USE OF THE VERBAL REINFORCEMENT TECHNIQUE (VRT) IN DEVELOPMENTAL, PERSONALITY, AND SOCIALIZATION STUDIES OFTEN RESTS ON TENUOUS AND UNTESTED ASSUMPTIONS. THIS STUDY EXAMINED FIVE VARIABLES WHICH HYPOTHETICALLY RELATE TO PERFORMANCE UNDER REINFORCEMENT--SELF-ESTEEM OF S. TASK-INVOLVEMENT, EXPERIMENTER, ORDINAL POSITION, AND FAMILY SIZE. THE METHOD CONTROLLED AGAINST INFORMATIONAL ADVANTAGES FOR THE NEGATIVE OR POSITIVE REINFORCEMENT CONDITIONS. SS WERE COLLEGE MALES. POSTEXPERIMENTAL DATA (FROM AN INTERVIEW) INDICATED INDUCTION OF ALL THE MPUTED VARIABLES. RESULTS WERE THAT VR HAD A MAIN EFFECT ON PERFORMANCE (DIGIT-SYMBOL TASK). NEGATIVE REINFORCEMENT WAS SUPERIOR TO BOTH POSITIVE REINFORCEMENT AND TO NONEVALUATIVE INSTRUCTIONS, SUGGESTING A PSYCHOLOGICAL DIFFERENCE BETWEEN CRITICISM AND PRAISE. LOW SELF-ESTEEM SS IMPROVED MORE UNDER CRITICISM THAN DID HIGH SELF-ESTEEM SS. THE ONLY OTHER VARIABLE INTERACTING WITH REINFORCEMENT WAS FAMILY SIZE. SAMLL-FAMILY SS RESPONDED MORE TO REINFORCEMENT PER SE AND SPECIFICALLY TO POSITIVE REINFORCEMENT THAN LARGE-FAMILY SS. THIS FINDING APPEARS INCONSISTENT WITH THE SATIATION HYPOTHESIS. FAILURE OF OTHER VARIABLES TO INTERACT WITH VR, IN SPITE OF EVIDENCE THAT THESE VARIABLES WERE SUCCESSFULLY MANIPULATED, SUGGESTS THE ASSUMPTIONS OF MUCH VR RESEARCH NEED MORE DIRECT TESTING. (AUTHOR)