Curtailment of artificial attribute sampling plans

Abstract

Curtailed sampling with one or two artificial ‘compressed limits’ (which are more severe than specification limits) are analysed for a single specification limit of a normally distributed variable, and the magnitude of the reduction in the sample size due to curtailment is evaluated. The computational methods for derivation of an efficient non-curtailed sampling plan are reviewed. A few curtailed sampling procedures are presented and numerical results are given.