Use of Predictive Distributions in Acceptance Sampling

Document Type : Original Article

Author

Abstract

Two possible acceptance criteria are considered. The first is based on the predictive distribution of the first failure time of the unsampled items in a batch, and the second based on the proportion of unused items having specified minimum lifetime. In both cases the underlying distribution is assumed to be the one-parameter exponential. The discussion is restricted to a classical approach which happens to coincide in each case with the noninformative Bayesian approach. Numerical examples are provided to illustrate our method.

Keywords