Fourth-Generation Evaluation

Fourth-generation evaluation, as its name implies, is the successor model incorporating three earlier generations of evaluation models (objectives, description, judgment) and moving beyond them to include intensive stakeholder participation in determining both the course of the evaluation and, as well, what actions should be taken on the evaluation results. In fourthgeneration evaluation, the definition of stakeholder is considerably expanded to include not only program funders and managers but also targets of the program (those who, but for the program, might have had a program of their own) and other members of the community who have an interest or stake in the outcomes of the program being evaluated.

Fourth-generation evaluation begins with sharply different assumptions, which lead, in turn, to different modes of operating in the context. First, such evaluation efforts begin with the premise that it is not merely a physical, tangible reality to which stakeholders respond, but to their social-psychological constructions—that is, the mental meanings, values, beliefs, and sensemaking structures in which humans engage to make meaning from events, contexts, activities, and situations in their lives. Thus, the collection of data on constructions assumes equal importance with collection of data on tangible realities (e.g., test scores, number of program participants, and the like). The collection, analysis, and evaluation of those constructions are central activities of the evaluation effort.

Second, it is assumed that the interaction between evaluators and stakeholders is an interactive epistemological exercise in which both sides arrive at a position that is more informed, more factual, more sophisticated, more data-rich, and more subtle. This is in sharp contrast to older models, in which stakeholders were assumed to be sources of data but not contributors to the nomination of critical issues or shaping of the evaluation effort. In fourth-generation evaluation, original program objectives are not the sole focus of evaluation data collection. Instead, objectives are joined with critical claims, concerns, and issues nominated by stake-holding audiences to enlarge the range of data gathering, discussion, and negotiation points between and among stakeholders. Furthermore, an epistemological commitment is made [Page 162]to expand the range of audiences that have access to data, information, and interpretations. Information is no longer concentrated in the hands of a small number of individuals but rather is shared widely in an effort to provide maximum evaluative responsiveness and the increased participation of stakeholders who have sophisticated information with which they might negotiate a program's future.

Third, fourth-generation evaluation enjoys an expanded methodological repertoire. Some evaluation efforts confine themselves to experimental and randomized-trial models, which rely heavily on mathematical and statistical models to generate data, but fourth-generation evaluation uses a data collection and analytic design that also incorporates, in additional to statistical approaches, a qualitative repertoire of methods. Although statistical and mathematical models do handle some forms of data well, it is only qualitative methods that are capable of collecting and analyzing the social constructions of stakeholders. Thus the methodological strategies of fourth-generation evaluators are often mixed in design, with fit of method to question rather than evaluator persuasion or training being the central methodological issue.

...