Approach to evaluating the impact of AID projects

staff, go into the field for 3-4 weeks to make observations and to collect and analyze data. An overview of this methodology from its inception in 1979 to 1985 is presented. The report is in two parts. Part one describes the rationale, activities, and products that typify CDIE impact studies in contradistinction to other kinds of evaluations. Included is a detailed description, intended as an overview of the methodology for development workers, of the organization of evaluation teams and the procedures used in the impact studies (topic selection, topic coordinator, background studies, planning workshops, site selection, team composition, team workshops, fieldwork, conferences, and publications). Part two explores more specific methodological and operational issues, and is intended for those who anticipate direct participation in impact evaluations and who therefore require additional "how-to" guidance. The first of its three sections discusses evaluation design, including developing conceptual frameworks, identifying impacts, designing the study and developing working hypotheses, and selecting evaluatory techniques, resources, and measures. A second section discusses procedures for collecting and analyzing data from field observations and interviews. The final section presents strategies for reporting evaluation findings, including written reports, debriefing sessions, and the responsibilities of team members in concluding the evaluation process. The study emphasizes several characteristics which make the use of impact evaluations ideal for A.I.D. projects. (1) Although it is difficult to assess impact, the studies are valuable reminders of the difference development assistance has made. (2) Impact evaluations are important in helping A.I.D. policymakers address broad policy concerns. (3) The studies give development experts the opportunity to compare what was done with what should have been done and to develop alternatives for the future - a stance less common with more narrowly defined evaluations.