EVALUATION PROBLEMS ARISE WHEN PROGRAM OUTCOMES ARE MEASURED AGAINST CERTAIN GOAL STATEMENTS USED TO INITIATE THE PROGRAM. THESE PROBLEMS ARE RELATED TO GOAL STRUCTURE; THE CUMULATIVE NATURE OF PROGRAMS AND KNOWLEDGE; THE CONTEXT IN WHICH PROGRAMS ARE DESIGNED, IMPLEMENTED, AND MAINTAINED; AND INTRAORGANIZATIONAL AND INTERORGANIZATIONAL REALITIES. IN ORDER TO HELP DECISIONMAKERS UNDERSTAND THE EFFECTS OF THEIR POLICIES AND THE FACTORS IMPINGING ON THESE EFFECTS, AND TO INCREASE THEIR KNOWLEDGE OF WHY THEY REPEATEDLY MAKE THE SAME MISTAKES, EVALUATION RESEARCH NEEDS TO FACILITATE THE PROCESS OF ORGANIZATIONAL LEARNING. ALSO EVALUATION RESEARCH SHOULD BE BASED ON AN UNDERSTANDING OF (1) THE INTERACTION AMONG GOALS, CONTEXT, AND THEORY; (2) THE NATURE OF PROGRAM DEVELOPMENT AS AN ONGOING, INTERACTIVE, AND CUMULATIVE PROCESS; AND (3) THE NATURE OF THE INNOVATION PROCESS. WHEN AN AGENCY FUNDS THE EVALUATION OF A SPECIFIC PROGRAM, THE FOCUS MUST BE THE PROGRAM DETAILS. HOWEVER, THE WAY A PARTICULAR PROGRAM NESTS INTO THE TOTAL SET OF AGENCY PROGRAMS IS IMPORTANT AND CAN BE SEEN FROM THREE PERSPECTIVES: (1) PROGRAM PORTFOLIOS, HOW THE PROGRAM COMPLEMENTS OR CONFLICTS WITH OTHER AGENCY PROGRAMS; (2) BALANCE ACROSS THE VARIOUS FUNCTIONAL RESPONSIBILITIES OF AN AGENCY; AND (3) HOW PROGRAMS INTERACT OVER TIME IN AN ONGOING, CUMULATIVE MANNER. CHARACTERISTICS OF THE INNOVATION PROCESS CAN AFFECT THE NATURE OF PROGRAM DEVELOPMENT, ITS FINAL OUTCOME, DEVELOPMENT, AND THE DISSEMINATION, ACQUISITION, AND UTILIZATION OF PROGRAM OUTCOMES. THESE, IN TURN, LEAD TO AN EXAMINATION OF THE FEASIBILITY AND ATTAINMENT OF GOALS, THE APPROPRIATENESS OF THE PROGRAM DESIGN, AND EVEN THE APPROPRIATENESS OF THE POLICY DECISION TO DEVELOP A PROGRAM. CASE STUDIES AND SCENARIOS ARE USED TO ILLUSTRATE THESE IDEAS. NO REFERENCES ARE PROVIDED. (PRG)