Explanatory Item Response Models: A Generalized Linear and Nonlinear Approach <i>edited by P. de Boeck and M. Wilson, reviewed by Cees A. W. Glas</i>
Besides classical test theory models, item response theory (IRT) models have become the most prominent measurement models in the behavioral and social sciences.Most applications are found in the field of educational measurement, but in other fields, such as psychology and epidemiology, IRT models receive more and more attention.The main feature of IRT is the separation of the influence of the items and the persons on the item responses.This has two consequences.Firstly, IRT supports the evaluation of local test reliability, that is, the reliability of inferences about the relations between specific test scores.One could think of the reliability with which a test score can be distinguished from a minimum passing score.In contrast, in classical test theory, test reliability is an index for the extent to which randomly sampled individuals from some well-defined population can be distinguish...
Cees A. W. Glas & Mark Wilson (2005). Explanatory Item Response Models: A Generalized Linear and Nonlinear Approach <i>edited by P. de Boeck and M. Wilson, reviewed by Cees A. W. Glas</i>. Journal of Educational Measurement, 42(3), 303-307. https://doi.org/10.1111/j.1745-3984.2005.00016.x