“…To examine the possibility of systematic bias in the representation of the classification groups in the laboratory subsample, analyses were conducted on the degree of aggression and withdrawal m the sample that agreed to participate relative to those who refused to participate or were not found A oneway multivanate analysis of variance (MANOVA) was calculated for each peer classification group The mdepjendent variable was retrieval status, and the dependent variables were the aggression and the withdrawal scores There were no significant differences in scores on aggression or on withdrawal as a function of retneval status for any of the groups Thus, the subsample tested m the laboratory appears to be representative of the sample from which it was drawn with respect to degree of scores on aggression and withdrawal Procedure Parents and subjects were contacted first by letter descnbing the study and then by a follow-up telephone call Wntten parental consent was obtained for individuals younger than 18 years Individuals older than 18 provided wntten consent for themselves Individuals were paid for their participation Intelligence To assess intelligence, six subtests from the WISC-R or the WAIS-R (for individuals 17 years and older) were administered, and the scaled scores were summed The subtests were Information, Vocabulary, Comprehension, Similanties, Block Design, and Picture Arrangement The first five are the best measures of general intelligence (Kaufman, 1979), and the sixth was added to have a second measure of intelligence in addition to Block Design that is relatively less affected by verbal abilities Scaled subtest scores were denved using the norms provided by the WISC-R and the WAIS-R manuals behavior problems and social competence Individuals completed the Youth Self-Report (YSR, Achenbach & Edelbrock, 1983) This measure has two components, one examining behavior problems and the other examining social competence From this measure the following scores were analyzed the total scores for behavior problems and social competence and the subscale scores for internalizing, externalizing, activities, school competence, 3nd sociai aspects of sociai competencê ocioeconomic status Socioeconomic status was measured using the Household Prestige Scale (Nock & Rossi, 1978 This measure of socioeconomic status utilizes information concerning the occupational and educational status of both parents This index has been reliably coded in several studies and has high test-retest reliability (Mueller & Parcel, 1981) Social desirability Social desirability was assessed with a translation of the approval-seeking measure developed by Crowne and Marlowe (1964) This scale was used to control for the tendency to provide responses that seek social approval in the analysis of the self-reports of behavior problems and social competence (YSR) As demonstrated by Paulhus (1984), this scale measures both the self-deception and the impression management components of social desirability…”