Includes bibliographical references and index.

This book considers the problem of estimating the percent of nonconforming items in a process from the ground up. It examines methods based on both attribute data and variable data. For variable data, it looks at the basic techniques that were initially developed for data from normal distributions and considers how they must be modified to deal with non-normal data. The importance of capability indices and their relationship to the percent of non-conforming items is discussed, as is the use of statistical tolerance limits. Finally, the book covers an extension of univariate capability analysis to the multivariate situation, which is often ignored.