Product-specific adaption of conformity assessment criteria and their financial consequences

Abstract

The conformity of products is usually assessed on the basis of Gaussian distributions of the test results. For wire screen products, as an example, this assumption is only valid if the apertures are quite close to the nominal value and if the dispersion of the apertures is small compared to the tolerance band. In cases where these provisions are not given beta distributions are better distribution approaches. The application of a modified beta distribution leads to significantly lower producer’s and consumer’s risks allowing to expand safeguard limits without changing the levels of confidence. A financial model enables the user to create scenarios for such different input data as the effort for testing, the number of products in line with the production target, the prices of products and scrap prices as well as quality classifications of products. The pivots chosen for the model are the producer’s and consumer’s risks. For reasons of comparison the risks were determined both on the basis of normal and modified beta distributions. Product-specific conformity assessment procedures may have apparent financial benefits if the parameters of the model can be well defined.