Program Outcomes: Conceptual and Measurement Issues

Peter Rossi, Ph.D., Stuart A. Rice Professor Emeritus, Director Emeritus, Social and Demographic Research Institute, University of Massachusetts at Amherst

The identification and specification of the desired outcomes of a program are critical features of program design. Failure to identify desired outcomes makes evaluation and accountability virtually impossible. Imprecise outcomes lead to contentious evaluations and uncertain accountability. Overly ambitious outcomes are an invitation to failure. The paper reviews the several definitions of outcomes, makes distinction among kinds of outcomes, such as proximal and distal, and examines some common measurement problems, especially those involving endogeneity and corruptibility. Guidelines are offered for specifying outcomes that are feasible, useful, and unlikely to be corrupted.

Return to Program

Return to Abstracts