Subjective and Objective Methods of Evaluating Social Programs

Abstract

A practical problem in evaluation studies with limited resources is that all methodologies cannot be tried. Thus the evaluator must decide what questions must be answered and then choose which method would best provide the answers. The choice, however, is not simple. It involves several different trade-offs. These trade-offs are articulated through a discussion of the differences between (a) Bayesian and traditional statistics, (b) decision and cost benefit analysis, and (c) anthropological and traditional case studies. The intent of the article is to make evaluators sensitive to the implicit trade-offs associated with choosing a subjective or an objective method.