerical estimations.