اشکال پنهان و آشکار رابطه پرخاشگری واکنشی در نوجوانان: روابط با واکنش خون گرم و خودتنظیمی

This study examined whether overt and relational forms of reactive aggression were differentially related to adolescents’ temperament. Measures of adolescents’ temperament and aggression were completed by 670 adolescents (369 females), ages 10–17, and their mothers. Effortful control and fearfulness were inversely associated only with reactive–overt aggression, whereas frustration proneness was more strongly linked with reactive–relational aggression. Furthermore, amongst younger adolescents, effortful control had a larger association with reactive–overt aggression when fearfulness was low, whereas frustration proneness had a stronger relation to reactive–relational aggression when effortful control was high. The differential relations between the two forms of reactive aggression (i.e., overt and relational) and effortful control or fearfulness are discussed with respect to variations in the riskiness and the social competence required to implement these aggressive actions.

Aggressive behavior in childhood predicts a host of negative developmental outcomes in adolescence and adulthood, including substance dependence, high school drop-out, mental health problems, financial problems, unemployment, interpartner violence, criminal offending, and imprisonment (Fergusson et al., 2005 and Moffitt et al., 2002). In addition, conduct problems are the primary reason that children and adolescents are referred to mental health clinics (Kazdin, 1995). Consequently, it is crucial to study the development of aggressive behavior with a view to developing effective prevention and treatment programs. Several models of aggressive behavior specify that temperament or personality-based predispositions affecting emotional reactivity and self-regulation abilities are important risk factors (e.g., Anderson and Bushman, 2002, Berkowitz, 2012 and Dodge and Pettit, 2003). Temperament is defined as biologically-based individual differences in emotional, attentional and motor reactivity, and in self-regulation processes, such as effortful control and reactive control (Rothbart, 2011 and Rothbart and Bates, 2006). Proneness to frustration is a facet of emotional reactivity that has been shown to increase risk of aggression (Rothbart, 2011), though two aspects of self-regulation—effortful control or reactive control – may diminish this association. Effortful control involves conscious, voluntary, effortful, and cognitive strategies (e.g., stopping and thinking about consequences prior to acting) that facilitate the inhibition of a dominant impulse to permit the performance of a subdominant response, as well as planning and error detection (Eisenberg et al., 2010 and Rothbart and Bates, 2006). Reactive control, on the other hand, entails automatic, unconscious, and emotion-related self-regulation processes that reflect the balance in the reactivity of approach and avoidance motivation systems (Eisenberg et al., 2010 and Rothbart, 2011). For example, an easily frustrated adolescent who is relatively fearless may be predisposed to engage in an approach-related aggressive response to provocation rather than avoiding confrontation (Rothbart, Ellis, & Posner, 2011). Temperament contributes indirectly to the development of conduct problems by influencing the mental processes or internal states (e.g., cognitions, emotions, arousal levels, action tendencies) that arise in social situations, such as a perceived provocation by parents or peers (e.g., Anderson and Bushman, 2002, Berkowitz, 2012 and Dodge and Pettit, 2003). Temperament has been shown to differentiate reactive and proactive functions of aggression. In contrast to proactively aggressive attacks, which are goal-directed, instrumental and deliberate, reactively aggressive acts are described as provoked, retaliatory, defensive, anger-driven, emotionally dysregulated, and impulsive (Card and Little, 2006, Hubbard et al., 2010 and Vitaro et al., 2006). Consistent with the Frustration–Aggression and Cognitive Neoassociation model of aggressive behavior (Berkowitz, 1993 and Berkowitz, 2012), reactive aggression has been associated with predispositions that contribute to difficulties with the regulation of negative affect (Hubbard et al., 2010), which include anxiety, angry reactivity, emotional dysregulation and inattention (see Vitaro et al., 2006). Furthermore, a recent meta-analysis indicates that reactive aggression, rather than proactive aggression, is independently associated with measures of emotional dysregulation and ADHD symptoms (Card & Little, 2006). However, much of the previous research that has examined distinctions between reactive and proactive aggression has employed measures that emphasize physical aggression (Vitaro et al., 2006). Therefore, the purpose of the present study is to investigate the link between temperament and both overt and relational forms of reactive aggression. Currently, there is very little research differentiating reactive–overt from reactive–relational aggression. In contrast to the physical or verbal attacks that characterize overt aggression, relational aggression involves harming the victim by damaging relationships or social status, and by virtue of being more indirect and covert, it allows the aggressor to avoid detection as the perpetrator (Card et al., 2008 and Crick and Grotpeter, 1995). A reactive–relationally aggressive response involves retaliation through acts such as social exclusion or rumor spreading. Amongst adolescent participants, reactive–relational aggression is associated with perceived popularity and social preference, whereas reactive–overt aggression is negatively related to these same outcomes (Prinstein & Cillessen, 2003). Furthermore, reactive–overt aggression also has a stronger association than reactive–relational aggression with adolescents’ violent delinquency and arrest history (Marsee et al., 2011). Finally, in a sample of adolescent girls in a detention centre, reactive–overt but not reactive–relational aggression was uniquely associated with emotional dysregulation (Marsee & Frick, 2007). Additional research is required to illuminate the temperament dimensions that distinguish the heterogenous forms of reactive aggression. We expected that two facets of self-regulation – effortful control and reactive control – would constitute protective factors for reactive–overt aggression as opposed to reactive–relational aggression. Although the ability to inhibit impulses and regulate emotions is negatively related to the frequency of overt and relational aggression (Card et al., 2008), the use of relational aggression also appears to be facilitated by social competence (Björkqvist, 1994). For example, it is done more often by youth with high levels of social intelligence or social status (Cillessen and Mayeux, 2004 and Kaukiainen et al., 1999). Furthermore, a meta-analysis indicates that measures of poor self-regulation are more strongly associated with overt aggression (Card et al., 2008). Thus, there are empirical and theoretical reasons to suggest that complex relationally aggressive responses to provocation, such as social exclusion, may be facilitated by good effortful control. Therefore, because effortful control may both inhibit and facilitate reactive–relational aggression, we hypothesized that effortful control would be more strongly related to reactive–overt than to reactive–relational aggression, and that it would significantly moderate the association between frustration proneness and only the overt form of reactive aggression. We further predicted that fearful youth would display lower frequencies of reactive–overt aggression. We reasoned that fear would affect participation in reactive–overt aggression rather than reactive–relational aggression because the former is more dangerous and risky, given that the perpetrator is more likely to sustain a physical injury, to be identified as the aggressor and, consequently, to face retaliation (Björkqvist, 1994). Indeed, some writers taking an evolutionary perspective have stated that females prefer to use relational rather than overt aggression primarily because it is a safer option (Vaillancourt, 2005 and Volk et al., 2012). Consistent with this expectation, Terranova, Morris, and Boxer (2008) found that temperamental fearfulness was inversely associated with future overt but not relational bullying.

Analyses indicated that missing data were well under the recommended cutoff point of 3%, both at the item and variable level (Tabachnick & Fidell, 2001). As shown in Table 1, reactive–overt aggression and reactive–relational aggression were strongly associated, though the correlation was somewhat lower than the average for pure overt and relational aggression (Card et al., 2008). As predicted, there were moderate, inverse links between effortful control and both forms of reactive aggression, whereas fearfulness was negatively associated only with reactive–overt aggression. Frustration proneness was positively associated with both forms of reactive aggression. Demographic variables such as ethnicity, family income, maternal education, marital status and maternal age were not associated with temperament or aggression measures, and thus were excluded from analyses.