TABLE OF CONTENTS As noted in the preface, we have two objectives for writing this book. First, we provide a protocol for testing the capability of a measurement system. Second, we provide an up-to-date summary of confidence interval estimation for variance components and means in normal-based random and mixed analysis of variance (ANOVA) models.
Although confidence intervals for ANOVA models are useful in many fields, we focus on the important application of gauge repeatability and reproducibility (R&R) studies. Confidence intervals are used in this application to determine if a measurement system is capable of monitoring a manufacturing process. This is an important application because reliable measurements are needed to monitor any manufacturing process.
Consistent with the second objective, this book extends the review of modified large-sample (MLS) confidence intervals for variance components presented in the book by Burdick and Graybill [10]. In addition, we report generalized confidence intervals (GCIs) for functions of variance components and means. GCIs are based on the process of generalized inference first proposed by Tsui and Weerahandi [65].
We hope the gauge R&R application considered in this book will be easily understood by readers in all fields of discovery. To help ensure this outcome, we provide background on this manufacturing application in this chapter. More detail on gauge R&R studies is provided in the article by Burdick, Borror, and Montgomery [8].
To properly monitor and improve a manufacturing process, it is necessary to measure attributes of the process output. For any group of...
