Child abuse, self-control, and delinquency: A general strain perspective

Journal of Criminal Justice xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Journal of Criminal Justice journal homepage: www.elsevier.com/locate/jcrimjus

Child abuse, self-control, and delinquency: A general strain perspective Jackson M. Buncha,1, Amaia Iratzoquib, Stephen J. Wattsc,⁎ a b c

Department of Sociology, University of Montana, Department of Sociology, Social Science 333, 32 Campus Dr., Missoula, MT 59812, United States Department of Criminology and Criminal Justice, University of Memphis, 311 McCord Hall, Memphis, TN 38152, United States Department of Criminology and Criminal Justice, University of Memphis, 311 McCord Hall, Memphis, TN 38152-3530, United States

A R T I C L E I N F O

A B S T R A C T

Keywords: Child abuse Delinquency Self-control General strain theory Add health

Purpose: This study examined the independent effects of child abuse on self-control and delinquency and explored whether self-control mediates the child abuse-delinquency relationship. Methods: We employed path modeling in Mplus to examine the relationship between child abuse, self-control, and delinquency using data from the National Longitudinal Study of Adolescent to Adult Health (Add Health). Results: Net of theoretically important controls, child abuse has direct and indirect effects on delinquency, and the relationship is partially explained by low self-control. Conclusions: Child abuse has an independent influence on levels of self-control, supporting a proposition made by general strain theory, and self-control partially mediates the oft-observed relationship between child abuse and delinquency.

1. Introduction

child abuse, self-control, and delinquency. First, we examine GST and its predictions regarding the relationship between victimization and delinquency, paying particular attention to the research on the link between child abuse and delinquency. Then, we discuss the role selfcontrol plays in GST. We focus on the GST claim that the strain of victimization—such as child abuse—can reduce self-control, discussing the research examining the influence of victimization on self-control, the impact of childhood trauma on brain development, and the association between negative parenting and self-control.

Research consistently demonstrates that low self-control is an important cause of both offending (see Pratt & Cullen, 2000) and victimization (see Pratt, Turanovic, Fox, & Wright, 2014). In contrast, few studies have examined the impact of victimization on self-control, and the limited results are mixed (Agnew et al., 2011; Monahan, King, Shulman, Cauffman, & Chassin, 2015; Sullivan, Farrell, Kliewer, VulinReynolds, & Valois, 2007). Agnew's (1992, 2006) general strain theory (GST) predicts, in part, that low self-control mediates the relationship between victimization and delinquency. GST highlights child abuse, in particular, as a severe form of strain likely to lead to delinquency (Agnew, 2001, 2013). Despite this fact, no research has examined the extent to which these GST predictions regarding the nature of the relationship between victimization, self-control, and delinquency are observed within the context of child abuse victimization. Here, we use a GST framework to examine how child abuse impacts self-control and whether self-control mediates the relationship between child abuse and adolescent delinquency. We test these predictions using data from the National Longitudinal Study of Adolescent to Adult Health (Add Health). We begin with a review of GST and the relevant literature. 2. Theoretical framework and prior research In this section, we discuss GST (Agnew, 1992, 2006) as it relates to

⁎

1

2.1. General strain theory and child abuse GST explains why certain people engage in delinquent and criminal behavior. Agnew claimed that individuals commit crime as a way to relieve psychological strain (Agnew, 1992, 2006). People experience three broad categories of strains: (1) losing something valued, (2) being exposed to something unpleasant, and (3) failing to achieve valued goals. Strain causes an individual to experience negative emotions, such as anger, depression, or fear. According to the theory, if people experience strain and lack alternative coping mechanisms, then they may resort to crime in order to relieve their negative emotions. Agnew (2001) asserted that certain types of strain are more likely to lead to criminal behavior. In particular, he stated that the most criminogenic strains are those that are perceived as unjust, strains that are high in magnitude, strains that are caused by or associated with low social

Corresponding author. E-mail addresses: [email protected] (J.M. Bunch), [email protected] (A. Iratzoqui), [email protected] (S.J. Watts). The first two authors contributed equally to this manuscript.

http://dx.doi.org/10.1016/j.jcrimjus.2017.09.009 Received 5 June 2017; Received in revised form 29 September 2017; Accepted 30 September 2017 0047-2352/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Bunch, J.M., Journal of Criminal Justice (2017), http://dx.doi.org/10.1016/j.jcrimjus.2017.09.009

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

rather than children), they also examined the link between maltreatment by parents and delinquency. Using cross-sectional data from the NSC, the researchers found that adolescent maltreatment—which included measures of physical punishment, parental withdrawal of love, emotional abuse, and absence of parental support—was associated with general delinquency, serious violent delinquency, and substance use. The study also suggested that some of the effect of maltreatment on delinquency was mediated by negative emotions, such as anger, anxiety, and depression. Finally, another study of adolescent maltreatment lends further support to GST (Brezina, 1998). Brezina utilized two waves of the Youth in Transition survey, a national study of male public high school students. He found that adolescent maltreatment by parents (physical and verbal abuse) was positively associated with delinquency, and these effects worked through a variety of mechanisms, including anger. Taken together, these studies lend support to the GST prediction that child abuse leads to delinquency.

control, and strains that create a pressure/incentive to engage in criminal coping behaviors. Agnew (2001, 2002) has argued that victimization is one of the most consequential forms of strain. In fact, Agnew (2001, 2013) discussed a specific form of victimization—child abuse—as a particularly criminogenic form of strain. Victimization possesses the characteristics of strains most likely to lead to crime (Agnew, 2001). Victims are likely to perceive themselves as having been treated unfairly, viewing the perpetrator's actions as unjustified. Victimization—particularly violent victimization—is generally very high in magnitude: it is an incredibly noxious experience likely to produce severe strain. Agnew (2001) also pointed out that victimization is more likely to take place in situations characterized by low levels of social control. Finally, victimization can create pressure to engage in criminal coping, perhaps by providing a justification for delinquent behavior or because it is associated with the social learning of crime (Agnew, 2001). Numerous studies examining GST demonstrate that victimization is associated with various forms of offending (e.g., Agnew, 2002; Hay & Evans, 2006; Iratzoqui, 2015; Lin, Cochran, & Mieczkowski, 2011; Neff & Waite, 2007; Watts & McNulty, 2013). Using panel data from the National Survey of Children (NSC), Hay and Evans (2006) found that violent victimization was positively associated with future substance use and violent/property offending. Agnew (2002) looked at the relationship between victimization and offending using a nationally representative sample of male high school students. He found that, controlling for prior delinquency, violent victimization was associated with higher levels of delinquency. Lin et al. (2011) examined nationally representative cross-sectional data from the National Survey of Adolescents and found that violent victimization was associated with both violent/property crime and drug use. In addition to direct victimization, research suggests that indirect victimization (e.g., the victimization of friends or family) is also associated with offending (Agnew, 2002; Lin et al., 2011). Research examining the relationship between child abuse and delinquency reflects the broader findings of victimization and offending, providing support for GST arguments related to child abuse: Child abuse is associated with a wide range of future delinquent and criminal behaviors, including violent crime and aggression (Chapple, 2003; Herrenkohl, Huang, Tajima, & Whitney, 2003; Salzinger, Rosario, & Feldman, 2007), substance use (Bergen, Martin, Richardson, Allison, & Roeger, 2004; Brems, Johnson, Neal, & Freemon, 2004; Gutierres & Van Puymbroeck, 2006; Ompad et al., 2005), intimate partner violence (Godbout et al., 2017; Gómez, 2011), sexual offending (DeLisi, Kosloski, Vaughn, Caudill, & Trulson, 2014), and general antisocial behavior (Dembo, Schmeidler, & Childs, 2007; Gao, Wong, & Yu, 2016; Klika, Herrenkohl, & Lee, 2013; Smith, Park, Ireland, Elwyn, & Thornberry, 2013; Teague, Mazerolle, Legosz, & Sanderson, 2008; Watts, 2016). A meta-analysis examining the negative impacts of non-sexual child maltreatment by Norman et al. (2012) found that across over 100 studies, non-sexual child maltreatment was consistently associated with future drug abuse, suicide attempts, and risky sexual behavior. Several studies have examined child abuse from a GST perspective (e.g., Brezina, 1998; Hollist, Hughes, & Schaible, 2009; Iratzoqui, 2015; Watts & McNulty, 2013). Using Add Health data, Iratzoqui (2015) found that victims of child abuse and neglect were more likely to engage in delinquent coping strategies during adolescence, including substance abuse and selling drugs. Watts and McNulty (2013) also tested GST predictions on the impact of child abuse using the Add Health survey. The authors found that child abuse—including physical and sexual abuse by parents or caregivers—was positively associated with adolescent delinquency (an index of criminal behavior including violent crimes, property crimes, and drug selling). Watts and McNulty (2013) also found that depression partially mediated the effects of physical abuse on offending, providing some support for another GST prediction. Though Hollist et al. (2009) utilized a slightly older sample (adolescents

2.2. General strain theory and self-control Though self-control is a concept primarily associated with Gottfredson and Hirschi's (1990) self-control theory, self-control plays an important role in GST, as well. GST posits two main claims regarding self-control. First, GST suggests that self-control influences how people react to strain. Research generally supports this claim: self-control appears to moderate the effects of strain—including victimization—on offending (e.g., Agnew, Brezina, Wright, & Cullen, 2002; Hay & Evans, 2006; Turanovic & Pratt, 2013). Second, GST suggests that repeated exposure to strain can reduce self-control, thus leading to delinquent behavior (Agnew, 2006; Agnew et al., 2011). This argument implies that negative emotions are not the only mechanisms through which strain leads to delinquency; rather, reduced self-control also mediates some of the influence of strain on criminal behavior. There is little research examining this proposition, though a few studies have examined how victimization impacts selfcontrol (Agnew et al., 2011; Monahan et al., 2015; Sullivan et al., 2007). Overall, the results are mixed. Agnew et al. (2011) found some support for the GST claims regarding victimization and self-control. The authors utilized several waves of the GREAT data to examine whether victimization reduced self-control. They found evidence that victimization within the past year was associated with decreased self-control among adolescents, but victimization was unrelated to self-control in subsequent waves. This suggests that any influence of victimization on self-control may be fairly contemporaneous. Sullivan et al. (2007) also found no support for victimization reducing future self-control. The researchers examined the impact of violence on self-control and subsequent delinquency within a sample of rural sixth graders. They found no relationship between victimization reported at the beginning of the year and self-control measured at the end of the school year. Similarly, indirect victimization (i.e., witnessing violence) was unrelated to subsequent self-control. More recently, Monahan et al. (2015) examined the impact of victimization on both impulse control and future orientation among a sample of juvenile offenders. The researchers found that victimization was associated with a slower growth in future orientation during adolescence and early adulthood. They also found that an increase in victimization during adolescence was associated with a reduction in impulse control the following year, but this pattern did not hold in young adulthood. The potential contemporaneous impact of victimization on self-control suggested by some of these studies is not necessarily inconsistent with GST predictions (Agnew, 2006; Agnew et al., 2011). Much of the GST discussion on this topic focuses on how strain impacts self-control in the near-term, with the deleterious effects assumed to diminish over time as the negative emotions abate. However, GST also argues that chronic victimization—such as child abuse—could result in a more enduring decrease in self-control. Taken together, these studies provide mixed evidence as to whether victimization influences self-control, particularly in the long term. 2

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

utilizing a large, nationally representative sample to test GST hypotheses concerning whether self-control mediates the relationship between child abuse and delinquency.

Additionally, this literature does not address how child abuse in particular impacts future self-control. Research examining the impacts of childhood trauma suggests mechanisms through which child abuse may have an enduring impact on low self-control. Childhood exposure to violence, such as child abuse (Glaser, 2000; Heim & Nemeroff, 2001; Schore, 2001; Teicher et al., 2003) or bullying victimization (Vaillancourt, Hymel, & McDougall, 2013) can affect the development of a child's brain. The stress response is associated with changes to the structure and function of the brain, altering neuroendocrine systems (Glaser, 2000; Golier & Yehuda, 1998; Teicher et al., 2003). These types of neurobiological systems are related to a variety of negative behavioral and mental health outcomes, including low self-control (Davidson, Putnam, & Larson, 2000; Heim & Nemeroff, 2001; Schore, 2001; van der Kolk & Fisler, 1994). On a closely related topic, research has also examined the impact of harsh forms of parenting—but not child abuse, per se—on self-control and subsequent delinquency (Hay, 2001; Meldrum, Verhoeven, Junger, van Aken, & Deković, 2016; Simons, Simons, Chen, Brody, & Lin, 2007). At its most severe end, the construct of “harsh parenting” used in research can overlaps somewhat with child abuse, but it is generally viewed as distinct from child abuse and maltreatment (see Simons, Whitbeck, Conger, & Wu, 1991). Therefore, though this research does not directly address the focus of the current study, it does inform the broader discussion of negative parenting practices, self-control, and delinquency. Overall, this body of research suggests that harsh parenting is associated with low self-control, which is in turn associated with delinquency. This implies that child abuse could have a similar impact, decreasing self-control of the victim and increasing future involvement in crime; however, this particular prediction of GST remains an open question.

4. Methods 4.1. Data The National Longitudinal Study of Adolescent to Adult Health (Add Health) is a longitudinal study of a nationally representative schoolbased sample of youth who were in grades 7–12 in the United States beginning in 1994–95. Add Health employed cluster and systematic sampling methods to select 80 public and private high schools and 52 middle schools to be representative of U.S. regions, urban composition, school size and type, and ethnicity (Harris et al., 2003). Add Health collected data first via an in-school survey, then, for a select subsample, followed up with a series of in-home interviews conducted in 1994–95 (Wave I; n = 20,745), 1996 (Wave II; n = 14,738), 2001–02 (Wave III; n = 15,170), and 2007–08 (Wave IV; n = 15,701).2 Our current study utilizes data from the restricted data set, which is identical to the public-use data set but contains the full sample of respondents. We use data only from respondents who completed the Wave I in-school survey and the in-home surveys for Waves I, II, and III, and for whom sampling weights were available (N = 11,621).3 An important point to note is that Add Health is limited in its generalizability, in part because it is a sample of American youth drawn from schools, born in a particular time period, and may not be representative of all American youth across time. Additionally, Add Health, like any longitudinal survey, experienced significant attrition from sample respondents. To attempt to handle attrition, the path model was run in Mplus, which relies upon a direct maximum likelihood (ML) estimator to handle missing data on individual items and retain the full sample in the model. Direct ML is comparable with multiple imputation. The Mplus software automatically imputes values for these cases unless listwise deletion is specified. Direct ML averages out over all the data and generates likelihoods for each individual given the data available for that individual. As a result, the models retained individuals with missing data on individual items, avoiding the bias associated with listwise deletion, and improving the ability to generate unique parameter estimates and standard errors.4 Table 1 provides a description of the sample on each of the relevant variables.

3. Current study In order to address gaps in the empirical literature, our study examines the mediating role self-control plays in the relationship between childhood experiences of abuse and involvement in delinquent behavior in adolescence. We use path modeling in Mplus Version 7 (Muthén & Muthén, 2006) with nationally representative, longitudinal data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) to test several theoretical predictions derived from GST regarding the nature of the relationship between child abuse, selfcontrol, and delinquency. Specifically, we predict that:

4.2. Measures

Hypothesis 1. Child abuse is associated with increased involvement in adolescent delinquency.

Hypothesis 4. Low self-control partially mediates the relationship between child abuse and adolescent delinquency.

4.2.1. Dependent variable: delinquency Delinquency was constructed from twelve items capturing violent and nonviolent forms of delinquent behavior in the twelve months prior to Wave II. Respondents were asked the frequency of their participation in twelve behaviors: (1) graffiti on another's property or public space, (2) damaging property that did not belong to the respondent, (3) taking something from a store without paying, (4) driving a car without the owner's permission, (5) stealing something worth more than $50, (6)

This study makes several unique contributions over previous research. First, prior research examining the impact of victimization on self-control relied on samples that are not nationally representative (Agnew et al., 2011; Monahan et al., 2015; Sullivan et al., 2007). Second, of these studies, only Sullivan et al. (2007) also examined whether self-control mediated the relationship between victimization and delinquency. Third, these studies examined the impact of victimization in general on self-control among adolescents. While prior research has examined the effect of various forms of harsh parenting on self-control and delinquency (Hay, 2001; Meldrum et al., 2016; Simons et al., 2007), no research to date has focused specifically on the impact of child abuse victimization on self-control and delinquency. The current study overcomes several limitations of previous research by

2 For more information on the Add Health data collection, see http://www.cpc.unc. edu/projects/addhealth/documentation/guides. 3 The child abuse questions were drawn from a series of questions asked retrospectively about childhood experiences during the Wave III survey. 4 A number of variables were recoded before the data file was converted to Mplus. The child abuse, delinquency, and self-control measures were created in Stata (StataCorp, 2013), a program which has a default of listwise deletion. In other words, the program automatically drops cases if a case is missing values on any one variable. In an effort to avoid this issue, each index was coded such that Stata only dropped cases missing on all indicators within a given index. For example, if a case scored 1 on an indicator, 0 on a second indicator, and was missing data on the remaining indicators, the case would score a value of 1 on the index. Missing values were not imputed. Once in Mplus, the direct ML estimator applied in cases missing values across indices, as well as the other variables of interest. Of note, there was very little missing data for the three key study variables.

Hypothesis 2. Child abuse is associated with lower levels of selfcontrol. Hypothesis 3. Low self-control is involvement in adolescent delinquency.

associated

with

increased

3

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

Table 1 Descriptives (N = 11,621). Variable

Description

Mean

SD

Min

Max

Delinquency Child abuse Self-control Maternal supervision Maternal warmth Perceived early fatality Deviant peers Binge drinking Receiving public assistance Prior violent victimization Prior delinquency Age Female White

12-item measure of a sum of standardized general delinquency items from Wave II Dichotomous item coded 1 for any physical or sexual abuse 21-item summed measure of standardized items capturing elements of low self-control 3-item index of maternal presence in the home 5-item index on relationship with mother Single item asking respondents their chances of being killed before age 21 3-item index of peer self-reported deviance Single item on how often respondents reported drinking 5 or more drinks in a row Dichotomous item coded 1 if family reported receiving welfare Dichotomous item coded 1 if respondent reported violent victimization at Wave I 12-item summed measure of standardized general delinquency items from Wave I Age at Wave I Dichotmous item coded 1 for Female Dichotmous item coded 1 for White

− 0.026 0.287 − 0.021 5.83 7.84 0.642 1.26 0.569 0.089 0.193 − 0.0659 15.28 0.528 0.635

6.83 0.452 8.027 2.39 3.12 0.805 1.17 1.23 0.285 0.395 7.21 1.61 0.499 0.481

−3.13 0 −28.21 1 2 0 0 0 0 0 −4.19 11 0 0

84.24 1 38.71 15 25 4 3 6 1 1 67.14 21 1 1

Add Health data. However, Pratt and Cullen's (2000) meta-analysis revealed that the association between low self-control and antisocial outcomes is not contingent on utilizing the Grasmick et al. (1993) scale, and the low self-control index utilized in the current study has been shown to have predictive validity (Beaver et al., 2008; Beaver et al., 2009; Belsky & Beaver, 2011; Watts & McNulty, 2016).7

entering a house or building to steal something, (7) using or threatening to use a weapon to get something from someone, (8) selling marijuana or other drugs, (9) stealing something worth less than $50, (10) taking part in a group fight, (11) pulling a knife or gun on someone, and (12) shooting or stabbing someone. Responses on the first 10 items ranged from 0 (never) to 3 (5 or more times), while responses on the last 2 items ranged from 0 (never) to 2 (more than once). The items were standardized and then summed in a global delinquency measure at Wave II where higher scores indicate more delinquency (Cronbach's α = 0.81). 5

4.2.3. Independent variable: child abuse Child abuse was constructed from two questions at Wave III that retrospectively asked respondents to recall their experiences of physical and sexual abuse before starting the sixth grade. Respondents were asked how often a parent or other caregiver had (1) slapped, hit, or kicked the respondent, or (2) touched them in a sexual way or forced the respondent into sexual activity. Most respondents reported no experiences of child abuse, and the vast majority of respondents who reported experiencing child abuse were the victims of physical abuse only, with only a very small percentage of respondents reporting experiencing sexual abuse or both physical and sexual abuse.8 The final measure was dichotomized, with a 0 indicating no experiences of either physical or sexual abuse and a 1 indicated experiences of either or both of these behaviors at least once.

4.2.2. Mediating variable: low self-control The appropriate method for measuring self-control has been the source of much debate in the criminological literature (Beaver, DeLisi, Mears, & Stewart, 2009; Beaver, Wright, DeLisi, & Vaughn, 2008; Longshore, 1998; Longshore & Turner, 1998). In the current study, low self-control is measured as an index constructed from 21 items in the Wave I survey that related to the different components of self-control proposed by Gottfredson and Hirschi (1990). This index is a slightly altered version of one developed by Beaver et al. (2008, 2009) for use in the Add Health survey. The majority of questions asked respondents to rank themselves on their self-centeredness, decision-making style, and attention span. The parents of respondents were also asked four questions related to whether their child had a bad temper, was doing well, and the state of their relationship with the parent (see the Appendix A for a complete list of the items that make up the self-control measure). Responses for these items ranged from 1 (always) to 5 (never) and 1 (very well) to 4 (not well at all) on the parent items and 1 (strongly agree) to 5 (strongly disagree) on the self-reported items. The individual items were standardized and then summed into an index, with higher scores indicating overall lower levels of self-control (Cronbach's α = 0.74).6 While the Grasmick, Tittle, and Arneklev (1993) scale is most widely used to assess self-control, these items are not available in the

4.2.4. Control variables The analyses also included a battery of control variables to better isolate any potential associations found between child abuse, self-control, and delinquency, and to help avoid model misspecification. While numerous controls are included, the ones of key import include measures of parenting, prior delinquency, and violent victimizations (separate from child abuse). Including parenting is especially important, as parenting is the primary predictor of self-control according to Gottfredson and Hirschi (1990). Controlling for aspects of parenting means that the independent effect of child abuse on self-control can be measured. Maternal supervision was coded from three items at Wave I asking respondents how often their mother was at home (1) when they left for school, (2) when they returned from school, and (3) when they went to bed. Maternal warmth summed respondents' responses to five items at the Wave I survey asking (1) whether they felt close to their mother, (2) that she cared for them, (3) was warm and loving towards them, (4) that they communicated well, and (5) that they were satisfied

5 Some work has argued that violent and nonviolent delinquency are conceptually different behaviors, particularly in their associations with child abuse (see, e.g., Savage, Palmer, & Martin, 2014). In the current analysis, a universal construct was preferred, for several reasons. One is that antisocial behavior as a whole tends to be highly correlated, such that considering violent and nonviolent delinquency as separate independent behaviors would not reflect the true propensity for offending. Second, there is far less evidence of specialization in offending, particularly in behavior linked to low self-control (i.e., Gottfredson & Hirschi, 1990; see also Clinkinbeard, 2014; Perrone, Sullivan, Pratt, & Margaryan, 2004). 6 One issue with using a universal measure of self-control is that it fails to capture the six components of self-control as developed by Gottfredson and Hirschi's (1990) theoretical argument. A potential solution is to use a latent construct of self-control, but the model fit statistics ultimately revealed the additive index to be a better fitting construct for the model and data used in the analyses. Also, if any of the 21 items were removed from this additive index, Cronbach's alpha was reduced.

7 It is also worth noting that this low self-control index has high construct validity within this study. Low self-control significantly correlated with maternal warmth, maternal supervision, deviant peers, and other variables in the expected manner. 8 The number of respondents reporting any sexual abuse were very low (n = 524), which limited the power of the model to estimate this variable as an independent predictor of low self-control and delinquency. However, an alternate series of analyses were estimated examining physical and sexual abuse separately, and found smaller effect sizes (i.e., a smaller but still significant beta for physical abuse, no significant effect for sexual abuse) and lower model fit, so the combined construct was preferred.

4

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

with their relationship with their mother. For both maternal supervision and maternal warmth higher scores denote worse parenting. Controls were also included for prior victimization and prior delinquency. Prior victimization was a single dichotomous item that records whether a respondent experienced a violent victimization in the twelve months prior to Wave I (1 = yes). Prior delinquency was a twelve-item measure identical to the dependent variable capturing delinquency occurring in the twelve months prior to Wave I (Cronbach's α = 0.83). We include several other theoretically supported covariates in the model. Perceived early fatality captured whether respondents reported at Wave I if they thought they would be killed by age 21. Anticipated early death may play a role in whether youth will take the risk in committing deviant behavior, as well as whether they can focus on long-term goals, a component of self-control (Clinkinbeard, 2014; Haynie, Soller, & Williams, 2014). Binge drinking was identified from a single item asking respondents at Wave I how often they drank five or more alcoholic drinks in a row within the past twelve months. Deviant peers was a three-item scale that asked respondents' self-identified friends at the Wave I survey about their use of (1) cigarettes, (2) alcohol, and (3) marijuana, with a higher score denoting more peer deviancy. Receiving public assistance was identified from a single item asking parents whether they were currently receiving any form of public assistance, including welfare (1 = yes). Finally, demographic variables included the respondent's gender, race, and age reported at Wave I. Race was constructed as a single dichotomous variable, White (with non-White as the reference category).

Goodness-of-fit for structural equation models is generally based on multiple measures of model fit. Tankebe (2013) provided a review of the most commonly used measures, including the chi-square likelihood ratio statistic (χ2), the Comparative Fit Index (CFI), the Tucker Lewis Index (TLI), and the root mean square error of approximation (RMSEA). The goal of any statistical model is to be able to replicate the “true” population covariance matrix with the model estimated covariance matrix by isolating the outcome from the influences of all other factors, except those included in the model and a random disturbance term. Generally, good model fit provides a gauge for how well the model succeeds at matching the relationships as they exist in the true population (e.g., a nonsignificant chi-square, RMSEA value below 0.06, and CFI and TLI above 0.95) (Hu & Bentler, 1999). The model presented in the current study represents the best-fitting model to the data that is theoretically appropriate for the modeled relationships. Taken together, the fit statistics indicate the current model provides a good fit to the data (χ2 = 20.57, df = 5, p = 0.001; RMSEA = 0.019; CFI = 0.997; TLI = 0.987). While the χ2 is significant, this statistic is vulnerable to a number of factors, particularly sample size, the number of variables in the model, and the presence of a non-normal distribution, so a significant χ2 is not considered the main determinant of model fit (Bollen, 1989, pp. 269–70). Alternative statistics are not vulnerable to these issues, and are thus considered better determinants of model fit. All of these statistics point to excellent model fit. Table 2 reports the results for the full model. Mplus provides both unstandardized and standardized coefficients. Standardized coefficients are preferred, because they allow for the comparison of the effects of each measure to the overall model. Accordingly, we report standardized coefficients in Table 2. The findings in Table 2 generally support the theoretical relationships between child abuse, low self-control, and delinquency as proposed by GST. In the first path, experiences of child abuse are directly and significantly related to lower levels of self-control (β = 0.040, p ≤ 0.001). In the second path, low self-control is also positively and significantly related to delinquent behavior (β = 0.046, p ≤ 0.001). Child abuse also indirectly increases the risk of delinquent behavior because of lower levels of self-control. Lower self-control partially mediates the relationship between child abuse and delinquency (β = 0.002, p = 0.001). However, the direct relationship between child abuse and later involvement in delinquent behavior remains significant (β = 0.021, p < 0.05). These findings suggest that part of the reason that child abuse leads to adolescent offending is because it lessens self-control, and low self-control leads to delinquent behaviors. This supports the broad GST argument that a reduction in self-control is one of the ways in which strain leads to offending. Fig. 2 presents a graphical representation of only the paths relevant to the four hypotheses. The remaining significant coefficients for the theoretical controls found in Table 2 were in directions consistent with prior research. In the first path, the strongest predictors of low self-control are the maternal warmth (β = 0.333, p ≤ 0.001) and prior delinquency (β = 0.166, p ≤ 0.001) measures. These factors support both the early parental influence in the development of self-control as well as the stability of delinquent involvement, particularly at the peak of offending risk in adolescence. The only variable not statistically significant in either model is the indicator of the family receiving public assistance (β = 0.008, p = 0.414). In the second path, prior

4.3. Analytic strategy Regression methods typically require multiple models to be estimated in a precise order to “establish” a mediating or moderating effect (e.g., Baron & Kenny, 1986). However, even when all three variables are included in the regression model, the mediating variable is treated no differently from a covariate or control variable. The true set of relationships is not modeled, and the actual mediating effect (i.e., the effect of child abuse on delinquency, net of the effect of child abuse on low self-control) is not modeled. With the omission of this key relationship, two regression assumptions (e.g., independence of error terms and omitted variables) are violated, and effect estimates will be neither valid nor efficient (Freedman, 1987). Structural equation modeling methods—specifically path analysis—offer a unique contribution in being able to estimate direct, indirect, and total effects simultaneously within a single model, providing a better understanding of the underlying causal processes. Under this approach, each relationship can be estimated independently and then simultaneously within a single model, enabling for the examination of the change in effects within the inclusion of the additional paths. First, the individual relationships are modeled independently to illustrate these relationships, and then the full model is built to represent the span of the relationships from child abuse through the delinquency outcome. The hypotheses are supported if the causal inferences made by the theoretical model are consistent with the data used to model the relationships (Bollen, 1989). Specifically, in the final model, four relationships were simultaneously estimated, including the direct relationships between: (1) child abuse and low self-control, (2) low selfcontrol and delinquency, (3) and child abuse and delinquency and the mediated relationship between child abuse, low self-control, and delinquency.

(footnote continued) would be appropriate to include the same set of controls in both paths of the model, Mplus (i.e., path analysis) presents an advantage over traditional regression methods in this regard. The covariates were not included in both models because Mplus has difficulties with overspecifying models, i.e., when too many relationships are specified. However, Mplus will automatically specify covariances between each variable as part of estimating all models simultaneously, so these covariances (e.g., the association between child abuse and prior victimization) were de facto controlled in the final path model, but not reported.

5. Results Fig. 1 illustrates how the final empirical model was estimated, and the accompanying controls included within each path of the model.9 9

Several variables were included in only one path of the model. While traditionally it

5

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

Fig. 1. Empirical Model.

in the formation of self-control. By using a structural equation model, the model simultaneously accounts for all of the variable's influences at different points in time to be able to explain both low self-control and involvement in delinquency, which provides a more complete picture of the pathways between child abuse and delinquency. As a whole, the full model explains roughly 30% of the variance in the delinquency outcome (R2 = 0.304), while also explaining 25% of the variance in low self-control (R2 = 0.250).

Table 2 The direct and indirect effects of child abuse (N = 11,621).a

Predicting low self-control Child abuse Maternal supervision Maternal warmth Perceived early fatality Deviant peers Receiving public assistance Prior delinquency Age Female White R2 = 0.250 Predicting delinquency Child abuse Low self-control Maternal supervision Deviant peers Binge drinking Prior delinquency Prior victimization Age Female White Indirect child abuse effect R2 = 0.304 χ2 = 20.57, df = 5, p = 0.001 RMSEA = 0.019 CFI = 0.997 TLI = 0.987 a

β

SE

p

0.040 0.066 0.333 0.084 0.136 0.008 0.166 −0.039 0.020 0.108

0.009 0.009 0.009 0.009 0.010 0.009 0.010 0.009 0.010 0.009

0.000 0.000 0.000 0.000 0.000 0.414 0.000 0.000 0.031 0.000

0.021 0.046 0.024 0.061 0.023 0.473 0.038 −0.073 −0.048 −0.019 0.002

0.009 0.009 0.009 0.010 0.010 0.010 0.010 0.009 0.009 0.009 0.001

0.019 0.000 0.007 0.000 0.026 0.000 0.001 0.000 0.000 0.031 0.001

6. Discussion The current study examined the GST prediction that victimization—specifically child abuse—leads to delinquency, in part, because abuse reduces self-control, and low self-control increases delinquency. This claim was tested using nationally representative longitudinal data from Add Health. Overall, the results demonstrated the utility of GST in understanding the impacts of child abuse. Path analysis in Mplus indicated support for each of the GST predictions regarding the relationship between child abuse, self-control, and delinquency. Results showed that child abuse was positively associated with adolescent delinquency, providing support for our first hypothesis. This is consistent with the large body of research linking child maltreatment with future criminal offending (e.g., Brems et al., 2004; Brezina, 1998; Hollist et al., 2009; Iratzoqui, 2015; Klika et al., 2013; Teague et al., 2008; Watts & McNulty, 2013). Child abuse was also associated with lower self-control, supporting our second hypothesis. The claim that victimization leads to low self-control has received little attention in the literature, and the extant research provides mixed evidence for this assertion (Agnew et al., 2011; Monahan et al., 2015;

Standardized coefficients reported.

Fig. 2. Direct and indirect effects of child abuse.

Sullivan et al., 2007). We also found that low self-control is associated with higher levels of adolescent delinquency, supporting hypothesis three. This was expected, given the consistent findings linking selfcontrol and criminal behavior (see Pratt & Cullen, 2000). Finally, the relationship between child abuse and future delinquency was partially mediated by self-control: the indirect effect of abuse on delinquency via

delinquency is the strongest predictor of delinquent behavior (β = 0.473, p ≤ 0.001), with deviant peers (β = 0.061, p ≤ 0.001) and age (β = −0.073, p ≤ 0.001) also strongly affecting the delinquency outcome. It should additionally be noted that the measure of maternal presence in the home—the maternal supervision variable—plays less of a role in predicting actual involvement in delinquency than 6

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

effects of maltreatment/harsh parenting on antisocial outcomes can be moderated by genotype (e.g., Caspi et al., 2002; Kim-Cohen et al., 2006), future research should attempt to account for genetics in studies observing the relationship between child abuse, self-control, and antisocial outcomes. The current study raises a number of questions that future research should examine. Previous research suggests that the impact of child abuse on delinquency can differ for males and females (e.g., Bergen et al., 2004; Brems et al., 2004; Jang, 2007; Watts & McNulty, 2013). Does the relationship observed differ based on gender? Additionally, does the relationship found in the current analysis apply to victimization in later adolescence or adulthood? Does the type of victimization or the victim's relationship to the offender affect the observed relationships, with some forms of victimization having a more pronounced deleterious impact on self-control and future criminal behavior? Does chronic victimization have a different impact from a more isolated incident of victimization? We found that child abuse was associated with low self-control, leading to adolescent delinquency. Does this low self-control last beyond adolescence, or do victims of child abuse recover from this deficit? Research suggests that self-control continues to shift throughout adolescence and early adulthood (Burt et al., 2006, 2014; Hay & Forrest, 2006; Piquero, Jennings, & Farrington, 2010; Turner & Piquero, 2002; Winfree, Taylor, He, & Esbensen, 2006). How enduring is the impact of child abuse on self-control? Does low self-control partially explain the link between child abuse and adult criminality? Finally, GST suggests that decreased self-control is one of the ways in which victimization leads to delinquency. Of course, the other mediating factors highlighted by GST are negative emotions. Some research suggests the validity of this claim, with anger (Hay & Evans, 2006; Mazerolle, Piquero, & Capowich, 2003; Robertson, Stein, & Schaefer-Rohleder, 2010), depression (Carson, Sullivan, Cochran, & Lersch, 2009; Jang & Johnson, 2003; Manasse & Ganem, 2009; Watts & McNulty, 2013), and fear/anxiety (Hollist et al., 2009; Iratzoqui, 2015) mediating at least part of the relationship between victimization and delinquency. Future research should examine these two GST claims together, examining how child abuse impacts both selfcontrol and negative emotions, and how these, in turn, affect criminal behavior.

low self-control was statistically significant. This provided support for our fourth hypothesis. To our knowledge, this is the first evidence supporting the important but understudied GST proposition that part of the reason victimization leads to delinquent behavior is due to low selfcontrol. Our research has important implications for both policy and theory. The evidence suggests that part of the reason why victims of child abuse engage in adolescent delinquency is that abuse leads to low self-control. Knowledge of the mechanisms connecting child abuse to future criminal behavior may be used to help victims. These victims could be targeted with interventions designed to increase their self-control, thereby limiting the heightened risk of delinquency associated with abuse. Our findings also provide support for a relatively untested claim of GST. This study is the first evidence that low self-control partially mediates the influence of victimization on offending, adding nuance to our understanding of the well-documented relationship between victimization and offending (e.g., Averdijk, Van Gelder, Eisner, & Ribeaud, 2016; Lauritsen, Sampson, & Laub, 1991; Papachristos, Braga, Piza, & Grossman, 2015; Shaffer & Ruback, 2002). Our research highlights an understudied mechanism connecting victimization with criminal behavior and provides support for a GST understanding of the victim-offender overlap. The current study also demonstrates the applicability of this GST explanation to child abuse victimization and delinquency. Finally, though our study was not focused on examining Gottfredson and Hirschi's (1990) self-control theory (SCT), these results lend support to the predictions of their theory, as well. SCT argues that self-control is the result of socialization: effective parenting during childhood instills an individual with self-control. Child abuse is an extreme form of ineffective parenting as defined in SCT; therefore, it follows that abuse by parents or caregivers would not lead a child to develop self-control. This lack of self-control would ultimately lead to delinquency. Though we viewed our findings from a GST perspective, these results are not inconsistent with SCT. 6.1. Limitations and future research No research is without limitations. Our study has limitations related to data and measurement, specifically the measures of child abuse and self-control. The measure of child abuse we utilized lacks specificity: it does not indicate either the time period wherein events occurred or how long the abuse lasted. Additionally, the exact perpetrator of the abuse was not identified. The age at onset of abuse, the duration, and relationship to abuser are all potentially important factors that could shape how impactful child abuse is on future behavior and functioning. Another limitation is that the abuse items were retrospective in nature. The potential error introduced in terms of memory recall was partly dealt with through the use of a simple dichotomous measure, making for very conservative estimates of child abuse before the sixth grade. Nevertheless, prospective measures would have been more desirable. Finally, we should note that, based on the age groups targeted by Add Health, we cannot capture and measure self-control until middle adolescence in these data. Essentially, we cannot control for earlier selfcontrol. According to Gottfredson and Hirschi (1990), self-control is established by approximately age 10, but research has found that selfcontrol can change over other periods of the life course (e.g., Burt, Simons, & Simons, 2006; Burt, Sweeten, & Simons, 2014). Still, not being able to measure self-control until middle adolescence is a limitation of the Add Health, and thus, the present study. Finally, we should also note that our study does not account for potential genetic confounds. Research in the biosocial criminological literature has found that the association between parenting and selfcontrol/delinquency can partly be explained and is possibly moderated by genetics (e.g., Barnes, Boutwell, Beaver, Gibson, & Wright, 2014; Belsky & Beaver, 2011; Watts & McNulty, 2016; Wright, Beaver, Delisi, & Vaughn, 2008). Given that we partly cast child abuse as an extreme form of harsh parenting, and past studies have shown that the

7. Conclusion In sum, our study suggests the validity of the relatively unexamined GST claim that the influence of victimization on offending is partially mediated by lowered self-control. This lends further support to GST and indicates the utility of the GST framework in understanding the relationship between child abuse and offending. This study contributed to the literature by directly testing this GST proposition using nationally representative data, with a focus on the interrelationships between child abuse, self-control, and delinquency. We hope that future research continues to examine GST and its predictions concerning victimization, self-control, and offending, exploring the applicability of these predictions across different ages and with other types of victimization. Appendix A. Items for low self-control measure 1. 2. 3. 4. 5. 6.

All things considered, how is your child's life going? You get along well with your child. You can trust your child. Does your child have a bad temper? You never argue with anyone. When you get what you want, it's usually because you worked hard for it. 7. You never criticize other people. 8. You usually go out of your way to avoid having to deal with problems in your life. 7

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

weak parental controls? Violence and Victims, 18(2), 143–162. Clinkinbeard, S. S. (2014). What lies ahead: an exploration of future orientation, selfcontrol, and delinquency. Criminal Justice Review, 39(1), 19–36. Davidson, R. J., Putnam, K. M., & Larson, C. L. (2000). Dysfunction in the neural circuitry of emotion regulation—A possible prelude to violence. Science, 289(5479), 591–594. DeLisi, M., Kosloski, A. E., Vaughn, M. G., Caudill, J. W., & Trulson, C. R. (2014). Does childhood sexual abuse victimization translate into juvenile sexual offending? New evidence. Violence and Victims, 29, 620–635. Dembo, R., Schmeidler, J., & Childs, K. (2007). Correlates of male and female juvenile offender abuse experiences. Journal of Child Sexual Abuse, 16(3), 75–94. Freedman, D. A. (1987). As others see us: A case study in path analysis. Journal of Educational Statistics, 12(2), 101–128. Gao, Y., Wong, D. S., & Yu, Y. (2016). Maltreatment and delinquency in China: Examining and extending the intervening process of general strain theory. International Journal of Offender Therapy and Comparative Criminology, 60, 38–61. Glaser, D. (2000). Child abuse and neglect and the brain—a review. Journal of Child Psychology and Psychiatry, 41(1), 97–116. Godbout, N., Vaillancourt-Morel, M. P., Bigras, N., Briere, J., Hébert, M., Runtz, M., & Sabourin, S. (2017). Intimate partner violence in male survivors of child maltreatment: A meta-analysis. Trauma, Violence, & Abuse. Advance online publication https://doi.org/10.1177/1524838017692382. Golier, J., & Yehuda, R. (1998). Neuroendocrine activity and memory-related impairments in post-traumatic stress disorder. Development and Psychopathology, 10, 857–869. Gómez, A. M. (2011). Testing the cycle of violence hypothesis: Child abuse and adolescent dating violence as predictors of intimate partner violence in young adulthood. Youth Society, 43, 171–192. Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press. Grasmick, H. G., Tittle, C. R., Bursik, R. J., Jr., & Arneklev, B. J. (1993). Testing the core empirical implications of Gottfredson and Hirschi's general theory of crime. Journal of Research in Crime and Delinquency, 30(1), 5–29. Gutierres, S. E., & Van Puymbroeck, C. (2006). Childhood and adult violence in the lives of women who misuse substances. Aggression and Violent Behavior, 11(5), 497–513. Harris, K. M., Florey, T., Tabor, J., Bearman, P. S., Jones, J., & Udry, J. R. (2003). The national longitudinal study of adolescent health: Research design. Retrieved from http://www.cpc.unc/edu/projects/addhealth/design. Hay, C. (2001). Parenting, self-control, and delinquency: A test of self-control theory. Criminology, 39, 707–736. Hay, C., & Evans, M. M. (2006). Violent victimization and involvement in delinquency: Examining predictions from general strain theory. Journal of Criminal Justice, 34, 261–274. Hay, C., & Forrest, W. (2006). The development of self-control: Examining self-control theory's stability thesis. Criminology, 44(4), 739–774. Haynie, D. L., Soller, B., & Williams, K. (2014). Anticipating early fatality: Friends', schoolmates' and individual perceptions of fatality on adolescent risk behaviors. Journal of Youth and Adolescence, 43(2), 175–192. Heim, C., & Nemeroff, C. B. (2001). The role of childhood trauma in the neurobiology of mood and anxiety disorders: Preclinical and clinical studies. Biological Psychiatry, 49, 1023–1039. Herrenkohl, T. I., Huang, B., Tajima, E. A., & Whitney, S. D. (2003). Examining the link between child abuse and youth violence: An analysis of mediating mechanisms. Journal of Interpersonal Violence, 18(10), 1189–1208. 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. Hu, L. T., & Bentler, P. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. Iratzoqui, A. (2015). Strain and opportunity: A theory of repeat victimization. Journal of Interpersonal Violence. Advance online publication https://doi.org/10.1177/ 0886260515615146. Jang, S. J. (2007). Gender differences in strain, negative emotions, and coping behaviors: A general strain theory approach. Justice Quarterly, 24(3), 523–553. Jang, S. J., & Johnson, B. R. (2003). Strain, negative emotions, and deviant coping among African Americans. Journal of Quantitative Criminology, 19, 79–105. Kim-Cohen, J., Caspi, A., Taylor, A., Williams, B., Newcombe, R., Craig, I. W., & Moffitt, T. E. (2006). MAOA, maltreatment, and gene-environment interaction predicting children's mental health: new evidence and a meta-analysis. Molecular Psychiatry, 11(10), 903–913. Klika, J. B., Herrenkohl, T. I., & Lee, J. O. (2013). School factors as moderators of the relationship between physical child abuse and pathways of antisocial behavior. Journal of Interpersonal Violence, 28(4), 852–867. Lauritsen, J. L., Sampson, R. J., & Laub, J. H. (1991). The link between offending and victimization among adolescents. Criminology, 29, 265–292. Lin, W. H., Cochran, J. K., & Mieczkowski, T. (2011). Direct and vicarious violent victimization and juvenile delinquency: An application of general strain theory. Sociological Inquiry, 81, 195–222. Longshore, D. (1998). Self-control and criminal opportunity: A prospective test of the general theory of crime. Social Problems, 45(1), 102–113. Longshore, D., & Turner, S. (1998). Self-control and criminal opportunity: Cross-sectional test of the general theory of crime. Criminal Justice and Behavior, 25(1), 81–98. Manasse, M. E., & Ganem, N. M. (2009). Victimization as a cause of delinquency: The role of depression and gender. Journal of Criminal Justice, 37(4), 371–378. Mazerolle, P., Piquero, A. R., & Capowich, G. E. (2003). Examining the links between strain, situational and dispositional anger, and crime. Youth Society, 35, 131–157.

9. Difficult problems make you very upset. 10. When making decisions, you usually go with your “gut feeling” without thinking too much about the consequences of each alternative. 11. When you have a problem to solve, one of the first things you do is get as many facts about the problem as possible. 12. When attempting to find a solution to a problem, you usually try to think of as many different ways to approach the problem as possible. 13. When making decisions, you generally use a systematic method for judging and comparing alternatives. 14. After carrying out a solution to a problem, you usually try to analyze what went right and what went wrong. 15. You like yourself just the way you are. 16. You feel like you are doing everything just about right. 17. You feel socially accepted. 18. Do you have trouble getting along with your teachers? 19. Do you have trouble paying attention in school? 20. Do you have trouble keeping your mind focused? 21. Do you have trouble getting your homework done? References Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30, 47–88. Agnew, R. (2001). Building on the foundation of general strain theory: Specifying the types of strain most likely to lead to crime and delinquency. Journal of Research in Crime and Delinquency, 38, 319–361. Agnew, R. (2002). Experienced, vicarious, and anticipated strain: An exploratory study on physical victimization and delinquency. Justice Quarterly, 19, 603–632. Agnew, R. (2006). Pressured into crime: An overview of general strain theory. Los Angeles: Roxbury. Agnew, R. (2013). When criminal coping is likely: An extension of general strain theory. Deviant Behavior, 34, 653–670. Agnew, R., Brezina, T., Wright, J. P., & Cullen, F. T. (2002). Strain, personality traits, and delinquency: Extending general strain theory. Criminology, 40, 43–72. Agnew, R., Scheuerman, H., Grosholz, J., Isom, D., Watson, L., & Thaxton, S. (2011). Does victimization reduce self-control? A longitudinal analysis. Journal of Criminal Justice, 39, 169–174. Averdijk, M., Van Gelder, J., Eisner, M., & Ribeaud, D. (2016). Violence begets violence… but how? A decision-making perspective on the victim-offender overlap. Criminology, 54, 282–306. Barnes, J. C., Boutwell, B. B., Beaver, K. M., Gibson, C. L., & Wright, J. P. (2014). On the consequences of ignoring genetic influences in criminological research. Journal of Criminal Justice, 42(6), 471–482. 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. Beaver, K. M., DeLisi, M., Mears, D. P., & Stewart, E. (2009). Low self-control and contact with the criminal justice system in a nationally representative sample of males. Justice Quarterly, 26(4), 695–715. Beaver, K. M., Wright, J. P., DeLisi, M., & Vaughn, M. G. (2008). Genetic influences on the stability of low self-control: Results from a longitudinal sample of twins. Journal of Criminal Justice, 36(6), 478–485. Belsky, J., & Beaver, K. M. (2011). Cumulative-genetic plasticity, parenting and adolescent self-regulation. Journal of Child Psychology and Psychiatry, 52(5), 619–626. Bergen, H. A., Martin, G., Richardson, A. S., Allison, S., & Roeger, L. (2004). Sexual abuse, antisocial behaviour and substance use: Gender differences in young community adolescents. The Australian and New Zealand Journal of Psychiatry, 38(1/2), 34–41. Bollen, K. A. (1989). Causality and causal models. Structural Equations With Latent Variables (pp. 40–79). New York: Wiley. Brems, C., Johnson, M., Neal, D., & Freemon, M. (2004). Childhood abuse history and substance use among men and women receiving detoxification services. The American Journal of Drug and Alcohol Abuse, 30(4), 799–821. Brezina, T. (1998). Adolescent maltreatment and delinquency: The question of intervening processes. Journal of Research in Crime and Delinquency, 35, 71–99. Burt, C. H., Simons, R. L., & Simons, L. G. (2006). A longitudinal test of the effects of parenting and the stability of self-control: Negative evidence for the general theory of crime. Criminology, 44, 353–396. Burt, C. H., Sweeten, G., & Simons, R. L. (2014). Self-control through emerging adulthood: Instability, multidimensionality, and criminological significance. Criminology, 52, 450–487. Carson, D. C., Sullivan, C. J., Cochran, J. K., & Lersch, K. M. (2009). General strain theory and the relationship between early victimization and drug use. Deviant Behavior, 30(1), 54–88. Caspi, A., McClay, J., Moffitt, T. E., Mill, J., Martin, J., Craig, I. W., ... Poulton, R. (2002). Role of genotype in the cycle of violence in maltreated children. Science, 297(5582), 851–854. Chapple, C. L. (2003). Examining intergenerational violence: Violent role modeling or

8

Journal of Criminal Justice xxx (xxxx) xxx–xxx

J.M. Bunch et al.

outcomes of young adults exposed to maltreatment: The role of educational experiences in promoting resilience to crime and violence in early adulthood. Journal of Interpersonal Violence, 28, 121–156. StataCorp (2013). Stata statistical software: Release 13. College Station, TX: StataCorp LP. Sullivan, T. N., Farrell, A. D., Kliewer, W., Vulin-Reynolds, M., & Valois, R. F. (2007). Exposure to violence in early adolescence: The impact of self-restraint, witnessing violence, and victimization on aggression and drug use. The Journal of Early Adolescence, 27, 296–323. Tankebe, J. (2013). Viewing things differently: The dimensions of public perceptions of police legitimacy. Criminology, 51(1), 103–135. Teague, R., Mazerolle, P., Legosz, M., & Sanderson, J. (2008). Linking childhood exposure to physical abuse and adult offending: Examining mediating factors and gendered relationships. Justice Quarterly, 25, 313–348. Teicher, M. H., Andersen, S. L., Polcari, A., Anderson, C. M., Navalta, C. P., & Kim, D. M. (2003). The neurobiological consequences of early stress and childhood maltreatment. Neuroscience and Biobehavioral Reviews, 27, 33–44. Turanovic, J. J., & Pratt, T. C. (2013). The consequences of maladaptive coping: Integrating general strain and self-control theories to specify a causal pathway between victimization and offending. Journal of Quantitative Criminology, 29, 321–345. Turner, M. G., & Piquero, A. R. (2002). The stability of self-control. Journal of Criminal Justice, 30, 457–471. Vaillancourt, T., Hymel, S., & McDougall, P. (2013). The biological underpinnings of peer victimization: Understanding why and how the effects of bullying can last a lifetime. Theory Into Practice, 52, 241–248. van der Kolk, B. A., & Fisler, R. E. (1994). Childhood abuse and neglect and loss of selfregulation. Bulletin of the Menninger Clinic, 58(2), 145. Watts, S. J. (2016). The link between child abuse and neglect and delinquency: Examining the mediating role of social bonds. Victims & Offenders, 12(5), 700–717. Watts, S. J., & McNulty, T. L. (2013). Childhood abuse and criminal behavior: Testing a general strain theory model. Journal of Interpersonal Violence, 28, 3023–3040. Watts, S. J., & McNulty, T. L. (2016). Genes, parenting, self-control, and criminal behavior. International Journal of Offender Therapy and Comparative Criminology, 60(4), 469–491. Winfree, L. T., Taylor, T. J., He, N., & Esbensen, F.-A. (2006). Self-control and variability over time: Multivariate results using a 5-year multi-site panel of youths. Crime & Delinquency, 52, 253–286. Wright, J., Beaver, K., Delisi, M., & Vaughn, M. (2008). Evidence of negligible parenting influences on self-control, delinquent peers, and delinquency in a sample of twins. Justice Quarterly, 25(3), 544–569.

Meldrum, R. C., Verhoeven, M., Junger, M., van Aken, M. A., & Deković, M. (2016). Parental self-control and the development of male aggression in early childhood: A longitudinal test of self-control theory. International Journal of Offender Therapy and Comparative Criminology. Advance online publication https://doi.org/10.1177/ 0306624X16662921. Monahan, K. C., King, K. M., Shulman, E. P., Cauffman, E., & Chassin, L. (2015). The effects of violence exposure on the development of impulse control and future orientation across adolescence and early adulthood: Time-specific and generalized effects in a sample of juvenile offenders. Development and Psychopathology, 27, 1267–1283. Muthén, L. K., & Muthén, B. O. (2006). Mplus Version 7 user's guide. Los Angeles, CA: Muthén & Muthén. Neff, J. L., & Waite, D. E. (2007). Male versus female substance abuse patterns among incarcerated juvenile offenders: Comparing strain and social learning variables. Justice Quarterly, 24(1), 106–132. Norman, R. E., Byambaa, M., De, R., Butchart, A., Scott, J., & Vos, T. (2012). The longterm health consequences of child physical abuse, emotional abuse, and neglect: A systematic review and meta-analysis. PLoS Medicine, 9(11), e1001349. Ompad, D. C., Ikeda, R. M., Shah, N., Fuller, C. M., Bailey, S., Morse, E., & Strathdee, S. A. (2005). Childhood sexual abuse and age at initiation of injection drug use. American Journal of Public Health, 95(4), 703–709. Papachristos, A. V., Braga, A. A., Piza, E., & Grossman, L. S. (2015). The company you keep? The spillover effects of gang membership on individual gunshot victimization in a co-offending network. Criminology, 53, 624–649. Perrone, D., Sullivan, C. J., Pratt, T. C., & Margaryan, S. (2004). Parental efficacy, selfcontrol, and delinquency: A test of a general theory of crime on a nationally representative sample of youth. International Journal of Offender Therapy and Comparative Criminology, 48(3), 298–312. Piquero, A. R., Jennings, W. G., & Farrington, D. P. (2010). On the malleability of selfcontrol: Theoretical and policy implications regarding a General Theory of Crime. Justice Quarterly, 27, 803–834. Pratt, T. C., & Cullen, F. T. (2000). The empirical status of Gottfredson and Hirschi's general theory of crime: A meta-analysis. Criminology, 38(3), 931–964. Pratt, T. C., Turanovic, J. J., Fox, K. A., & Wright, K. A. (2014). Self-control and victimization: A meta-analysis. Criminology, 52, 87–116. Robertson, A., Stein, J., & Schaefer-Rohleder, L. (2010). Effects of hurricane Katrina and other adverse life events on adolescent female offenders: A test of general strain theory. Journal of Research in Crime and Delinquency, 47(4), 469–495. Salzinger, S., Rosario, M., & Feldman, R. S. (2007). Physical child abuse and adolescent violent delinquency: The mediating and moderating roles of personal relationships. Child Maltreatment, 12, 208–219. Savage, J., Palmer, J. E., & Martin, A. B. (2014). Intergenerational transmission: Physical abuse and violent vs. nonviolent criminal outcomes. Journal of Family Violence, 29(7), 739–748. Schore, A. N. (2001). The effects of early relational trauma on right brain development, affect regulation, and infant mental health. Infant Mental Health Journal, 22(1–2), 201–269. Shaffer, J. N., & Ruback, R. B. (2002). Violent victimization as a risk factor for violent offending among juveniles. Juvenile Justice Bulletin. Washington, D.C.: U.S Department of Justice, Office of Juvenile Justice and Delinquency Prevention. 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. Simons, R. L., Whitbeck, L. B., Conger, R. D., & Wu, C. I. (1991). Intergenerational transmission of harsh parenting. Developmental Psychology, 27, 159–171. Smith, C. A., Park, A., Ireland, T. O., Elwyn, L., & Thornberry, T. P. (2013). Long-term

Jackson M. Bunch is an assistant professor in the Department of Sociology at the University of Montana. His research examines the causes and consequences of victimization, including routine activities, fear of crime, and the relationship between offending and victimization. Amaia Iratzoqui is an Assistant Professor in the Department of Criminology and Criminal Justice at the University of Memphis. Her research interests include the gendered nature of victimization, repeat victimization, intimate partner violence, and theoretical development. Stephen J. Watts is an Assistant Professor in the Department of Criminology and Criminal Justice at the University of Memphis. His research focuses on the victimization/ offending overlap, the testing of criminological theories, and the integration of these theories into a biosocial framework for explaining antisocial behavior.

9