Estimating outgoing quality using the quality measurement plan

Consider a sequence of lots from a manufacturing process that is put through an acceptancesampling inspection program, the accepted lots being shipped to a customer. Taken together. the accepted lots form a population whose overall quality, called the outgoing quality, is of importance to the customer. This article describes a method for estimating this outgoing quality using data from both the accepted and rejected lots. The method is based on a very general definition of outgoing quality and on the quality measurement plan (QMP)—a hierarchical Bayes model of the underlying process. Our estimator is the only one known to us that works when sample sizes and acceptance numbers vary from lot to lot and the acceptance-sampling rules are sometimes broken—a practical reality. In the case of fixed sample size, the QMP estimator of outgoing quality has a very simple form. It is the weighted average of two naive estimators—the sample qualities observed in the incoming and outgoing lots, respectively. The weight o...