Identifying dominant causes using leveraged study designs

Abstract

Dorian Shainin’s group comparison procedure is a way to find the dominant cause(s) of variation in a process, based on an investigation plan that exploits the idea of leveraging. In this paper, we study whether Shainin’s procedure is sound. Our main finding is that Shainin’s procedure is unreliable in identifying the dominant cause, and in addition, the procedure is inefficient. We propose a new efficient and sound analysis procedure based on the method of maximum likelihood. A critical evaluation shows our proposed method is superior. We also provide a tangible example and compare both methods’ outcomes.