ALTHOUGH EVALUATORS TEND TO BE VIEWED BY OTHERS AND SOMETIMES BY THEMSELVES AS SCIENTIFIC AGENTS WHO REVEAL THE FINAL TRUTH ABOUT THE IMPACT OF PROGRAM INTERVENTIONS, EVALUATORS ARE THEMSELVES EXPERIMENTERS WHOSE EVALUATION DESIGNS, INSTRUMENTS, AND CRITERIA ARE RIDDLED WITH AMBIGUITIES AND LIMITATIONS. FOR EXAMPLE, FREQUENTLY THE TIME NEEDED TO DEVELOP THE INSTRUMENTS NECESSARY TO CONDUCT AN ADEQUATE EVALUATION OF A GIVEN PROGRAM IS NOT AVAILABLE. THE EVALUATOR THEN APPLIES A BATTERY OF INSTRUMENTS DEVELOPED BY OTHER INVESTIGATORS FOR SOMEWHAT DIFFERENT PURPOSES. FURTHER, GOVERNMENT GRANTS, UNDER WHICH MANY INTERVENTION PROGRAMS ARE FUNDED, HAVE TIME LIMITS THAT RESTRICT THE EXTENT AND CHARACTER OF CHANGE THAT CAN REALISTICALLY OCCUR OVER THAT TIME. WHICH GROUPS TO STUDY, WHAT SAMPLES ARE AVAILABLE, AND THE CHARACTERISTICS OF SUBGROUPS NEEDED FOR PROGRESSIVELY FINER ANALYSES ARE ALSO CRITICAL ASPECTS OF PREVENTION PROGRAM EVALUATION. ALSO, THE REPRESENTATIVENESS OF A SAMPLE MAY BE HAMPERED BY THE RESISTANCE OF SOME GROUPS TO PROGRAM PARTICIPATION. SAMPLE SIZE MAY BE LIMITED BY THE EVALUATION BUDGET. THE SEVERITY OF THE SOCIAL AND PSYCHOLOGICAL PROBLEMS OF SUBJECTS IN A GIVEN SAMPLE WILL ALSO DETERMINE THE KIND OF CHANGE THAT CAN REALISTICALLY BE EXPECTED OVER A GIVEN PERIOD OF TIME. WHILE THE SEVERELY HANDICAPPED MAY BE THE TARGET OF A GIVEN PROGRAM, MODERATE CASES MAY SHOW THE GREATEST IMPROVEMENT OVER THE SHORT TERM. EVALUATION CRITERIA SHOULD BE DEVELOPED WITH THE PERSPECTIVES OF VARIOUS INVOLVED OR INTERESTED GROUPS IN MIND (CLIENTS, FUNDING SOURCES, CLIENT PARENTS, EDUCATORS, ETC.). FINALLY, THE EVALUATION OF ANY PREVENTIVE PROGRAM IS ADVISED TO HAVE SEVERAL LEVELS OR KINDS OF CRITERIA, BECAUSE THE MORE WAYS IN WHICH A PROGRAM'S OUTCOME CAN BE INTERPRETED WITHIN THE FRAME OF REFERENCE OF INVOLVED OR INTERESTED GROUPS, THE MORE LIKELY IT IS TO RECEIVE CONTINUED SUPPORT.