THE DECISION-THEORETIC PARADIGM TAKES INTO ACCOUNT THE GOALS AND OBJECTIVES OF FORMAL AND INFORMAL DECISIONMAKING GROUPS INVOLVED IN THE PROGRAM BEING EVALUATED AND MAKES IT POSSIBLE TO RELATE WHAT IS HAPPENING IN THE PROGRAM TO THESE GOALS AND OBJECTIVES. IT IS POINTED OUT THAT A PARADIGM IS AN OVERALL MODEL THAT ENCOMPASSES ALL EVALUATIONS -- A WAY OF STRUCTURING THE CONCEPTUAL AND METHODOLOGICAL STEPS IN ANY EVALUATION. WITHIN THE DECISION-THEORETIC PARADIGM OF EVALUATION, EXPERIMENTAL METHODS ARE BUT ONE OF MANY POSSIBLE METHODS THAT CAN BE USED TO GENERATE CERTAIN TYPES OF DATA. NO A PRIORI CRITERA ARE IMPOSED ON PROGRAMS. WHATEVER THE PROGRAM IS, IT CAN BE EVALUATED, REGARDLESS OF THE CERTAINTY OF THE INFORMATION ABOUT THE PROGRAM. IF ONE TAKES THE EXPERIMENTAL METHOD AS AN OVERALL PARADIGM FOR EVALUATION, THEN MANY PROGRAMS CANNOT BE EVALUATED BECAUSE THEY DO NOT CONFORM TO THE REQUIREMENTS FOR THE USE OF THE EXPERIMENTAL METHOD. POINTS OF AGREEMENT AND DISAGREEMENT WITH THE ARGUMENTS PRESENTED IN FAVOR OF THE EXPERIMENTAL, AS OPPOSED TO THE DECISION-THEORETIC, PARADIGM (SEE NCJ-47350) ARE NOTED. IT IS EMPHASIZED THAT THE AIM OF EVALUATIONS IS TO HELP DECISIONMAKERS MAKE THE BEST INFORMED JUDGMENTS ABOUT PROGRAMS. (AUTHOR ABSTRACT MODIFIED--LKM)