Abstract
ABSTRACT Corporations of all sizes form teams in order to tackle quality problems. Often, these teams are given a quantifiable goal such as “reduce defects by 40% within the next two months.” In order to assess whether or not the goal has been met, data must be gathered. In this article, we develop methods for determining the sample sizes necessary for detecting relative quality improvements, with specified probability, in finite and infinite populations. We provide formulas for the calculations and use the Solver function in Excel to implement normal approximations to the solutions. In addition, we provide methods, based on the hypergeometric distribution, for finding the exact error rates (α and β) for given samples sizes. A real-life example is discussed and modified to illustrate cases with finite and infinite populations.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call