مسیر های توسعه از سوء استفاده پدر و مادر تا بزهکاری: نقش واسطه افسردگی و پرخاشگری

Abstract This study investigated the long-term relationship between abusive parenting and adolescent mental health, and the path to delinquent behavior. Longitudinal data from 5th through 7th graders from the Korean Children and Youth Panel Survey (KCYPS) were analyzed to examine if abusive parenting was a predictor of early adolescent delinquency behavior, via aggression and depression as mediating factors. The results were as follows. First, parental abuse (both emotional and physical) was found to have significant effects on children's psychosocial factors (aggression and depression), while parental neglect (both emotional and physical) had significant effects on depression alone and not on aggression. Second, aggression exerted significant effects on both violent and non-violent delinquent behaviors, while depression had a significant effect on only non-violent delinquent behaviors. Third, children's psychosocial factors (aggression and depression) played significant mediating roles between earlier abusive parenting and delinquent behaviors. Fourth, for children living in a family with their grandparents, paths from abusive parenting, psychosocial adaptation, and later delinquent behaviors were not significant, implying that living with grandparents played a protective factor in these relationships.

Introduction In Korea, legislative revisions for child abuse prevention are relatively recent. It was not until January 2000 that clauses related to child abuse were added to official protection laws and child protection institutions were systematized. Thus, a systematic approach in preventing children from risk of abuse has not yet been fully formulated (Kim & Yoo, 2012). The cultural perception of child abuse in Korea is that it should remain more or less a family issue, that is, a private matter (Yon, 1992). This perception makes it difficult to prevent child abuse and protect child victims from abusive situations. In Article 3 No. 7 of the Child Welfare Act of Korea, child abuse is defined as “The behavior of adults or guardians inflicting physical, psychological, and sexual abuse that can damage the child's health, welfare, or normal development, and the act of child neglect or abandonment”. The National Child Protection Agency (2012) is in charge of all national child abuse prevention programs. According to their guidelines, there are four child abuse subtypes: Physical abuse, Emotional abuse, Physical neglect, and Emotional neglect. Different types of abuse have distinctive physical and behavioral characteristics. Child abuse occurs when the guardian's hostile, premeditated, and aggressive behavior intentionally harms the child. To be specific, physical abuse is to inflict deliberate or accidental bodily injury on a child. In cases of physical abuse, the parent or guardian inflicts physical injury and pain, such as intentionally and deliberately hitting or inflicting burns, and causing harm through physical violence (Kwack, 2005). Emotional abuse is to cause emotional wound and pain to a child and refers to abusive acts that hinder the self-esteem or social interaction of the child. For example, an emotionally abusive parent or guardian may ignore the child or say words that insult and threaten the child (Kwack, 2005). While abuse is related to direct action toward the child, neglect has to do with lack of action or care. Child neglect is when the guardian does not respond to or adequately care for the needs of the child and remains indifferent toward the child's welfare (National Child Protect Agency, 2012). Physical neglect is not providing or permitting the appropriate measures needed for a child's nurturing and growth, or behaviors that delay them (Ahn, 2000). Emotional neglect occurs when the parent does not provide the emotional support or protection needed for the healthy development of a child's personality (Ahn, 2000). While there are differences in the types of abuse, the common link among them is the negative outcome resulting from all forms of abuse. The negative effects of abusive parenting on adolescence have been widely reported. Childhood abuse was found to have negative impacts on individual behavioral problems, including alcohol and drug addiction, family violence, and suicidal tendencies, as well as social problems such as juvenile delinquency and criminal acts (Dube et al., 2001, Felitti et al., 1998 and Osofsky, 1999). Empirical studies showed approximately one-third to two-thirds of juvenile delinquents have experienced some form of child abuse (Wiebush, Freitag, & Baird, 2001); and those who have experienced child abuse are more likely to engage in crime (Currie & Tekin, 2006) and adolescent delinquency (Hwang, 2009). For example, many of those who had committed serious criminal acts, such as violent and sexual crimes, were found to have experienced chronic abuse in childhood (Hamilton, Falshaw, & Browne, 2002). The negative impact that abusive parenting has on adolescent development may extend beyond disruptive behavioral and social problems to deeper levels of psychological and emotional maladjustment. Adolescent psycho-emotional problems are another layer related to abusive parenting. Previous studies have shown that children who have experienced family abuse are generally more aggressive than those who have not (Chung et al., 2006 and Salzinger et al., 1993), and higher levels of aggression eventually lead to violent behaviors (Finzi et al., 2001, Kim and Choi, 2011, Lee and Yoo, 2011, Swogger et al., 2011 and White and Widom, 2003). According to social learning theory (Bandura, 1973), these findings imply that domestic violence could create the environment in which children might develop and internalize favorable values on violence, which later leads to delinquent behaviors (Luntz and Widom, 1994, Smith and Thornberry, 1995 and Um, 2001). Moreover, adolescents’ psychological malfunctions, including depression and aggression, are manifested as external problems, such as adjustment problems at school, hypergasia, and delinquency. That is, depressed and aggressive children might choose external delinquent actions as a possible way of showing their internal problems (Cantwell and Baker, 1991, Kim and Nam, 2012 and Koening, 1988). These results indicate the need to focus on depression and aggression, which are common psycho-emotional struggles of abused and delinquent adolescents. The evidence of the damaging effects of abusive parenting on abuse victims lends support for the idea that there are different levels of impact. The negative effects of abuse include maladjustments on physical, behavioral, and psychological levels. For adolescents in particular, abuse inhibits physical, psychological, emotional, and social growth and development. The general belief of parents and guardians as protectors, providers, role models or nurturers for adolescents’ well-being is antithetical in meaning for circumstances where it is the parents or guardians themselves who are the perpetrators of abuse. Rather than protecting, abusive parents are inflicting harm that impact children beyond adolescence into adulthood. Victims of abuse could possibly experience lifelong behavioral and social problems that may affect their quality of life. Therefore, this study aims to explain the relationships from abusive parenting to psycho-emotional problems and delinquent behaviors. Literature review Child Maltreatment and Its Impact on Delinquency The family is often the foundation from which people first begin to learn the basic patterns of behaviors and roles; it is also within the family unit that people experience relationships and these experiences play a significant role in determining his or her future behavior (Um, 2001). Thus, family environment and parenting are significant factors in studying adolescent delinquency. According to previous research, parenting behaviors exerted significant effects on delinquency (Ahn, 2010, Baldry, 2003 and Kim and Park, 2002). Chronic and repeated child abuse resulted in severe youth crimes such as violent crime and sexual crime (Hamilton et al., 2002), and abused children had higher chances of experiencing juvenile imprisonment (Currie & Tekin, 2006). In particular, studies showed that physical abuse had the biggest impact on adolescent delinquency (Hwang, 2009 and Um, 2001), and that abused children showed a higher level of aggression (Mersky & Reynolds, 2007). Nevertheless, the direct influence of abusive parenting is not solely responsible for adolescents’ delinquency, as not all victims of child abuse show delinquent behaviors. In fact, a large body of research has argued for the indirect effect of abuse on delinquency (Cho and Kang, 2010, Chung and Kim, 2012, Kim and Nam, 2012, Stuewing and McCloskey, 2005 and Yoder et al., 2001). Thus, the influence of abusive parenting on delinquency needs to be understood not as a simple causal relationship, but rather a complex process of abusive parenting and victims’ experiences. These studies seem to suggest a mediating mechanism within the patterns of abuse as it affects victims’ experiences and negative outcomes. For example, it may not be as straightforward as physically abuse directly causes adolescents to become violent. The indication of indirect effects takes into account a more comprehensive impact on adolescents’ well-being that includes physical, psychological, emotional, and social development. Among these adolescent development factors, one of the indicators mentioned in discussions on the indirect effects of abuse on adolescent behaviors is psycho-social development. Child Abuse and Psycho-Social Adaptation Child abuse has been reported to affect psycho-social adaptation. Among the many psycho-social adaptation problems, we focus especially on depression and aggression. Depression, an internal problem behavior, was shown to be continuously related to child abuse (Gilbert et al., 2009, Harkness and Lumley, 2008, Kim, 2009 and Springer et al., 2007). According to the National Comorbidity Survey on 873 participants aged between 15 and 54 who had experienced child abuse, child abuse was found to have a significant relationship with depressive disorder and anxiety disorder in adolescence and adulthood (Afifi, Boman, Fleisher, & Sareen, 2009). In the same vein, Kim and Choi (2012) used the Korea Welfare Panel Study based on 612 Korean participants aged between 12 and 15 to investigate the relationship between child abuse and negative emotion. They reported that children who had more experiences of abusive parenting tended to display a higher level of depression than those who had not. Furthermore, a survey on 1,310 children aged 18 and below who had experienced abuse in the past year indicated that high-risk children had a higher level of depression than low-risk children (Ministry of Health & Welfare, 2012). The other factor within psycho-social adaptation is aggression. Aggression, an external problem behavior, is reportedly influenced by child abuse and can be present from childhood through adulthood (Brodsky et al., 2001, Chung and Lee, 2012, Gilbert et al., 2009, Hodgdon, 2009 and Kim, 2009). For instance, in Korea, Chung (2008) studied 904 children in poverty and found that child abuse had effects on aggression, leading to isolation from the peer group. The Ministry of Health and Welfare (2012) also indicated that high-risk children had a higher level of aggression than low-risk children. The suggested relationships between abuse and adolescent depression and aggression, as well as, impacts on later negative behaviors seem to suggest that psycho-social adaptation may be related to adolescents’ participation in other maladaptive behaviors such as delinquency. Psycho-Social Adaptation and Delinquency Delinquency has a complex relationship with diverse emotional and behavioral problems (Angold et al., 1999, Kim, 2000 and Lee et al., 2004). Kim's study (2000) results showed that the delinquency of adolescents on probation was significantly correlated with each of the depression, anxiety, contraction, and body symptoms, indicating that adolescent delinquency is closely related to mental health. Also, research has shown that juvenile delinquents displayed a higher level of depression than other adolescents, and that adolescent depression was manifested as external problems such as defiance, self-abuse, delinquency, behavioral problems at school, learning difficulties, hypergasia, and aggressive behaviors (Kim & Baik, 2000). Study results also showed that adolescent depression had an influence on delinquency through the mediation of rejection from peer groups and drug use (Vaske & Gehring, 2010). Also, the idea that adolescents who were depressed and had experienced rejection and harsh parenting was linked to crime acts (Stuewing & McCloskey, 2005) also supports the suggestion that depression may precede delinquency experience. Previous studies also addressed aggression as another emotional factor of juvenile delinquents; they reported that aggression is an important predictor of adolescent delinquency (Kim and Lee, 2001 and Van Manen et al., 2004). Jin and Bae's (2012) meta-analysis study on the related variables of adolescent delinquency revealed that aggression had a larger effect than other psychological factors, and was a major factor for adolescent delinquency over time. Furthermore, aggression was identified as an important predictor of bullying, as bullying adolescents displayed a higher level of aggression (Cho, 2011). The presence of adolescent depression and aggression related to types or patterns of abuse and adolescent delinquency shows the complexity in understanding how parental abuse cycles through the development of children affecting layers of behaviors and psycho-social factors. For example, being a victim of abuse may leave a child more prone to developing depressive symptoms or sensitive to developing aggressive behaviors. Depression and aggression, in turn, may influence the child to engage in delinquent behaviors. Of concern within this repeated system of abuse and its effects on delinquency is where in the process are there factors that can prevent or reduce the negative effects of abuse. Protective Factor The current study aimed to identify possible protective factors and mechanisms that influence adolescent delinquency. With regard to children exposed to stressful life events, Garmezy (1985) attributes resilience to protective factors that mediate adverse rearing conditions. Personality factors, family practices, and external support system have been mentioned as three sources that contribute to resilience. A study by Lee and Kim (2007) concluded that the negative effects from child maltreatment could be weakened or countervailed by the presence of familial protection, and accordingly, the negative effects from child neglect and maltreatment can be prevented by familial protection. Thus, this study included the factor of familial protection as a possible factor influencing adolescent aggression, depression, and delinquency related to experiences of being abused. The presence of the familial protection factor may mitigate the negative effects or protect children from such effects of child maltreatment. In this case, familial protection includes family structure composition. The traditional Korean family structure which typically includes grandparents has transitioned to a generalized nuclear family structure consistent with a more Westernized idea of family structure (Kim, 2006). The increase in child abuse and neglect may be influenced by changes within the family structure. Therefore, this study aimed to examine factors related to the relationship between child abuse and neglect, and adolescent aggressiveness, depression, and delinquency together with the role of grandparents as a protective factor against the negative effects of abuse. This study perceived familial protection, in the form of the presence of grandparents for Korean families, to be a potential barrier against the negative effects of child maltreatment and thus advance the understanding of sources of support in determining methods of intervention and prevention for decreasing child abuse and neglect. Types of Abusive Parenting and Delinquency As previous discussed, the relationship between parental abuse and the negative outcomes for adolescent growth and development contains a multitude of factors that range from type of abuse to family structure and to varying personal and social outcomes for the victims. It is important to consider any differences these factors may have in order to reveal the layers of physical, psychological, emotional, and social impact on adolescents’ well-being. The complexity and multi-level perspective of parental abuse and its relationship to adolescent delinquency behaviors underscores the necessity for examining the distinct characteristics of abuse types and possible differential effects on negative behaviors or outcomes. Furthermore, different types of abuse may affect different types of delinquency behaviors. For instances, reports have indicated that parental abuse was related to adolescents’ delinquent behaviors such as engaging in aggressive fights and sexual assaults, while negligent parenting was linked to adolescents’ adjustment problems such as truancy (National Child Protect Agency). Specifically, parenting with physical abuse was found to be related with adolescents’ serious crime acts (Widom, 1995, Widom, 1996 and Widom and Maxfield, 2001). However, Evans and his colleagues (2012) reported that emotional abuse by parents had larger impacts on both violent and non-violent youth delinquency, as compared to physical abuse. Therefore, to understand child maltreatment and delinquency, the subtypes of parental abuse needs to be investigated. In addition, following Broidy et al. (2003), the current study divides the subtypes of delinquent behaviors into violent and non-violent behaviors.

Results The correlations, mean, standard deviation, skewness, and kurtosis of the variables in the study are provided in Table 1. Significant correlations were found among the study variables except for the relationship between physical neglect parenting and violent delinquency behavior. According to the guidelines of severe non-normality (i.e., skewness > 3; kurtosis > 10) proposed by Curran, West, and Finch (1996), the normality assumption of all the variables was well met, with skewness values of <3 and kurtosis values of <10. Table 1. Descriptive statistics for study variables and correlation. Emotional neglect Physical neglect Emotional abuse Physical abuse Aggression Depression Non-violent Violent Physical neglect 0.57*** Emotional abuse 0.16*** 0.20*** Physical abuse 0.18*** 0.23*** 0.56*** Aggression 0.12*** 0.13*** 0.17*** 0.21*** Depression 0.15*** 0.15*** 0.15*** 0.15*** 0.44*** Non-violent 0.06* 0.06* 0.05** 0.07** 0.10** 0.09** Violent 0.07** .02 0.06** 0.05* 0.12*** 0.05* 0.47*** M 1.66 1.50 1.62 1.50 2.08 1.70 1.12 1.13 SD 0.75 0.63 0.81 0.77 0.64 0.61 0.43 0.61 Skewness 0.91 1.15 1.20 1.46 0.27 0.75 2.76 2.94 Kurtosis 0.22 1.28 0.75 1.39 0.11 0.25 9.50 9.98 * p < .05. ** p < .01. *** p < .001. Table options Testing the Mediational Models We used 2 mediational models to test the hypothesis that the relationship between abusive parenting and delinquency behaviors is mediated by students’ psycho-social assets. The initial structural model reflecting partial mediation was specified with both direct and indirect paths from abusive parenting factors to delinquency behaviors via 2 mediators. The second structural model represented the full mediational model, which did not include direct effects of abusive parenting factors on delinquency behaviors. Results indicated that both models showed a good fit for the sample; the partial mediational model yielded an overall χ2(169) value of 2,527.37, with CFI = .969, NNFI = .966, and RMSEA = .057, and the full mediational model yielded an overall χ2(177) value of 2,551.63, with CFI = .969, NNFI = .974, and RMSEA = .055. A χ2 difference test was conducted to decide the better fitting model to the data. The χ2 difference and difference in degrees of freedom between the full mediation model and the partial mediation model determined the model selection. The χ2 difference value was statistically significant at the .05 probability level: Δχ2 (8) = 14.26. A χ2 difference test supported the full mediational model. Thus, we chose the full mediational model as the final theoretical model. The fit of the final model was considered acceptable in terms of the 3 fit indices. We tested for significant direct and indirect effects using Preacher and Haye's (2008) nonparametric bootstrapping approach. The standardized parameter estimates for this model are presented in Table 2. Table 2. Results for direct and indirect effect in the final model. Unstandardized coefficient Standard error Standardized coefficient 95% CI (bootstrap with bias correction) Direct effects EN → A .041 .023 .053 (−.002, .108) EN → D .070* .020 .097 (.042, .164) PN → A .048 .027 .053 (−.002, .109) PN → D .060* .024 .070 (.021, .129) EA → A .064* .022 .084 (.019, .151) EA → D .069* .020 .098 (.031, .152) PA → A .114** .021 .158 (.108, .224) PA → D .044* .019 .066 (.017, .135) A → NV .066* .018 .090 (.025, .147) A → V .138** .026 .133 (.067, .200) D → NV .050* .018 .064 (.004, .124) D → V .008 .026 .007 (−.046, .072) Indirect effects EN → A → NV .003 .002 .006 (.000, .006) EN → A → V .004* .003 .008 (.001, .009) PN → A → NV .003* .002 .006 (.001, .008) PN → A → V .005 .003 .008 (.000, .010) EA → A → NV .004** .002 .009 (.002, .008) EA → A → V .007* .003 .013 (.002, .014) PA → A → NV .008** .002 .017 (.004, .020) PA → A → V .012** .003 .023 (.009, .025) EN → D → NV .005* .002 .010 (.004, .017) EN → D → V .003* .002 .006 (.001, .008) PN → D → NV .004** .002 .007 (.003, .015) PN → D → V .003* .002 .004 (.001, .006) EA → D → NV .005** .002 .010 (.004, .016) EA → D → V .003* .002 .006 (.001, .007) PA → D → NV .003** .002 .007 (.003, .018) PA → D → V .002* .001 .004 (.001, .008) EN: emotional neglect, PN: physical neglect, EA: emotional abuse, PA: physical abuse, A: aggression, D: depression, NV: non-violent, V: violent. * p < .05. ** p < .01. ***p < .001. Table options Table 2 shows that negligent parenting did not exert significant effects on aggression but exerted significant effects on depression (emotional β = .097, p < .05; physical β = .07, p < .05). Abusive parenting had significant effects on depression (emotional β = .098, p < .05; physical β = .066, p < .05) and aggression (emotional β = .084, p < .05; physical β = .158, p < .01). Aggression had significant effects on non-violent and violent delinquency behaviors (non-violent β = .090, p < .05; violent β = .133, p < .01), while depression did not exert a significant effect on violent delinquency behavior but exerted a significant effect on non-violent delinquency behavior (β = .064, p < .05). In addition to these direct effects, the bootstrap results indicated that parental abuse and neglect had significant indirect effects on delinquency behaviors via 2 mediators. Specifically, there were indirect effects of negligent parenting on delinquency behaviors via aggression (emotional neglect to violent β = .008, p < .05; physical neglect to non-violent β = .006, p < .05). There were indirect effects of negligent parenting on violent (emotional β = .006, p < .05; physical β = .004, p < .05) and non-violent delinquency behaviors (emotional β = .010, p < .05; physical β = .007, p < .01) via depression. There were indirect effects of abusive parenting on non-violent delinquency behavior via aggression (emotional β = .009, p < .01; physical β = .017, p < .01) and depression (emotional β = .010, p < .01; physical β = .007, p < .01), as well as indirect effects of abusive parenting on violent delinquency behavior via aggression (emotional β = .013, p < .05; physical β = .023, p < .01) and depression (emotional β = .006, p < .05; physical β = .004, p < .05). Comparisons Between the Two Groups Table 3 provides the standardized parameter estimates in the model for children from traditional families and nuclear families. Results showed that all the paths were non-significant for children living with their grandparents. However, for children from nuclear families, most paths were statistically significant. Specifically, emotionally negligent parenting exerted a significant effect on depression (β = .113, p < .01). Abusive parenting had a significant effect on depression (emotional β = .107, p < .01; physical β = .060, p < .05) and aggression (emotional β = .071, p < .05; physical β = .166, p < .01). Aggression had significant effects on both non-violent and violent delinquency behaviors (non-violent β = .104, p < .01; violent β = .150, p < .01). Parental abuse and neglect had significant indirect effects on non-violent delinquency behaviors via 2 mediators. Specifically, there were indirect effects of physical abusive parenting on delinquency behaviors via aggression (non-violent β = .019, p < .01; violent β = .023, p < .01) and depression (non-violent β = .006, p < .05). There were indirect effects of emotional abusive parenting on delinquency behavior via aggression (non-violent β = .010, p < .01; violent β = .006, p < .05). There were indirect effects of emotionally negligent parenting on non-violent (β = .011, p < .01) and violent delinquency behaviors (β = .006, p < .05) via depression. Table 3. Results for direct and indirect effect in the final model across two groups. Children from traditional family (N = 168) Nuclear family (N = 1,781) Unstandardized coefficient Standard error Standardized coefficient 95% CI (bootstrap with bias correction) Unstandardized coefficient Standard error Standardized coefficient 95% CI (bootstrap with bias correction) Direct effects EN → A .025 .066 .035 (−.205, 225) .041 .024 .052 (−.101, .118) EN → D −.037 .061 −.053 (−.240, 101) .082** .022 .113 (.065, .179) PN → A .098 .080 .117 (−.119, 324) .040 .029 .044 (−.018, .131) PN → D .124 .075 .167 (−.053, .418) .047 .026 .055 (−.012, .113) EA → A .173 .072 .226 (−.073, .401) .054* .023 .071 (.003, .127) EA → D .039 .065 .053 (−.236, .288) .076** .021 .107 (.033, .166) PA → A .078 .072 .100 (−.144, .270) .118** .022 .166 (.120, .242) PA → D .070 .067 .094 (−.130, .287) .040* .020 .060 (.011, .135) A → NV −.020 .088 −.020 (−.180, .182) .072** .018 .104 (.028, .160) A → V .046 .140 .028 (−.096, 260) .146** .025 .150 (.079, 228) D → NV .140 .087 .131 (−.070, 392) .040 .018 .053 (−.004, .135) D →　 V .120 .136 .071 (−.047, 296) −.006 .025 −.005 (−.084, .061) Indirect effects EN → A → NV .001 .009 .001 (−.009, .034) .003 .002 .006 (−.001, .018) EN → A → V .002 .012 .002 (−.010, .041) .005 .004 .007 (−.001, .021) PN → A → NV .003 .011 .003 (−.008, .062) .003 .003 .005 (−.002, .017) PN → A → V .009 .016 .007 (−.005, .000) .005 .004 .006 (−.002, .022) EA → A → NV .005 .013 .007 (−.019, .062) .004 .002 .008 (−.002, .016) EA → A → V .017 .023 .013 (−.008, .094) .007 .004 .010 (−.002, .020) PA → A → NV .002 .019 .003 (−.008, .057) .010** .003 .019 (.011, .035) PA → A → V .007 .015 .006 (−.004, .066) .016** .005 .023 (.012, .043) EN → D → NV −.005 .013 −.006 (−.001, .007) .006** .002 .011 (.005, .025) EN → D → V −.005 .011 −.004 (−.074, .005) .005* .003 .006 (.001, .016) PN → D → NV .017 .024 .020 (−.007, .037) .003 .002 .005 (−.001, .014) PN → D → V .017 .022 .013 (−.006, .020) .003 .002 .003 (−.001, .008) EA → D → NV .045 .015 .011 (−.011, .077) .005** .003 .010 (.004, .025) EA → D → V .005 .015 .004 (−.010, .044) .004* .003 .006 (.000, .015) PA → D → NV .083 .013 .006 (−.005, .000) .003* .002 .006 (.002, .019) PA → D → V .009 .015 .007 (−.005, .001) .002 .002 .003 (−.001, .011) EN: emotional neglect, PN: physical neglect, EA: emotional abuse, PA: physical abuse, A: aggression, D: depression, NV: non-violent, V: violent. * p < .05. ** p < .01. ***p < .001.