The effects of variables samples on component reliability

The effect of statistical quality control (SQC) by variables sampling on the probability distributions of controlled variables is considered and related quantitatively to measures of post-SQC reliability. The techniques permit the assessment of a particular SQC procedure with regard to the degree of protection obtained against the occurrence of specific defined events (e.g. failures), which constitutes a more direct assessment than those generally employed (e.g. producer and consumer risks). Assumptions regarding the statistics of controlled variables are not critical, provided that they are not clearly inappropriate, since the above assessment is on the basis of comparison of risks rather than absolute values. If certain comparative costs are known, or can be estimated, it is possible to determine optimum variables sampling plans, corresponding to minimum cost, on the basis of a cost balance between additional SQC and resulting reduction in risk of failure, or other malfunction.<
>