The Difficulty of a Test and Its Factor Composition Revisited

Abstract

In the 30 years since Fred Lord earned his doctorate in psychology from Princeton University, we have witnessed solid progress in the elaboration of mathematical models for conceptualizing the nature of psychological tests and in the sharpening of methodologies for analyzing test data. These scholarly and theoretical advances will, I am sure, be fully developed and debated by others at this conference. Yet severe limitations have been imposed on the application of these advances to the practical task of developing tests as a result of a number of social and political developments during this same period. As my contribution to this gathering, I have elected to identify some of these limiting developments and to comment on their impact on the endeavors of the practical test developer.