Abstract

The article describes a method for developing input and output scales and for eliciting a subjective transformation function between input and output as measured by these scales. The inability to measure program achievement satisfactorily is well recognized as one of the principal problems in evaluating public programs. The goals to be reached are difficult to specify, are attributable to many origins, and these origins are hard to name. Precision in the analysis of input-output relationships and of the effect of program controls on output is also discouraged by this measurement problem. Even if the program's process is understood and well defined and controlling variables identified, evaluation still will be difficult. The validity of such an output scale would be no greater than the validity of the referents themselves. The scale should at least reflect empirical indicators used by the program manager as proxie criteria. Assuming that the scale has some reasonable validity reflecting the experience and understanding of the program manager, it would have begun to provide a framework within which the decision maker can express his judgment on the transformation function.