A LACK OF CLEARLY STATED PROGRAM OBJECTIVES HAS BEEN ONE OF THE MAJOR PROBLEMS ENCOUNTERED IN EVALUATING PROGRAM EFFECTIVENESS. PROBLEMS ARE ALSO INHERENT IN THE USE OF SOCIAL INDICATORS AND IN THE APPLICATION OF THE MEASUREMENT PROCESS TO SOCIAL PROGRAMS. NECESSARY CHARACTERISTICS OF SOCIAL INDICATORS ARE VALIDITY AND RELIABILITY. IN ORDER TO REACH RATIONAL DECISIONS ABOUT PROGRAM PERFORMANCE, ACCURACY IS IMPORTANT AS WELL. ACCURACY PERTAINS TO MEASUREMENTS MADE WITH INDICATORS RATHER THAN TO INDICATORS THEMSELVES. GROUPS CAPABLE OF CONTRIBUTING TO DATA INACCURACY ARE PROGRAM DESIGNERS, ADMINISTRATORS, TREATMENT STAFF, PROGRAM PARTICIPANTS, AND PROGRAM RECORDKEEPING STAFF. THE TECHNIQUE OF CORRUPTING PROGRAM DATA CAN BE EXTREMELY POWERFUL SINCE THERE ARE SEVERAL LEVELS AT WHICH DATA CAN BE RENDERED INACCURATE: PROGRAM PARTICIPANT CHARACTERISTICS OR BACKGROUND DATA, PROGRAM TREATMENT DATA, AND PROGRAM OUTCOME DATA. ACCOUNTANTS CAN SERVE TO PREVENT ABUSES ASSOCIATED WITH MEASUREMENTS DERIVED FROM INDICATORS OF ANY DEGREE OF VALIDITY AND RELIABILITY THAT CAN BE MADE INACCURATE. THEY HAVE LITTLE CONTROL OVER INVALID OR UNRELIABLE SOCIAL INDICATORS. TWO SPECIFIC FUNCTIONS OF ACCOUNTANTS IN SOCIAL EVALUATION ARE INTERNAL CONTROL (DATA INACCURACY PREVENTION) AND AUDITING (DATA INACCURACY DETECTION). TYPES OF AUDIT EVIDENCE TO SUPPORT ASSERTIONS RELATED TO DATA COLLECTION AND EVALUATION ARE NOTED IN A FIGURE. NOTES ARE INCLUDED. (DEP)