Review of Item-analysis data, their computation, interpretation

Abstract

Reviews Item-analysis data, their computation, interpretation, and use in test construction by Fred B. Davis (1946). The extensive literature and the number of computationally simple techniques of item analysis developed over the past fifteen years make questionable the need for a new method. If the need be granted, however, this bulletin presents a sound and accurate summary of the underlying logic and a method which should prove useful. The most serious omission is the complete lack of any reference to cross-validation or to the need for it. However, despite this and a few other omissions, the bulletin has a definite value in summarizing the uses and limitations of item analysis techniques and in presenting them as tools to be used with judgment rather than as a machine into which one can pour data, then turn a crank and extract a finished test. (PsycInfo Database Record (c) 2025 APA, all rights reserved)