- Shopping Bag ( 0 items )
The author clearly explains how to reduce measurement error, presents numerous practical examples of the interobserver agreement approach, and emphasizes measures of agreement among raters for categorical assessments. The models and methods are considered in two different but closely related contexts: 1) assessing agreement among several raters where the response variable is continuous and 2) where there is a prior decision by the investigators to use categorical scales to judge the subjects enrolled in the study. While the author thoroughly discusses the practical and theoretical issues of case 1, a major portion of this book is devoted to case 2. He explores issues such as two raters randomly judging a group of subjects, interrater bias and its connection to marginal homogeneity, and statistical issues in determining sample size.
Statistical analysis of real and hypothetical datasets are presented to demonstrate the various applications of the models in repeatability and validation studies. To help with problem solving, the monograph includes SAS code, both within the book and on the CRC Web site. The author presents information with the right amount mathematical details, making this a cohesive book that reflects new research and the latest developments in the field.