Anastasi's orignal textbook is a difficult text book for any graduate student. Ms Urbina simplifies and gives examples of typical exam questions. It reviews the nature of psychological tests, statistical concepts in reliability, validity, item analysis, ability testing, personality testing, and finally a good section on the application of testing.

A must for any student of psychology!

This text is relatively complete and is an excellent reference for graduate level courses but is frustratingly difficult for undergraduates. For undergraduates, the text is obscure, difficult to understand, and lacks sufficient examples, underlining, italics, case boxes, and other reader's aids. Many of my brightest and best undergraduate students found it very difficult to summarize and retain information abstracted from the text. Most have been highly critical not only of the text but of the very poor selection of test items in the instructor's test bank. Many distractors are nonfunctional, the wording is vague, and the correct alternative is sometimes obscure even on open book/open note tests. It is paradoxical that a text in test theory and construction has produced a test bank with poorly constructed multiple choice items, as my classes have consistently demonstrated through item analysis using the principles outlined in the text! Second, in some areas the text is insufficiently complete for graduate students. Frequently, the authors mention new technologies or approaches but with sufficient lack of detail to be frustrating. Finally, the text suffers from antiquated language (e.g., sentences that begin with: "It will be remembered that ...") and excessive use of the passive voice, making the text unnecessarily difficult to understand.

Inappropriate for undergrads but ideal for grad students. Anastasi's classic text (now updated), has transformed generations of "intuitive" clinicians-in-training into serious behavioral scientists. Note that this is _not_ a great sourcebook for how to administer or evaluate _specific_ tests... Rather, it is an outstanding discussion of the science of assessment, e.g., in _general_, what would it mean for a questionnaire to be "valid"? Or, in _general_, what would it mean for a test to be "biased"?