Consideration of Key Factors Affecting the Probability of False Acceptance of Testing and Calibration Laboratories

Laboratory False Acceptance Risk

Abstract

        The decision on laboratory test results is important because it affects consumers and manufacturers. It is well known that all measurements contain errors. In 2006, the American National Standards Institute (ANSI) issued ANSI/NCSL Z540.3 which specifies that the Probability of False Acceptance (PFA) should not exceed 2%. This study aims to elucidate the key factors affecting a false acceptance risk. The study was conducted by searching for academic information from widely published documents. The reviews found two important factors affecting the risk of false acceptance: 1) the capability of the laboratory measurements and 2) the capability of the product manufacturing process. If the laboratory has a Test Uncertainty Ratio (TUR) of 4.6:1 or higher, or the production process has an End-of-Period-Reliability (EOPR) of 89% or higher, the risk of false acceptance will not exceed 2%. Therefore, a testing and calibration laboratory accredited with ISO/IEC 17025:2017 can set the decision rule that considers only the laboratory measurement capability factors, without considering the production process capability factors, if the laboratory has sufficient measurement capability. The laboratory should ensure that the estimation of measurement uncertainty takes into account all important variables. In addition, practitioners should have knowledge and understanding of the nature of measurement and Probability Density Function (PDF) to determination of reliable and acceptable decision rules for both the laboratories and the customers.