Design and analysis of reliability studies

The statistical methods used to evaluate and compare different methods of measurement are a vital common component of all methods of scientific research. This book provides a practically orientated guide to the statistical models used in the evaluation of measurement errors with a wide variety of illustrative examples taken from across the sciences. After introducing basic concepts, such as precision, reproducibility and reliability, a detailed discussion of the sources of variability of measurements and associated variance components models is provided. The central chapters deal with the design and analysis of method comparison studies (concentrating primarily on quantitative measurements) ranging from simple paired comparisons to more complex studies involving three or more methods. This leads on to a review of methods for categorical measures.

Chapter 1: Setting the scene.

Chapter 2: Sources of variation.

Chapter 3: Method comparison 1 (paired observation).

Capter 4: Method comparison 2 (more informative).

Chapter 5: Methods for categorical (binary) data.

References and select bibliography.