ABSTRACT
So far in this volume we have only considered the selection problem, or the accuracy with which performance in a specific job can be predicted. Classification considers the problem of estimating aggregate outcomes if there can be a choice of job assignments for individuals. The objectives of the first part of this chapter are to (a) summarize the major issues involved in modeling the classification problem, (b) review alternative methods for estimating classification efficiency, and (c) outline the major alternative strategies for making differential job assignments. In the second part of the chapter, we report on the use of a recently developed method to estimate the potential classification gains of using the Project A Experimental Battery to make job assignments.
