Multilevel Analysis of Educational Data

Multilevel Analysis of Educational Data focuses on the principles and procedures used in the evaluation of educational progress.

The selection first offers information on some applications of multilevel models to educational data, empirical Bayes methods, and a hierarchical item-response model for educational testing. Discussions focus on the interface between levels, group-level model for content elements, an application of empirical Bayes, validity generalization, improving law school validity studies, and summarizing evidence in randomized experiments on coaching. The text then takes a look at difficulties with Bayesian inference for random effects and multilevel aspects of varying parameters in structural models.

The book elaborates on models for multilevel response variables with an application to growth curves and the issues and problems emerging from the application of multilevel models in British studies of school effectiveness, including enduring questions, two-level models, estimation and prediction, and econometric random coefficient modeling. The selection is a dependable reference for educators and researchers interested in the evaluation of educational progress.

• Bayesian methods
• Empirical Bayes
• Generalized least squares
• Profile likelihoods
• E-M algorithm
• Fisher scoring procedures
• Both educational and social science applications