Excessive variation in critical-to-quality process outputs is a common challenge in manufacturing industries. To mitigate it, determining the dominant cause(s) of variation is an effective and recommended first step. Shainin’s group comparison procedure is a way to identify these dominant cause(s) through a structured investigation plan that exploits the idea of leveraging. Our study evaluates the soundness of this procedure, and we find Shainin’s analysis procedure inefficient and, more importantly, unreliable in identifying the dominant cause. As an alternative, we propose a new, efficient, and reliable analysis procedure based on maximum likelihood. A critical evaluation demonstrates the superiority of our proposed analysis procedure. We also provide a tangible example and compare the outcomes of both methods.