Measuring Agreement

Preface

This book presents statistical models and methods for analyzing common types of data collected in method comparison experiments and illustrates their application through detailed case studies. The main aim of these trials is to evaluate agreement between two or more methods of measurement. Although such studies are particularly abundant in health-related fields, they are also conducted in other disciplines, including metrology, ecology, and social and behavioral sciences.

Currently, at least six books cover the topic of agreement evaluation, including von Eye and Mun (2004), Carstensen (2010), Dunn (2004), Shoukri (2010), Broemeling (2009), and Lin et al. (2011). Of these, the first focuses exclusively on categorical data, and the second on continuous data. Others consider both types of data with varying levels of depth and choice of topics. Our book also considers both but with a primary focus on continuous data and one chapter devoted to categorical data. By providing chapter-length treatments of the common types of continuous data, it offers a comprehensive coverage of the topic, and its scope is broader than any other book currently available. It, however, by no means offers a complete survey of the literature. For example, measurement error models, Bayesian methods, and approaches based on generalized estimating equations are not included.

Essentially two principles guided us while writing this book. The first was to view the analysis of method comparison data as ...