THE INSTRUMENT IN THE WAGNER HAND TEST CONSISTS OF 10 CARDS; 9 HAVE DIFFERENT DRAWINGS OF HANDS, AND 1 IS BLANK. THE TESTER PRESENTS THE CARDS, ONE AT A TIME, TO THE SUBJECT, WHO DESCRIBES WHAT THE HANDS ARE DOING. WHEN PRESENTED WITH THE BLANK CARD, THE SUBJECT IS TO IMAGINE A HAND AND DESCRIBE ITS ACTION. RESPONSES DURING TESTING ARE RECORDED, EACH RESPONSE BEING ASSIGNED TO ONE OF THE FOLLOWING CATEGORIES: AFFECTION, DEPENDENCE, COMMUNICATION, EXHIBITION, DIRECTION (DOMINATION OR CONTROL), AGGRESSION, ACQUISITIVENESS, ACTION (CONSTRUCTIVE OR MANIPULATIVE), PASSIVENESS, TENSION, DISABILITY, FEAR, BIZARRE DESCRIPTION, AND NO RESPONSE OR UNCLASSIFIABLE RESPONSE. A MEASURE OF AGGRESSIVENESS, THE ACTING OUT RATION (AOR), IS CALCULATED BY SUBTRACTING THE SUM OF RESPONSES ASSIGNED TO AFFECTION, DEPENDENCE, AND COMMUNICATION FROM THE SUM OF RESPONSES ASSIGNED TO AGGRESSION AND DIRECTION. AOR'S CALCULATED FROM THE TEST HAVE BEEN VERIFIED BY OTHER MEASURES. THE TEST'S ABILITY TO PREDICT AGGRESSIVENESS, HOWEVER, IS QUESTIONABLE. HAND TEST RESULTS COLLECTED FROM A STUDY OF AGGRESSIVENESS AND DELINQUENCY INDICATED NO CORRELATION BETWEEN THE AOR AND DELINQUENCY INDICATORS. SIMILARLY, IN A TYPOLOGICAL STUDY OF LEVELS OF INTERPERSONAL MATURITY, 70 SUBJECTS WERE PLACED INTO 3 HOMOGENEOUS GROUPS ACCORDING TO LEVEL OF INTERPERSONAL MATURITY (WHICH TAKE AGGRESSIVENESS INTO ACCOUNT). RESULTS OF THE HAND TEST DID NOT VARY SIGNIFICANTLY FOR THE THREE LEVELS, AS WOULD HAVE BEEN EXPECTED. A FURTHER STUDY OF 150 ADOLESCENTS IN QUEBEC COMPARED THE PREDICTED AGGRESSION FROM THE HAND TEST WITH THE ACTUAL INCIDENCE OF AGGRESSIVE BEHAVIOR; NO SIGNIFICANT CORRELATION WAS FOUND BETWEEN THE TWO. THE TEST'S APPARENT LACK OF VALIDITY MAY BE DUE TO TWO QUESTIONABLE ASSUMPTIONS IT MAKES: (1) THAT THERE IS A POSITIVE CORRELATION BETWEEN A SUBJECT'S VERBALIZED FANTASIES AND ACTUAL BEHAVIOR, AND (2) THAT ONE CAN PREDICT BEHAVIOR FROM SUCH FANTASIES. --IN FRENCH. (MBB)