“Above All Things, Be Glad and Young”: Advancing Research on Violence in Adolescence

Abstract

Adolescence is a unique developmental period in which youth are at heightened risk for violence perpetration and victimization, both of which can adversely affect the course of normal psychosocial development and health across the life course. The possible range of violent experiences during adolescence is broad, including physical and relational peer violence, bullying, sexual violence, cyber-aggression, child maltreatment, exposure to parental intimate partner violence, exposure to aggressive media, and more. The 8 articles in this special issue span this wide range of violence experiences. In helping to fill gaps in knowledge about the nature and processes by which violence develops, and how violence experiences affect adolescents, these articles as a group also offer direction for future research. They illustrate the need for research that cuts across the separate bodies of research on violence involving youth, that tackles thorny questions about the conceptualization of violence in its myriad forms, and that considers the ways in which the various forms and dimensions of violence operate together, across levels of the social ecology, to influence outcomes. The findings in this collection of studies also have implications for the development of prevention and intervention programs to address the problem of violence in adolescence.

Adolescent violence and victimization present tremendous public health challenges. Youth homicide is the second leading cause of death among adolescents aged 15–24 in the United States, and suicide is the third among the same age group (Centers for Disease Control and Prevention, 2010). Recent 1-year prevalence rates of physical fighting (32.8%), bullying at school (i.e., in-person bullying, 12%), cyberbullying (16%), physical dating violence (9.4%), attempted suicide (7.8%), and lifetime prevalence of forced sexual intercourse (8%) underscore the extent of the problem among U.S. high school students (Eaton et al., 2012). Today’s adolescents have ready access and opportunity for virtually constant exposure to violence through digital technology and mass media, and many have been or are currently exposed to violence in their communities and their families—either interparental violence, child maltreatment, or both (Finkelhor, Turner, Ormrod, Hamby, & Kracke, 2009). Adolescents who perpetrate violence, as well as those who are victims of violence and violence exposure, are at increased risk for a wide range of mental and physical health impairments, including changes in brain architecture and development, and risk behaviors throughout the life span (Dahlberg & Krug, 2002; Shonkoff, Boyce, & McEwen, 2009). Thus, understanding the unfolding of violence in adolescence, and the complexity of factors that influence its developmental course, are critical for prevention and intervention efforts, or as E. E. Cummings (1938) phrased it, helping adolescents to, “above all things, be glad and young.”

Adolescence places great demands on the adaptive capacities of youth. Key developmental tasks of adolescence involve substantial changes in roles and relationships necessary for a successful transition from childhood to emerging adulthood. Indeed, across all levels of the social ecological system, the forces that shape development change over the course of adolescence. Susceptibility to peer influence is heightened, parental influence weakens, and new forces, such as romantic and sexual relationships, emerge. These forces operate against a backdrop of a burst in brain development and reorganization of cognitive and emotional processes that can render adolescents’ problemsolving and decision-making capacities unstable and suboptimal (Albert, Chein, & Steinberg, 2013; Burnett, Sebastian, Cohen Kadosh, & Blakemore, 2011). These changing social and psychological factors contribute to adolescents’ heightened risk for perpetrating or being a victim of violence, each of which can complicate and disrupt the trajectory of psychosocial development, with effects that can persist throughout the life course (Tolan & Gorman-Smith, 2002). The open call for this special issue sought papers that would address key gaps in knowledge on violence in adolescence, the contexts within which such violence occurs, and the factors that contribute to its onset, recurrence, and outcomes.

Choosing from among the many excellent contributions submitted in response to the call for papers for this issue was challenging. The selected articles cut a wide swath across the broad field of violence involving youth. They address a range of types and dimensions of violence, risk and protective factors at different socioecological levels that shape the course of violence and victimization, and processes that help explain links between risk factors and outcomes. A couple of important themes are also apparent in this collection of articles, which may provide focus for future research. These themes pertain to the (a) conceptualization of violence and co-occurring types of violence and how both of those relate to youth outcomes, and (b) the multiple influences and pathways across levels of the social ecology that are important for understanding violence and its outcomes and for informing interventions designed to address them. We describe how the selected articles contribute to these themes below.

Conceptualizations of Violence and Its Relation to Outcomes

Research on violence in the lives of children and adolescents has often focused on a particular form of violence, such as child abuse, intimate partner violence (IPV), teen dating violence (TDV), bullying, or community violence. As knowledge from these separate bodies of research has accumulated, and research on polyvictimization has emerged, it has become clear that youth who are exposed to any single form of violence are commonly exposed to other forms as well (Finkelhor, Ormrod, Turner, & Hamby, 2005; Finkelhor et al., 2009). And although individual forms of youth victimization and violence exposure have been linked to poorer psychosocial adjustment, there is evidence that the breadth or cumulation of violence exposure may matter as much or more than the particular type of violence experienced (Anda et al., 2011; Finkelhor, Ormrod, & Turner, 2007; Finkelhor, Shattuck, Turner, Ormrod, & Hamby, 2011). Several articles in this issue extend our knowledge and propose new questions about the definition of violence, the breadth of victimization and violence exposure, and the effects of multiple forms of victimization on adolescent outcomes.

Two of the studies highlight the complexity of relationships among the various forms of violence and their effects on youth adjustment. Wigderson and Lynch (2013) examined whether cyber-victimization—the use of technology to harass, humiliate, threaten or insult someone purposely—exerts effects on adolescent adjustment, over and above the effects of physical and relational victimization. They found that cyber-victimization and relational victimization, but not physical victimization uniquely predicted emotional problems. Cybervictimization alone uniquely predicted grade point average (GPA), but this effect was moderated by physical victimization and by relational victimization. The pattern of the moderating effect is intriguing: The relation between cyber-victimization and GPA was stronger among youth who had experienced higher, rather than lower, levels of physical victimization, but the relation was lower among youth who had experienced higher levels of relational victimization. Extending the conceptualization of violence to include perpetration via digital technology, examining cyber-aggression together with more traditional forms of violent victimization, and testing for interactions among these forms of violence adds sophistication and clarity to knowledge on the topic of violent victimization. Greater attention to such potential complex relations is needed to understand the circumstances under which one form of violence exposure may exacerbate or perhaps override or mitigate the effects of other types of violence exposure on youth outcomes.

Friedlander, Connolly, Pepler, and Craig (2013) also addressed definitional issues in their study on youth exposure to aggressive media. Specifically, they considered breadth as well as persistence of adolescent exposure to violence across various forms of media, such as movies, magazines, music, TV, and websites, and their relations to TDV perpetration and victimization. They found that the sheer amount of adolescent exposure to aggressive media is great and is stable over time. They also found that breadth of exposure over a 3-year time period predicted TDV perpetration and victimization; however, the chronicity of exposure over time predicted TDV perpetration and victimization only among those who had been exposed to multiple (3 or more) types of aggressive media. Taken together, the findings of Wigderson and Lynch (2013) and Friedlander et al. can inform public discourse and policy regarding victimization and the effects of exposure to violence through technology and mass media.

In a study of the importance of the conceptualization of violence, Hamby and Turner (2013) illustrate how the prevalence of TDV victimization across males and females hinges upon how TDV is operationalized. Their study richly informs the debate about gender differences in TDV (and, by extension, IPV) victimization, by highlighting clear, gender-specific patterns of victimization that change as the conceptualization of TDV is broadened in very reasonable ways. Using a conventional operationalization of dating violence—acts of physical aggression—they found what others have found before: Rates of victimization were higher for males than females. However, when the conceptualization of TDV was broadened to also include sexual violence, injurious violence, or fear, rates of victimization were generally higher for women than men. Such research greatly informs our understanding of the phenomenon of dating violence and the experiences of those victimized by it. Together with the studies by Friedlander et al. (2013) and Wigderson and Lynch (2013), this study underscores the need for researchers to carefully attend to the conceptualization of violence and its dimensions, as well as the overlap among the types of violence that adolescent’s witness, experience, and perpetrate. The call for papers for the next special issue of Psychology of Violence, which will focus on measurement of violence, further highlights this need.

Pathways of Influence and Implications for Intervention

In adolescence, the types, potency, and combination of particular risk and protective factors pertaining to violence may differ from those of childhood. In addition, the nature and extent of previous effects of violence perpetration or victimization may be altered by more recent experiences as well as by physical and psychological maturation and the changing social contexts that characterize adolescent development. The processes that mediate and moderate the effects of violence experiences on outcomes may also undergo change during adolescence. Understanding how all of these forces act in concert and over time to influence youth outcomes is pivotal for knowing when and how to intervene with youth at risk for poor outcomes because of violence, as well as for developing theory that facilitates the identification of those most at risk and those who are resilient in the face of violence experiences.

Findings from a number of the studies in this issue highlight potential useful targets of intervention and extend knowledge on the pathways by which violent perpetration and victimization, and their effects, unfold over the course of adolescence. Miller and Esposito-Smythers (2013), for example, examined two individuallevel factors that may alter the relationship between child physical and sexual abuse experiences and suicidal ideation: cognitive distortions and substance-related problems. The cognitive–behavioral theory of adolescent suicide would suggest that child abuse may result in the development of maladaptive responses to acute stress, such as distorted cognitions and substance use, which fail to relieve distress and ultimately contribute to suicidal ideation. However, Miller and Esposito-Smythers found no evidence that substance-related problems or cognitive distortions mediated the relationship between child maltreatment and suicidal ideation in their psychiatric inpatient sample. In fact, they found that child abuse did not exert direct effects on suicidal ideation. Substance use problems, but not cognitive distortions, acted as moderators, however, such that among adolescents who had been maltreated, those with recent substance-related problems were at elevated risk for suicidal ideation. The authors raise intriguing questions about the persistence of effects of maltreatment on cognitive distortions and about the relative contributions of distal factors, such as child maltreatment, versus proximal factors, such as cognitive distortions and substance use, on current adolescent functioning. They also note that clinically, substance-related problems may be a potential intervention target for reducing suicide risk. Attending to issues such as these can help refine and extend theory and guide intervention development in the area of adolescent suicide prevention.

Garrido and Taussig (2013) focused on social influences in the family and community— positive parenting practices and prosocial peers—as moderators of the association between exposure to interparental IPV and adolescents’ TDV experiences. The focus on relationship-level factors as protective is noteworthy; it is more common to see examinations of the risks conferred by negative family and peer processes (e.g., harsh parenting, deviant peers) in the violence literature. Garrido and Taussig found that positive parenting yielded protective effects against TDV perpetration as well as victimization. Although most TDV prevention programs involve only the adolescents, and most are offered in the school setting (Jouriles, Platt, & McDonald, 2009), a few have been adapted to also involve caregivers and peers (e.g., Foshee et al., 2012; Tharp et al., 2011; Wolfe, Jaffe, & Crooks, 2006). Garrido and Taussig’s findings suggest that for adolescents exposed to IPV, it may be useful for interventions addressing this population to include a focus on strengthening positive parenting practices where they are deficient. The notion of including caregivers in an intervention targeting adolescent problem behavior also converges with other research in which a family based intervention has been found to be effective for treating anorexia, a very serious and often treatmentrefractory adolescent mental health problem (Lock, Agras, Bryson, & Kraemer, 2005; Lock, Couturier, & Agras, 2006).

Garrido and Taussig (2013) also found that prosocial peers yielded protective effects for TDV perpetration. This suggests that for adolescents who have been in the child welfare system and who have also been exposed to IPV, establishing relationships with prosocial peers may be particularly helpful in breaking the link between IPV exposure and TDV perpetration. Prosocial peers did not confer protection for TDV victimization, however. These findings help elucidate salient points of prevention and intervention efforts for adolescents by venturing outside of the self and the family of origin (given that two-thirds of this sample were in out-of-home care) into more exterior layers of the social ecology in which adolescents are not only spending more time but are increasingly invested and influenced. It may be fruitful to explore strategies that allow for peers to help protect vulnerable, IPV-exposed youth from physical TDV victimization—the research on bystander interventions for preventing sexual assault may provide some guidance in that regard. The sample for this study was small, but at high risk for TDV, and the results emerged after controlling for other violence experiences (child physical abuse and exposure to community violence) that can increase risk for TDV, thus the findings regarding protective factors are particularly noteworthy. Recent federal efforts have promoted enhancing protective factors in programs and planning for the primary and secondary prevention of child maltreatment (Centers for Disease Control & Prevention, 2013; U.S. Department of Health and Human Services, 2013); perhaps among maltreated youth, such strategies would also be useful in the prevention of TDV.

Ronis and Borduin (2013) note that individuals who perpetrate sexual violence are often treated differently from nonsexually violent of-fenders in mental health and criminal justice settings, following from the assumption that sexually violent offenders engage in more frequent and more serious criminal acts. They question this assumption in regard to adolescent offenders, highlighting that there is a dearth of developmental data attesting to its validity. In a secondary analysis of data gathered over a 10-year period from a large, national sample of youth, Ronis and Borduin examined whether the trajectories of antisocial behavior for adolescents with a history of sexual violence perpetration differed from those of adolescents with a history only of nonsexual violence perpetration. Similar to others’ findings, they identified three trajectories of antisocial behavior (low, moderate, and chronic). These did not differ across sexual versus nonsexual perpetrators, and there were similar proportions of sexual and nonsexual perpetrators on each of the trajectories. Ronis and Borduin conclude that the development of sexually violent behavior appears similar to that of other antisocial behavior, and that sexually violent adolescents may not be as different from their nonsexually violent counterparts as is often assumed. They also raise the intriguing question of whether interventions that are effective in addressing adolescent antisocial behavior might also be effective for sexually violent adolescents. Their study also provides another example of how broadening conceptualizations of violence and measuring various forms of violence experiences (in this case, sexual and nonsexual perpetration) can help extend theory and knowledge on violence in the lives of adolescents.

Borofsky, Kellerman, Baucom, Oliver, and Margolin (2013) conducted a longitudinal study to examine whether community violence exposure was negatively associated with academic outcomes. Their findings add to knowledge about community violence effects on adolescents in several ways. First, they demonstrated that community violence exposure negatively predicted later school engagement, but not vice versa. They also provided suggestive evidence of a cascade of effects proceeding from community violence exposure to adjustment problems, reductions in school engagement, and reductions in GPA. Specifically, they found that community violence exposure predicted GPA, and school engagement mediated this association. And they found that internalizing and externalizing problems, but not symptoms of traumatic stress, functioned as mediators of the association between community violence and school engagement. Finally, as others in this issue have done, they also controlled for another form of violence—parent–child aggression— that may have otherwise accounted for their findings. Their findings add to theory on the processes by which exposure to community violence can affect academic outcomes. They also suggest that clinically, psychological symptoms that accrue from exposure to community violence may herald reductions in school engagement and later decrements in GPA. This would suggest that in addition to addressing the psychological symptoms, interventions that focus on helping adolescents remain engaged with their schools and other safe and supportive environments may help prevent decreases in academic performance.

Fanti and Kimonis (2013) extend their previous cross-sectional work linking callous-unemotional (CU) traits of youth psychopathy with perpetration of bullying. They consider whether psychopathy predicts bullying perpetration and victimization prospectively (1 year later), and whether measuring narcissism and impulsivity—two dimensions of youth psychopathy that have received little attention in this area—along with CU traits, can help distinguish bullies, victims, and bully victims from one another. They found additive effects for each of the three dimensions of youth psychopathy in predicting bullying perpetration, after controlling for conduct problems (which also predicted perpetration). However, only impulsivity uniquely predicted victimization. Results of their person-centered analyses indicated that all three dimensions of psychopathy (as well as conduct problems) were useful in predicting group membership (bullies, victims, and bully victims), but that narcissism was particularly instrumental in identifying who was likely to bully others, and impulsivity predicted who was likely to be bullied. Fanti and Kimonis point out implications of their results for the design of prevention and intervention programs (e.g., youth high on narcissism may be less likely to respond to interventions that focus on developing empathy for victims; impulsive youth may benefit from social skills training) in schools. This appears to be a promising line of research for enhancing our understanding of psychopathic traits and their links to interpersonal violence and victimization.

As noted in the section on expanding conceptualizations of violence, Friedlander and colleagues (2013) found that the breadth of exposure to aggressive media predicted TDV perpetration and victimization, and that the effect was mediated by violence-tolerant attitudes. The authors suggest that incorporating this knowledge into existing TDV prevention programming may be helpful for enhancing their effectiveness. They also note that education about aggression in media, and media literacy programs, which have altered youth attitudes in other domains, may also be of benefit. Theirs is the first study to focus on effects of aggressive media exposure on TDV. And although the larger debate about the effects of violent media on youth aggression continues, Friedlander et al.’s findings parallel those of other studies, indicating that the extent of violent media consumption (e.g., video games; Willoughby, Adachi, & Good, 2012) is associated with youth violence and aggression. It is important that Friedlander et al. considered the influence of a factor at the outermost socialecological level on an outcome at the individual/peer level. Examining whether exposure to media aggression predicts TDV independently of other risk factors for TDV at other levels of the social ecology would contribute useful knowledge about the specificity of effects of media aggression. Research that includes measurement of other violence experiences may also help to determine whether the findings of this study are specific only to the domain of TDV, or whether they also apply to youth violence more broadly.

The Expanding Realm of Violence Research

In addition to summarizing the knowledge generated by the studies themselves, in this introduction we have tried to glean conceptual, methodological, and clinical considerations that may enhance future research efforts and inform the design of violence prevention and intervention programs. The two themes around which we organized the articles were simply those that were most apparent. There are certainly other aspects of the articles that are worth noting, in part because of the issues they raise. For example, we have highlighted that some studies in this issue examined variables within or across particular social–ecological levels. In predicting violence experiences and outcomes, it will likely be useful to give systematic attention to building and extending theory on key factors operating simultaneously, but at different levels of the social ecology (e.g., Borofsky et al., 2013; Friedlander et al., 2013; Garrido & Taussig, 2013), especially given the shifts in individual, relationship, community, and societal influences on behavior during adolescence.

Also, related to the issue of the conceptualization of violence, most of the studies in this issue examined more than one form of violence, whether to isolate effects of particular violence experiences, to examine their joint effects on outcomes, or their relations to one another. However, there is not much guidance available on when to include additional forms of violence and which ones to include. In some instances, it may be clear which types are most relevant for a particular study (e.g., parental IPV is routinely associated with child maltreatment and with youth aggression). In others, where it seems less clear—for example, in the case of technology-based aggression—utilizing knowledge of adolescent development and relevant theory in guiding such decisions is warranted. A developmentally sensitive understanding of the interplay among the multiple forms of violence experiences and how they relate collectively to outcomes would help refine theory and improve prediction of risk.

The sheer breadth of the literature on violence in the lives of adolescents is clear even from the few studies in this issue. It also seems clear that the literature will continue to broaden, to better accommodate the knowledge that violence and victimization experiences often emerge and recur in the context of other violence experiences. The contributions of the articles in this issue, and conducting research to address the issues that they collectively raise, may enhance our ability to better protect our youth from victimization and from becoming violent themselves—helping to keep them glad and young.