ABSTRACT
Chapter 4 deals with measurement error and reliability. It starts with a formal definition and basic properties of different estimates of reliability in the context of the Classical Test Theory (CTT). We cover the classical reliability estimates based on correlations and we present connections to Fisherian analysis of variance. Estimation methods of variance components are presented, including maximum likelihood methods within the random effect and mixed-effect models, Bayesian estimation, and bootstrap techniques. Extensions to more sources of error are provided within the framework of the generalizability theory and further extensions are discussed.
