THE GROWING DISILLUSIONMENT WITH SOCIAL SCIENCE EVALUATION IS DUE TO OVERLY HIGH EXPECTATIONS AND A NARROW VIEW OF THE FUNCTION OF EVALUATION. TO DATE, EVALUATION HAS BEEN SEEN AS A METHOD TO IMPROVE PERFORMANCE TOWARD PRACTICAL PROGRAM GOALS. HOWEVER, EFFORTS TO IMPROVE DECISIONMAKING THROUGH EVALUATION HAVE HAD DISAPPOINTING RESULTS. OFTEN LITTLE CHANGE IS MADE AS THE RESULT OF A STUDY. EVALUATIONS SELDOM TAKE INTO ACCOUNT THE DIVERGING GOALS INHERENT IN ANY SOCIAL PROGRAM. OFTEN THE ONLY WAY A PROGRAM GETS STARTED IS THROUGH AMBIGUITY AND VERBAL COMPROMISE, WHICH CREATES A SUCCESSFUL POLITICAL SITUATION BUT MAKES EVALUATION DIFFICULT. IN ADDITION, SINCE IT IS IMPOSSIBLE TO SEPARATE A PROGRAM FROM THE REST OF SOCIETY, ANY INFORMATION GATHERING WILL HAVE BUILT-IN DIFFICULTIES IN SEPARATING PROGRAM EFFECTS FROM SOCIETAL EFFECTS. THE POSSIBILITY THAT EVALUATIVE RESULTS CAN ALTER POLICY IN THE SHORT RUN IS ALSO SERIOUSLY DIMINISHED BY THE FACT THAT THE DECISIONMAKING STRUCTURE INCORPORATES GREAT INERTIA. EVALUATION CAN, HOWEVER, BE VERY EFFECTIVE AS A FORM OF CONFLICT RESOLUTION. THE FACT THAT A PROGRAM IS UNDER STUDY IS A SIGNAL THAT THE FORM IS OPEN TO NEGOTIATION AND COMPROMISE. SIMILARLY, EVALUATIONS IN THE FORM OF PILOT PROGRAMS WITH A SPECIFIC GRANT PERIOD AND EVALUATION OFTEN LEAD TO SOME PERMANENT FORM OF SOCIAL CHANGE. EVEN THOUGH THE ORIGINAL EXPERIMENT MAY DIE, A NUCLEUS IS LEFT FOR A NEW APPROACH. EVALUATION ALSO REDUCES COMPLACENCY AMONG PROGRAM ADMINISTRATORS AND REDUCES CITIZEN ANXIETY. THE VERY ACT OF EVALUATING CONVINCES THE PUBLIC THAT SOMEONE WILL BE HELD ACCOUNTABLE. IT IS THIS LAST ASPECT, THE CALMING OF THE CITIZENRY, THAT GIVES EVALUATION ITS RITUAL ASPECT. EVALUATION SERVES TO COMBAT WIDESPREAD FEELINGS OF HELPLESSNESS BY FOSTERING THE BELIEF THAT THE GOVERNMENT IS ACTING ON EVERYONE'S BEHALF TO FIND SOLUTIONS TO PRESSING PROBLEMS. AN EXTENSIVE BIBLIOGRAPHY IS APPENDED. (GLR)