THE CHALLENGE TO ADMINISTRATORS IN APPRAISING BROAD-AIM, BROAD-IMPACT UNDERTAKINGS SUCH AS PREVENTION AND REDUCTION OF CRIME AND DELINQUENCY IS TO APPLY MORE FLEXIBLE EVALUATIVE TECHNIQUES THAN THE TRADITIONAL ONE OF MEASURING THE PROJECT'S OBJECTIVES VERSUS ITS ACHIEVEMENTS. IN BROAD-AIM PROGRAMS, DESIRED OUTCOMES ARE UNDERSTOOD IN TERMS OF BROAD AND COMPREHENSIVE CHANGES, BUT THE SPECIFIC EVIDENCE OF THESE CHANGES ARE UNCERTAIN. AN ALTERNATIVE TO TRADITIONAL EVALUATION IS AN EXPERIMENTAL ONE, WHERE SELECTED ASPECTS OF THE SITUATION ARE CHOSEN FOR OBSERVATION AND ANALYSIS. MEASUREMENT IS APPLIED BEFORE AND AFTER THE PROGRAM. A CONTRARY ALTERNATIVE INCLUDES A RIGOROUS ATTENTION TO NON-INTENDED CONSEQUENCES, AN ACCEPTANCE OF 'NO-EFFECT' OUTCOMES, AND A BELIEF THAT SOME PROGRAMS ARE BEST EVALUATED QUALITATIVELY RATHER THAN QUANTITATIVELY. THE AUTHORS WHOSE THESES ARE CONTRASTED ARE WEISS AND REIN, AND CAMPBELL.