THE RESEARCH WAS PERFORMED IN TWO PHASES. IN THE FIRST PHASE, A MEASURE CONSISTING OF 44 BEHAVIORAL ROLEPLAYING AND PROBLEMSOLVING ITEMS--THE ADOLESCENT PROBLEMS INVENTORY (API)--WAS EMPIRICALLY DEVELOPED, ALONG WITH AN ITEM-SPECIFIC CRITERION-REFERENCED RATERS' MANUAL. THE INVENTORY WAS DESIGNED TO IDENTIFY STRENGTHS AND WEAKNESSES IN THE PERSONAL AND INTERPERSONAL SKILL REPERTOIRES OF ADOLESCENT BOYS. THE SECOND PHASE WAS THE VALIDATION OF THE API. INITIAL VALIDATION SHOWED THAT THE API RESPONSES OF INSTITUTIONALIZED DELINQUENT BOYS WERE RATED AS LESS COMPETENT THAN THE RESPONSES OF EITHER OF TWO NONDELINQUENT GROUPS OF TEENAGE BOYS FROM A PUBLIC HIGH SCHOOL. THE INVENTORY WAS SHOWN TO BE RELIABLE, COMPOSED OF ITEMS WITH LITTLE CLUSTER STRUCTURE, AND IT HAD EXTRAORDINARY DISCRIMINANT POWER. A SECOND VALIDATION STUDY COMPARED THE API RESPONSES OF TWO GROUPS: INSTITUTIONALIZED DELINQUENT BOYS WHO HAD FREQUENT BEHAVIORAL PROBLEMS WITHIN THE INSTITUTION AND INSTITUTIONALIZED DELINQUENT BOYS WHO HAD FEW ACTING-OUT PROBLEMS WITHIN THE INSTITUTION. THE FORMER GROUP RESPONDED LESS SKILLFULLY. A THIRD VALIDATION STUDY REPLICATED PREVIOUS GROUP DIFFERENCES BETWEEN DELINQUENTS AND MATCHED NONDELINQUENTS. TYPE OF DIRECTIONS GIVEN AND TEST FORMAT ALSO SIGNIFICANTLY AFFECTED PERFORMANCE. THE REPORT SUGGESTS THAT A SOCIAL RESEARCHER USING A SOCIAL SKILLS CONCEPTUALIZATION OF PERSONALITY SHOULD DO MORE THOROUGH ASSESSMENT STUDIES OF BEHAVIOR PATHOLOGIES BEFORE DEVELOPING LARGE-SCALE SOCIAL SKILLS TRAINING TRAINING PROGRAMS. REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--MJW)