Effectiveness of quality inspections of attributive characteristics – A novel and practical method for estimating the “intrinsic” value of kappa based on alpha and beta statistics

Abstract

The main contribution is to provide practitioners with a complete and practical method to plan the studies to evaluate the effectiveness of attribute assessment in quality inspections. Kappa statistic is one of the most used methods to evaluate the effectiveness of inspections based on attribute assessments in the industry. However, its estimation by available methods does not provide its “real” or “intrinsic” value. This article develops a method to estimate kappa's “intrinsic” value. In addition, type-II error (beta) needs to be estimated separately due to its impact on customer satisfaction. Therefore, practitioners must evaluate alpha (type-I error) and beta apart from kappa to interpret the results better. As a novelty, this method helps practitioners plan the study to estimate beta, alpha and kappa at once. Based on the graphical representation of the mathematical relationship between three statistics (kappa, beta and alpha) derived here, this paper provides a precise estimate of kappa’s one-sided lower bound. Additionally, a real-life example illustrates the use of the method for practitioners.