Bayesian Analysis of Item Response Theory Models Using SAS Reviews

"Bayesian Analysis of Item Response Theory Models Using SAS is a very nice modern book on Bayesian IRT, which discusses a wide range of IRT models with Bayesian methods for analysis. The introductory style of the text, the thorough explanation of the models and Bayesian methods, will make the book useful for (under) graduate students, researchers, and measurement practitioners who are interested in the Bayesian analysis of psychometric models. This book presents applications and explained SAS code, which will make the discussed methods readily accessible."

Jean-Paul Fox
Professor, University of Twente
Department: Research Methodology, Measurement, and Data Analysis

"Bayesian statistical analysis in the context of item response theory (IRT) models is becoming increasingly popular. However, popular textbooks on IRT models do not include much detail, either on the theory behind such analysis or on how such analyses can be implemented using popular software packages. Bayesian Analysis of Item Response Theory Models Using SAS, written by two experts on Bayesian inference for IRT models, fills that gap. An added bonus is detailed discussions on model fit and model comparison that is a crucial aspect of model-based inference and yet is often neglected in textbooks. Non-technical descriptions of difficult concepts and attractive graphical plots make this book a joy to read.

"Because of its tutorial and example-driven structure, the book includes enough details to enable a first-year graduate student in Psychology or Education to learn about the topic and then implement Bayesian Analysis of Item Response Theory Models Using SAS. Even experts on Bayesian IRT or non-users of SAS will find the in-depth coverage of various aspects of the topic, references to existing research, and the real data examples extremely valuable. Numerous programs for conducting these analyses are amply provided and annotated so that the readers can easily modify them for their own applications."

Sandip Sinharay
Pacific Metrics

Clement Stone's and Xiaowen Zhu's Bayesian Analysis of Item Response Theory Models Using SAS introduces and illustrates implementation of Bayesian estimation of item response theory (IRT) models by using PROC MCMC in SAS. Well written and thorough, this book begins with a review chapter on IRT models for the following cases: unidimensional, multidimensional, testlets, bifactor, and models with random effects. The next chapter is an introduction to Bayesian analysis. Both chapters assume the reader has a background in the subject matter of the chapters.

"Chapter 3 is a general introduction to using PROC MCMC for Bayesian estimation of IRT models; application of PROC MCMC to estimation of the one parameter logistic model is illustrated. The subsequent three chapters address application of PROC MCMC to IRT models for dichotomous data, polytomous data, and extensions (i.e. bifactor, testlet, and hierarchical) of these models, respectively. The final two chapters address Bayesian comparison of IRT models and Bayesian model checking. To apply the methods, readers will need prior experience using SAS, but not with PROC MCMC. For these readers, Chapters 3 through 8 provide firm grounding in carrying out Bayesian estimation in SAS for the most popular IRT models.

"Bayesian Analysis of Item Response Theory Models Using SAS is a first-rate book and should serve the important goal of promoting Bayesian estimation of IRT models."

Professor James Algina
University of Florida