The Influence of Computerized Adaptive Testing (CAT) on Psychometric Theory and Practice

Abstract

            This article presents a review of some of the issues that were prominent in the early days of large-scale assessment and presents an argument that the development of computerized adaptive testing (CAT) was influential in changing the way that these issues were addressed.  It is argued that work on CAT increased the rate of adoption of item response theory (IRT) as the basis for scaling and reporting for large-scale assessments.  It also shifted the emphasis of test analysis from the number-correct score to characteristics of test items and statistics related to those item characteristics.  Further, the emphasis on interval scaling has declined because specifying the form of an item characteristic curve subsumes the assumption of an interval scale.  These changes are the result of a paradigm shift in educational and psychological testing that resulted from the shift from classical test theory approaches to IRT-based approaches.