The Greater Avenues for Independence (GAIN) welfare-to-work program in California was evaluated by the Manpower Demonstration Research Corporation (MDRC). The evaluation relied on involvement of stakeholders, examination of program implementation, flexibility, the adaptation of evaluation questions and designs during the study, use of random…

Proposals satisfy a number of functions in evaluation work. This article examines how these functions differ depending on whether the proposal concerns a preordinate study in which methods are prespecified or an emergent study in which decisions about methods are made as the study proceeds. (Author/SLD)

Propositions about what works in program evaluation are presented: (1) evaluators should be involved from the beginning; (2) evaluation should be based on explicit programmatic theory; (3) quasi-experimental designs must have high internal validity; and (4) evaluators must combat the deterioration of experimental control that occurs in…

Evaluation as a field has moved beyond debating whether certain designs are good while others are bad to begin to appreciate that each of a group of designs may have its uses. The importance of good designs that reflect real-world complexities is increasingly being recognized. (SLD)

The experiences of a group of evaluators evaluating a multisite integrated human service delivery system illustrate many design and measurement considerations in the preimplementation and implementation phases of evaluation. The reemergence of human service integration approaches will make such evaluation issues increasingly common. (SLD)

The evaluation of the preschool program conducted by the High/Scope Educational Research foundation is an example of a program evaluation that employed a randomized design and also examined long-term outcomes. The study, which employed a 22-year follow-up study, is important for its findings and its methodology. (SLD)

Based on experiences in working with school district staff, suggestions for enhancing evaluation use through stakeholder involvement are discussed in terms of who should be involved, when involvement should occur, and how stakeholders should be involved via a dialogue-dependent process. (SLD)

An experimenting society is one that vigorously tries out proposed solutions to recurrent problems and then tries other alternatives if evaluation shows that previous efforts are ineffective or harmful. Methods that social scientists could use if the experimenting society becomes a reality are explored. (SLD)

The application of time series analysis is described, focusing on the use of regression analysis for analyzing time series in a way that may make it more readily available to an evaluation practice audience. Practical guidelines are suggested for decision makers in government, health, and social welfare agencies. (SLD)

This paper summarized contributors' ideas about evaluation as a field and where it is going. Topics discussed were qualitative versus quantitative debate; evaluation's purpose; professionalization; program failure; program development; evaluators as advocates; evaluation knowledge; evaluation expansion; and methodology and design. (SLD)