ONE OF THE GREATEST BARRIERS TO EVALUATION RESEARCH IS THAT TOOLS AND TECHNIQUES WITH WHICH SOCIAL AND BEHAVIORAL SCIENTISTS ARE EQUIPPED ARE OFTEN INADEQUATE FOR RESEARCH IN POLICE FIELD SITUATIONS. CONTROL GROUP-EXPERIMENTAL GROUP AND PRETEST-POSTTEST TECHNIQUES ARE NOT USEFUL FOR MEASURING SUCH THINGS AS THE QUALITY OF POLICE PROTECTION OR THE RESPONSIVENESS OF POLICING TO THE COMMUNITY. SINCE POLICE STUDIES DEPEND ON THE PERMISSION OF AUTHORITIES AND ACCEPTANCE BY OFFICERS, INVESTIGATORS IN THE FIELD OF POLICE SERVICE EVALUATION ADVISE A NOVICE TO BEGIN IMPROVING THE QUALITY OF THE RELATIONSHIP WITH OFFICIALS WHOSE COOPERATION IS VITAL TO THE PROJECT'S SUCCESS. PERSONNEL SELECTION IS ONE OF THE MOST FREQUENTLY RESEARCHED PROBLEMS. PSYCHOLOGISTS HAVE UTILIZED INTELLIGENCE TESTS, PERSONALITY TESTS, INTEREST TESTS, AND NUMEROUS OTHER INSTRUMENTS, BOTH SINGLY AND IN COMBINATION, TO IDENTIFY POTENTIALLY SUCCESSFUL POLICE OFFICER CANDIDATES. NO SURE PREDICTOR HAS BEEN DISCOVERED. THE INTERVENTION OF OUTSIDE FACTORS IN THE FORMATION OF A POLICE TRAINEE'S WORK CHARACTER ALSO CAN NEGATE A PRETRAINING PREDICTION. SEVERAL STUDIES ARE UNDERWAY THAT FOCUS ON TASK ANALYSIS, AND THESE MAY REDUCE THE MAGNITUDE OF THE CRITERION PROBLEM IN POLICE SERVICE EVALUATION. NOVEL APPROACHES TO POLICE SELECTION AND TRAINING HAVE BEGUN TO REVEAL DEFECTS IN TRADITIONAL SELECTION SYSTEMS. THE MOST IMPORTANT CONTRIBUTION OF EVALUATION MAY BE THAT IT UNCOVERS THE REAL PROBLEMS OF POLICE TRAINING AND SERVICE AND OFFERS GUIDANCE TOWARD THEIR RESOLUTION. (DEP)