Test Scoring and Analysis Using SAS by Ron Cody and Jeffrey Smith provides step-by-step instructions and examples for writing SAS procedures commonly used to carry out test and item analyses. These are supplemented with sage advice as to the rationale for conducting each procedure and guidance as to how to interpret and use results. A strength of the text is the care and attention given in three chapters to ensuring that test data is entered or transferred correctly and is checked for errors before it is analysed. Beginning with the basics of Classical Test Theory the authors show how to compute, display, and interpret CTT item statistics, distractor analysis, and quantile distributions. This leads nicely, and the authors clearly make the connection, to procedures used in Item Response Theory to calculate parameter values for item characteristic curves—this is a really nice pedagogical process so that the strong links between CTT and IRT are made apparent. Once poorly performing items are identified, the text shows how to drop offending items, recalculate test scores and test statistics; though this is done more thoroughly within the CTT approach than within IRT.

An unexpected strength, and perhaps an essential feature given the high-stakes of much testing in the United States, is the provision of procedures to detect possible collusion or cheating between candidates. Most texts talk about the possibility of cheating and presume that the provision of invigilation and multiple test forms is sufficient; the authors here provide three different approaches and code sets for checking for possible cheating. The summary chapter of codes will help those responsible for conducting institutional evaluations or for evaluating district or building tests quickly determine if the test items worked and if there was any cheating.

"This book is ideal for students and professionals who either begin to use SAS to perform test analysis, or have been doing this for years. It covers a range of basic and advanced topics on the process, evaluation, and analysis of testing data using SAS. For each topic, it introduces typical and common issues one would encounter, provides step-by-step approaches on solving the problems and presenting results, and offers an easily understandable guide on how to interpret the SAS output from the perspective of test measurement and assessment, which is rarely touched on in other general SAS programming books. The book also nicely introduces many useful SAS procedures and functions, and all sample programs are clearly explained and "macro-ized," so you don't have to be a SAS expert to be able to understand and use them.

"I wish I had the book when I started to work as a data analyst in the psychometric area, and I still found the book very informative and learned a great deal from it after over 10 years of working."