Journal of Child Psychology and Psychiatry 45:2 (2004), pp 405411

Subtypes of self-esteem and self-concept in adolescent violent and property oenders

Middelheimhospital, University of Antwerp, Belgium; 2Yale University, USA

Aim: To investigate the relationship between subtypes of self-esteem/self-concept (familial, academic, social, and personal security) and antisocial behavior in adolescents (violent and property offending). Method: A self-report survey was administered in a representative group of 1466 students aged 12 to 18. Results: Both low family acceptance and low academic competence were found to predict property and violent offending, while high peer popularity was found to predict mainly violent offending. Conclusion: The nding that subtypes of self-esteem/self-concept present differently in antisocial adolescents may have clinical and scientic consequences, and such differentiation should be considered in future research. It should also be investigated whether this nding has therapeutic consequences. Keywords: Antisocial behavior, adolescence, delinquency, violence.

Both self-esteem and self-concept are important and closely related personality constructs in psychology that grow out of life experiences. Whereas selfesteem indicates how a person evaluates himself (emotional dimension), self-concept refers to how a person describes himself (cognitive dimension). Because diagnostic assessments (e.g., self-report) of these personality dimensions always cover both aspects, it may be justied to perceive self-esteem and self-concept as just one construct, reecting the way an individual perceives and evaluates his own acceptance and/or competence. Although the relationship between self-esteem/self-concept and antisocial behavior in adolescents has been investigated for many decades, considerable controversy remains (Glueck & Glueck, 1950; Baumeister, Smart, & Boden, 1996). While initial studies agreed on the presence of low self-esteem in antisocial adolescents, more recent reports have disputed these ndings (Baumeister et al., 1996). Part of this controversy may be explained by the heterogeneity of the construct. Several aspects of self-esteem have been described, e.g., in relation to family functioning, school performance, social perception and physical/psychological well-being. Theoretically, it may be that an adolescent perceives and values him/herself differentially in relation to each of these aspects. In relation to self-perception, Garnefski (2000) has postulated an analogous view, which further supports the necessity to consider differentiation. Much has been reported on the familial, the academic, and the social functioning of delinquent adolescents; however, studies on the physical/psychological perceptions of antisocial youth are rather scarce. In this respect, it has been reported that delinquent boys did not differ from non-delinquents on perceived physical competence, but lower scores

were found with regard to general self-esteem (Cole, Chan, & Lytton, 1989). Family conict, harsh and inconsistent parenting and low levels of parentchild involvement have been documented in families of antisocial youths (Ary, Duncan, Duncan, & Hops, 1999; Fergusson & Lynskey, 1997; Neighbors, Forehand, & Bau, 1997; Ruchkin, Eisemann, & Sidorov, 1997). Also, parents of antisocial youths have more deviant personality traits and psychopathology (Klein, Forehand, Armistead, & Long, 1997; Nigg & Hinshaw, 1998). Academic achievement is also an area of concern, as most antisocial adolescents perform poorly (Tanner, Davies, & OGrady, 1999; Fergusson & Horwood, 1998), and a large proportion have learning disorders of some kind (Hinshaw, 1992). Consequently, as research has shown that both family and academic problems appear nearly universal in antisocial adolescents, it is expected that antisocial youth will describe and value these areas negatively. Although antisocial youth often have a history of rejection by peers, close relationships with likeminded peers have been described (Fergusson et al., 1998; Moftt, 1993; Vitaro, Tremblay, Kerr, Pagani, & Bukowski, 1997). Also, it has been shown that association with antisocial peers, but not antisocial behavior as such, has signicant self-enhancing effects (Jang & Thornberry, 1998). A sample of delinquents was shown to have lower self-ratings of social competence compared to non-delinquents (Cole et al., 1989); however, as all the delinquents were incarcerated in a maximum-security facility at the time of the study, the authors acknowledge that these circumstances may have accounted for the lower scores. Recently, it was found that overestimation of peer acceptance in elementary school children is related to peer nomination for aggression (David & Kistner, 2000). Hence, in relation to social

Association for Child Psychology and Psychiatry, 2004. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

406

Robert Vermeiren et al.

functioning, it is likely that antisocial individuals describe and value themselves as at least as competent and accepted as non-antisocial peers, especially when measured in a community context. Considerable debate exists about the relationship between self-esteem and aggressive antisocial behavior in particular (Baumeister et al., 1996; Heaven, 1996; Pliszka, Sherman, Barrow, & Irick, 2000). Initially it was thought that low self-appraisal, as a consequence of perceived social exclusion, leads to deviant behavior which in turn stimulates selfenhancement (Bynner, OMalley, & Bachman, 1981; Leary, Schreindorfer, & Haupt, 1995; Rigby & Cox, 1996; Rosenberg, Schooler, & Schoenbach, 1989). Although these ndings were rather consistent in cross-sectional studies, low self-esteem has shown only modest predictive validity for future antisocial behavior (Heaven, 1996). More recently, Baumeister et al. (1996) has contended that the relationship between self-esteem and aggression has rarely been elaborated and that low self-esteem may only lead to violence in situations in which the fear of retaliation is minimal. Alternatively, it has been stated that aggression goes hand in hand with an overly favorable self-appraisal, called egotism, rather than with decreased self-esteem (Baumeister et al., 1996; Bushman & Baumeister, 1998). The theory of egotism suggests that increased self-esteem may lead to aggression when the individuals views are disputed. However, as most reports focus on aggressive antisocial behavior, studies describing levels of self-esteem in non-aggressive antisocial adolescents are needed. The current study builds on the nding that research with antisocial adolescents has not yet focused on subtypes of self-esteem/self-concept and that specic aspects of self-esteem/self-concept have not been investigated in non-aggressive antisocial individuals. Also, no reports have described levels of selfesteem/self-concept specically in girls. This study investigated by gender the relationships between four self-esteem/self-concept subtypes (perceived family acceptance, academic competence, peer popularity, and personal security), and self-reported property offending and violent offending, after controlling for potential confounding factors such as socio-economic status (SES), race and age. Additionally, we examined the relationships between the self-esteem/selfconcept subtypes and antisocial modeling by friends. Our hypothesis was that antisocial behavior would be related to low academic competence and low family acceptance. With peer popularity, either no relationship was expected or a higher level of this characteristic in antisocial adolescents.

protective factors for adolescent adjustment. The current study will report on results from a survey of 1634 students from Antwerp, Belgium. Absentee rates were requested from the teachers at each assessment and estimated to be around 15% (estimated target sample 1922 students). One hundred and sixty-eight (11.8%) surveys were incompletely or inconsistently lled out and have been excluded, resulting in a nal group of 1466 students. Participants ranged in age from 12 to 18 years (14.9 1.9). The group as a whole is comparable to the Flemish population in socio-economic status. The racial distribution is similar to that of the general population in Antwerp: 73.1% native Flemish, 11.3% Moroccan, 5.6% Turkish, and 9.9% others. The other group is very heterogeneous, with relatively large proportions of adolescents from southeastern Europe and middle Africa.

Procedure
Data were collected in 8 schools (middle and highschool levels) in the city of Antwerp, Belgium during the spring of 2000. Approval of the survey was provided by the relevant boards of the school system, as well as by those of the individual schools. Students and their parents were informed of the planned date of survey administration. All students were surveyed unless they declined to participate or their parents had objections (<1%). Administration of the survey was conducted in the classrooms. Before starting the assessment, students signed assent forms. The surveys were administered on a regular school day during two academic hours (50 minutes each) by trained personnel (psychologists, MDs or medical students), who read all the questions aloud, while students followed along circling answers on their protocols.

Instruments Socio-economic status. Parental education and current employment were used to assess socio-economic status. The variable parental education was transformed into three categories: did not complete high school (0), completed high school (1), and higher education (2). The variable current employment consisted of the categories no employment (0), industrial worker (1), employee (2), self-employed (3), and managerial function or professional (doctor, lawyer, etc.) (4). The nal scores were added for adolescents reporting on both parents and doubled for those adolescents who reported that one of the parents was unknown (e.g., because of death, being unknown etc.). Based on the nal scores, ve socio-economic categories were created. Antisocial behavior. Antisocial behavior items were derived from the questionnaire designed by Junger-Tas (1994). Topics included in the current analysis were severe property offenses (shoplifting, robbery, car theft, burglary) and violent offenses (vandalism, threatening, weapon carrying, and beating up someone). When a specic behavior was admitted (yes/no), severity was assessed by asking about the frequency of occurrence

Methodology
Participants
Participants were adolescents selected from an ongoing cross-cultural project that aims to assess risk and

Self-esteem and antisocial behavior

407

(once, 24 times, more than 5 times, more than 10 times, or more than 20 times). Responses were scored from 0 to 5 and separate total scores for property offending and violent offending were obtained by adding relevant items. Cronbach as for these scales were .61 (property) and .76 (violent).

Self-esteem/self-concept. The self-esteem index (SEI) is an 80-item self-report questionnaire (Brown & Alexander, 1991) designed to elicit childrens perceptions of their personal traits and characteristics. It is appropriate for individuals ages 8.0 through 19.0. The respondent answers each item using a modied Likert-type scale to classify each item as Always True, Usually True, Usually False, Always False. Because the items allow the respondent to describe and also to value him- or herself with regard to a range of characteristics, both self-esteem and self-concept are assessed. The continuous scaled scoring allows for an evaluation of positive as well as negative aspects; the scale scores represent a multidimensional construct, rather than a single, global entity of self-esteem. The 80 items exhibited satisfactory criteria for item discrimination and item difculty, while validity was demonstrated through strong correlations with other tests of self-esteem, personality and behavior (Brown & Alexander, 1991). Reliability, as assessed by internal consistency, appeared to be excellent. Four 20-item scales comprise the SEI, including the perception of: 1) academic competence scale, 2) family acceptance scale, 3) peer popularity scale and 4) personal security scale. The factor structure corresponded to the structure of the four SEI scales (Brown & Alexander, 1991). Family acceptance describes an individuals perception of her or his abilities, relationships, attitudes, interests as they pertain to home, parents, siblings, and family activities and interactions. Negative deviant scores indicate that the child does not feel accepted, does not trust or is not trusted by family members, does not enjoy or is not included in family activities, and does not feel comforted or supported at home, while deviant positive scores indicate disturbingly positive, glowing perceptions of home and family. Examples of items from this scale include I am an important member of my family, My parents dont listen to me, and My family doesnt trust me. Academic competence taps an individuals perception of her or his abilities, relationships, attitudes, and interests as they pertain to school, education, academic skill, intelligence, learning, and other scholarly or intellectual pursuits. Students with low scores are reporting difculties at school or in academically loaded situations, and do not feel competent to meet expectations and requirements, while students with high scores have very positive images of themselves as learners or achievers. Examples of items from this scale include: Im not doing as well in school as Id like to, I like going to school, and I do as little work at school as I can get by with. Peer popularity investigates an individuals perception of her or his abilities, relationships, attitudes, and interests as they pertain to the quality, importance, and nature of relationships and interactions with peers outside the family unit. Deviant scores on the negative side indicate problematic perceptions of what friends, classmates, and other peers think about them, their social and interpersonal skills

and the ease with which they interact with peers, and their leadership traits and characteristics. Examples of items from this scale include: My friends think I have pretty good ideas, I learn a lot from other people, and I am not shy. The personal security dimension reects a persons feelings about her or his physical and psychological well-being. The items ask about the individuals perception of safety, condence, vulnerability, or anxiety concerning specic life situations and distinctive traits of body character, conduct, temperament, and emotions. Youngsters with deviant scores may be extremely overanxious, concerned with their health and wellness, report a tendency to develop physical symptoms in response to stressful situations, be overcontrolled, withdrawn, and internalizing, and feel threatened and unprotected. High scores indicate that the child feels comfortable, safe, protected, and cared for. Examples of items from this scale include: I get a lot of headaches and stomach aches, Sometimes I pretend to know more than I really do, and I am often afraid. Cronbach as for each scale in this study were respectively .86 for family acceptance, .85 for academic competence, .76 for peer popularity and .78 for personal security.

Friends modeling. In this scale, students reported about the number of friends (none, a few, some, most or all) involved in risk-taking behavior. The 5 items were: dropping out of school, smoking cigarettes, drinking alcohol, using marijuana, and having sexual intercourse. Cronbach a was .80. The translation of these scales into Dutch followed established guidelines, including appropriate use of back translations (Sartorius & Kuyken, 1994). The translation into Dutch was done by three independent persons, followed by discussion of the translated questionnaires with colleagues to comment on them. Also, questionnaires were pre-tested in a clinical sample. Finally, an independent interpreter made back translations, which were compared to the originals. Statistical analyses
SPSS 10.0 (1999) was used for all analyses. First, bivariate correlations were computed between selfesteem/self-concept subscores, violent and property offense scores, and modeling by friends scores. Because both violent and property offending scores were highly skewed, both underwent logarithmic transformation. Second, hierarchical regression analyses were performed to test for a linear relationship between subscores of self-esteem/self-concept (independent) and property and violent offending scores (dependent). Age, SES, and race were entered as a block in step 1. Third, the same hierarchical regression analyses were repeated, entering additionally antisocial behavior scores and/or modeling by friends in step 1 (e.g., violent delinquency added in step 1 when property delinquency is used as a dependent variable). This method of analysis allowed examination of the effects of self-esteem on delinquency after having removed the effects of the demographic variables and the additional variables of interest in step 1. Selection of variables in step 2 was done by excluding variables with p-values >.05.

408

Robert Vermeiren et al.

Table 1 Bivariate correlations between self-esteem/self-concept subtypes, antisocial behavior and modeling by friends girls and boys separately Boys\Girls FA AC PP PS Viol Prop Model FA .46* .29* .37* ).22* ).23* ).25* AC .44* .17* .23* ).45* ).39* ).40* PP .25* .18* .33* .15* .06 .10* PS .41* .33* .51 .01 ).07 .05 Viol ).14* ).35* .08 ).04 .57* .45* Prop ).20* ).27* .08 ).05 .44* .31* Model ).25* ).30* .17* ).02 .15* .22*

*p < .001. FA: family acceptance; AC: academic competence; PP: peer popularity; PS: personal security; Viol: violent antisocial behavior; Prop: non-violent antisocial behavior; Model: modeling by friends.

Results
Bivariate correlations
Violent and property offending were both inversely correlated with family acceptance and academic competence (Table 1). Peer popularity was not correlated with offending (property and violent) in girls and not with property offending in boys, but positively with violent offending in boys. No relationship with personal security was found. Modeling by friends showed a similar inverse correlation with family acceptance and academic competence and was in both genders positively correlated with peer popularity. As expected, the correlations between property offending, violent offending and modeling by friends were moderate to high.

When entering violent offending, modeling by friends or both variables together additionally in step 1, academic competence was the only self-esteem subtype that (negatively) predicted property offending signicantly.

Predicting antisocial behavior by subtypes of self-esteem (hierarchical regression) Property offending. Over and above socio-demographic characteristics (age, SES, and race; Table 2), property offending was predicted by low family acceptance and low academic competence and high peer popularity in both boys and girls. Low academic competence was the strongest predictor in boys and girls (respectively 16% and 7% of variance), while both peer popularity and family acceptance were weak predictors (1% to 2% of variance).

Violent offending. In boys, when controlling for the effect of socio-demographic variables (Table 3), violent offending was predicted by all subtypes of self-esteem/self-concept. Low academic competence carried the strongest contribution (20% of variance explained), high peer popularity an intermediate contribution (5% of variance explained), and both low family acceptance and high personal security the weakest contribution (1% of variance explained). In girls, only low academic competence (13% of variance explained) and high peer popularity (3% of variance explained) predicted violent offending over and above socio-demographic variables. When entering property offending, modeling by friends or both variables additionally in step 1, academic competence and peer popularity remained as signicant negative predictors in both boys and girls.

Discussion
In a community sample of adolescents, subtypes of self-esteem/self-concept showed differential relationships with property and violent offending. In

Table 2 Predicting property offenses Boys Beta St 1 Age SES Etn AC PP FA PS ).083 ).020 .004 ).378 .166 ).119 t;p )2.57* ).553 .098 )10.58*** 4.96*** )3.22** D R2 .00 Beta ).031 .138 .026 ).229 .177 ).176 Girls t;p ).81 3.16** .59 )5.55** 4.53** )4.12** DR2 .02

St 2

.16 .02 .01

.07 .02 .02

* < .05; ** < .01; *** < .001. SES: socio-economic status; FA: family acceptance; AC: academic competence; PP: peer popularity; PS: personal security. St 1: step 1 of the hierarchical regression; St 2: step 2 of the hierarchical regression.

Self-esteem and antisocial behavior Table 3 Predicting violent antisocial behavior Boys Beta St 1 Age SES Etn AC PP FA PS ).030 ).072 .089 ).470 .222 ).106 .102 t;p ).98 )2.13** 2.62** )13.98*** 6.88*** )2.98** 3.04** DR
2

409

Girls Beta ).152 ).041 .065 ).394 .164 t;p )4.15*** )1.02 1.62 )10.66*** 4.42*** DR2 .01

.03

St 2

.20 .05 .01 .01

.13 .03

** < .01; *** < .001. SES: socio-economic status; FA: family acceptance; AC: academic competence; PP: peer popularity; PS: personal security. St 1: step 1 of the hierarchical regression; St 2: step 2 of the hierarchical regression.

both genders, low academic competence had the strongest association with both property offending and violent offending, while low family acceptance was more weakly related to both types of offending. Particularly in boys, high peer popularity was related to violent offending. Although academic competence, family acceptance, and peer popularity were signicantly related to offending in both genders, the variance explained was consistently higher in boys than in girls. The main message of this study is that selfesteem/self-concept may not be a homogeneous construct, and that it may be of scientic and clinical interest to recognize specic subtypes that constitute this domain of personality. Some aspects (academic competence, familial acceptance) were negatively associated with antisocial behavior, while others carried a positive (peer popularity) or no (personal security) relationship. It was expected that low levels of academic competence and family acceptance would be present in antisocial adolescents, especially when considering the range of problems these youths face (Ary et al., 1999; Fergusson et al., 1997; Neighbors et al., 1997; Ruchkin et al., 1997; Tanner et al., 1999; Fergusson et al., 1998). Because age and SES are known to correlate with both self-esteem/self-concept and antisocial behavior, it was considered appropriate to use these variables as covariates. The results show that self-esteem/self-concept predicts offending over and above the effect of the covariates. However, one may question the extent to which antisocial adolescents perceive academic competence or family acceptance as part of their self-concept. Others have demonstrated previously that such youngsters are oriented predominantly towards leisure/unconventional activities (van Welzenis, 1997), implying a reduced interest and involvement in scholastic and academic activities. It may be that the perception of low self-esteem is an idea imposed by concerned adults, rather then a true worry of these youth. The current nding of elevated peer popularity in violent adolescents apparently contradicts numerous studies describing poor peer relationships prior

to antisocial behavior (Rutter, Giller, & Hagell, 1998), but is in accord with some recent ndings on self-esteem and self-perception in antisocial adolescents (David et al., 2000; Jang et al., 1998). However, peer popularity does not show a strong bivariate relationship with property offending or violent offending. Only when other aspects of selfesteem/self-concept are entered does the dark side of peer popularity appear. Interestingly, peer popularity also correlated positively with friends modeling of deviant behavior. Because the assessment adopted a cross-sectional design, developmental pathways cannot be determined. Suggestions on possible developmental relationships are warranted and may help future research. First, one may posit that deviant children, as a result of being rejected by conventionally oriented peers (and adults), develop social relationships with like peers, a consequence that helps to maintain, or even enhance, aspects of self-esteem. Second, an unreasonably favorable evaluation of peer popularity, if present early in life, may predispose to the development of interpersonal problems, and direct these youth toward a deviant subculture. Finally, in both paths, life with friends becomes a dominant interest, and favorable cognitions and feelings may develop and compensate for decits in other domains. This view corresponds to the concept of egotism (Baumeister et al., 1996) in which the presence of an overestimated (interpersonal) selfesteem leads to conicts with the outer world. Finally, as this aspect of life is a main investment, questioning or threatening it may lead to uncontrollable feelings of loss and anger. The nding on peer popularity may carry consequences for clinical intervention and treatment. A relationship may exist between our ndings and Dishion, McCord, and Poulins (1999) report that peer-group interventions increase adolescent problem behavior and negative life outcomes in adulthood. It may be suggested that group interventions with antisocial adolescents, especially when done in an unstructured way, stimulate further deviancy because of mutual direct or indirect peer inuences among participating youngsters. This hypothesis

410

Robert Vermeiren et al.

and the processes through which such reinforcement takes place should be investigated, as the development of future prevention and intervention programs would benet from such insight. Some limitations of the current study should be noted. First, all data were derived by self-report. Whereas this may be seen as a limitation, others have argued that self-report surveys are a valid source of information, particularly when used for assessing antisocial behavior (Junger-Tas, 1994). Second, peer popularity was assessed only as reported by adolescents themselves. The absence of other sources of information does not allow us to determine how the delinquent adolescents are actually perceived by their classmates, and thus to know whether their popularity is accurate or a misperception. Last, the cross-sectional nature of the study does not allow causal conclusions; however, suggestions for future prospective research may stem from these ndings.

Author note
This paper was written during a research fellowship of the rst author at the Child Study Center, Yale Medical School, sponsored by the Belgian American Educational Foundation. The authors gratefully acknowledge the support and helpful guidance of Donald J. Cohen, MD, the late Director, Yale Child Study Center.

Correspondence to
Robert Vermeiren, Middelheimhospital, University of Antwerp, Lindendreef 1, 2020 Antwerp, Belgium; Email: rvermeiren@europemail.com

References
Ary, D.V., Duncan, T.E., Duncan, S.C., & Hops, H. (1999). Adolescent problem behavior: The inuence of parents and peers. Behaviour Research and Therapy, 37, 217230. Baumeister, R.F., Smart, L., & Boden, J.M. (1996). Relation of threatened egotism to violence and aggression: The dark side of high self-esteem. Psychological Review, 103, 533. Brown, L., & Alexander, J. (1991). Self-esteem index. Austin, TX: Pro-Ed. Bushman, B.J., & Baumeister, R.F. (1998). Threatened egotism, narcissism, self-esteem, and direct and displaced aggression: Does self-love or self-hate lead to violence? Journal of Personality and Social Psychology, 75, 219229. Bynner, J.M., OMalley, P.M., & Bachman, J.G. (1981). Self-esteem and delinquency revisited. Journal of Youth and Adolescence, 10, 407441. Cole, P.G., Chan, L.K., & Lytton, L. (1989). Perceived competence of juvenile delinquents and nondelinquents. Journal of Special Education, 23, 294302.

David, C.F., & Kistner, J.A. (2000). Do positive selfperceptions have a dark side? Examination of the link between perceptual bias and aggression. Journal of Abnormal Child Psychology, 28, 327337. Dishion, T.J., McCord, J., & Poulin, F. (1999). When interventions harm. Peer groups and problem behavior. American Psychologist, 54, 755764. Fergusson, D.M., & Horwood, L.J. (1998). Early conduct problems and later life opportunities. Journal of Child Psychology and Psychiatry, 39, 10971108. Fergusson, D.M., & Lynskey, M.T. (1997). Physical punishment/maltreatment during childhood and adjustment in young adulthood. Child Abuse and Neglect, 21, 617630. Garnefski, N. (2000). Age differences in depressive symptoms, antisocial behavior, and negative perceptions of family, school, and peers among adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 11751181. Glueck, S., & Glueck, E. (1950). Unraveling juvenile delinquency. Cambridge, MA: Harvard University Press. Heaven, P.C. (1996). Personality and self-reported delinquency: A longitudinal analysis. Journal of Child Psychology and Psychiatry, 37, 747751. Hinshaw, S.P. (1992). Externalizing behavior problems and academic underachievement in childhood and adolescence: Causal relationships and underlying mechanisms. Psychological Bulletin, 111, 127155. Jang, S.J., & Thornberry, T.P. (1998). Self-esteem, delinquent peers, and delinquency: A test of the self-enhancement thesis. American Sociological Review, 63, 586598. Junger-Tas, J. (1994). Delinquent behaviour among young people in the western world. Amsterdam/New York: Kugler Publications. Klein, K., Forehand, R., Armistead, L., & Long, P. (1997). Delinquency during the transition to early adulthood: Family and parenting predictors from early adolescence. Adolescence, 32, 6180. Leary, M.R., Schreindorfer, L.S., & Haupt, A.L. (1995). The role of low self-esteem in emotional and behavioral problems: Why is low self-esteem dysfunctional? Journal of Social and Clinical Psychology, 14, 297314. Moftt, T.E. (1993). Adolescence-limited and lifecourse-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674701. Neighbors, B.D., Forehand, R., & Bau, J.J. (1997). Interparental conict and relations with parents as predictors of young adult functioning. Development and Psychopathology, 9, 169187. Nigg, J.T., & Hinshaw, S.P. (1998). Parent personality traits and psychopathology associated with antisocial behaviors in childhood attention-decit hyperactivity disorder. Journal of Child Psychology and Psychiatry, 39, 145159. Pliszka, S.R., Sherman, J.O., Barrow, M.V., & Irick, S. (2000). Affective disorder in juvenile offenders: A preliminary study. American Journal of Psychiatry, 157, 130132. Rigby, K., & Cox, I. (1996). The contribution of bullying at school and low self-esteem to acts of delinquency among Australian teenagers. Personality and Individual Differences, 21, 609612. Rosenberg, M., Schooler, C., & Schoenbach, C. (1989). Self-esteem and adolescent problems: Modeling

Self-esteem and antisocial behavior

411

reciprocal effects. American Sociological Review, 54, 10041018. Ruchkin, V., Eisemann, M., & Sidorov, P.I. (1997). Parental rearing: A comparison between juvenile delinquents and controls in Russia. International Journal of Circumpolar Health, 56, 8689. Rutter, M., Giller, H., & Hagell, A. (1998). Antisocial behavior by young people. New York, USA: Cambridge University Press. Sartorius N., & Kuyken, W. (1994). Translation of health status instruments. In J. Orley & W. Kuyken (Eds.), Quality of life assessment: International perspectives (pp. 318). Berlin/Heidelberg: Springer-Verlag. SPSS Inc. (1999). Statistical Package for the Social Sciences 10.0 for Windows. Chicago, IL: SPSS Inc.

Tanner, J., Davies, S., & OGrady, B. (1999). Whatever happened to yesterdays rebels? Longitudinal effects of youth delinquency on education and employment. Social Problems, 46, 250274. van Welzenis, I. (1997). The self-concept of societally vulnerable and delinquent boys within the context of school and leisure activities. Journal of Adolescence, 20, 695705. Vitaro, F., Tremblay, R.E., Kerr, M., Pagani, L., & Bukowski, W.M. (1997). Disruptiveness, friends characteristics, and delinquency in early adolescence: A test of two competing models of development. Child Development, 68, 676689. Manuscript accepted 3 December 2002