Economic Design of Lot-By-Lot Acceptance Sampling by Attributes

Risk-based sampling plans have long been the accepted standards in the practice of acceptance sampling for quality control. Usually the economic consequences of a given sampling plan are assessed by experience and intuition at best. With the availability of high speed computers and well-proven statistical methods, economically-optimal sampling pl a ns can now be developed so that much of the guesswork is removed. This study first surveys these methods for deriving economically-optimal sampling plans. Then, recognizing the reluctance in government and industry to abandoning ris k-based design of sampling plans, a method is explored which poses a compromise between the two. quality level is calculated at which the cost of inspection is equivalent to the cost of no inspection. A spr e adsheetbased model then selects an economically ne a r-optimal sampling p l a n from the MIL-STD-105D tables