ASQ Austin - July 2025 Lunch and Learn

Statistical Process Control (SPC) is the most widely used statistical application in the known universe. Many manufacturing sites operate thousands of SPC charts, often obtaining multiple observations per day on each chart. To operate SPC systems on such a large scale requires software, and there are many such commercially available systems. Many of these systems include options to apply an array of potential process control rules beyond the original rule proposed by Walter Shewhart in his 1931 book Economic Control of Quality of Manufactured Product. With the ready availability to apply these additional rules, often referred to as Western Electric (WECO) or Nelson rules, it is often tempting for practitioners to add some number of these rules to improve the sensitivity of the associated process control plan. While such an approach will indeed improve the capability of the plan to detect out-of-control operating conditions, the price that is paid for this is a higher false out-of-control signal rate. This presentation includes a review of the WECO rules and evaluates the sensitivity and associated false signal rates for a variety of combinations of these run-based rules, ultimately suggesting which of these additional rules might be of some value, as well as those which are of limited effectiveness.