اثرات متقابل فرد و اجتماع در بزهکاری نوجوانان: هیجان خواهی، انحراف همسالان و نظارت والدین

Abstract Sensation seeking is a personality trait that is robustly correlated with delinquent behavior in adolescence. The current study tested specific contextual factors hypothesized to facilitate, exacerbate or attenuate this risk factor for adolescent delinquency. Individual differences in sensation seeking, peer deviance, parental monitoring and self-reported delinquent behavior were assessed in a sample of 470 adolescents. Peer deviance partially mediated the effects of sensation seeking and parental monitoring on adolescent delinquency. We also found evidence for a three-way interaction between sensation seeking, peer deviance and parental monitoring, such that the highest rates of delinquency occurred from the concurrence of high sensation seeking, high peer deviance, and low levels of parental monitoring. Results highlight the importance of considering peer- and family-level processes when evaluating personality risk and problematic adolescent behavior.

. Introduction Sensation seeking, defined as a disposition to select and prefer novel, stimulating, or exciting experiences, is an intrapersonal risk factor for delinquent behavior (Harden et al., 2012, Harden and Tucker-Drob, 2011 and Popham et al., 2011). Population-average developmental increases in sensation seeking and delinquent behavior co-occur across adolescence (Moffitt, 1993 and Steinberg et al., 2008), and individual differences in longitudinal changes in sensation seeking account for much of the adolescent spike in delinquent behavior (Harden et al., 2012). As personality risk for adolescent delinquency, sensation seeking may index a “reaction range” for the emergence of delinquency (Nigg, 2006), with environmental contexts possibly mediating and/or moderating this risk. Researchers have therefore begun to examine specific contextual factors that facilitate, exacerbate or attenuate personality risk for delinquent behavior. In the current paper, we consider the relations between sensation seeking and two social contexts: deviant peers and parental monitoring. Peer deviance is a robust contextual correlate of adolescents’ delinquent behavior (Kandel, 1986), an association that reflects social selection and social influence ( Burk et al., 2011 and Wills and Cleary, 1999). Social selection is a process by which adolescents with dispositions toward delinquency select (and are selected into) deviant peer groups ( Gottfredson and Hirschi, 1990 and Kandel, 1978). Social influence occurs when befriending and socializing with deviant peers increases one’s likelihood to engage in delinquent behavior. Sensation seeking may play both mediating and moderating roles in these peer dynamics. As a mediator, sensation seeking may be a psychological mechanism of social selection, shaping who an adolescent’s friends are. For instance, affiliating with deviant peers may be one way that adolescents high in sensation seeking find a social–ecological niche that is conducive to their motivational and behavioral dispositions. Consistent with this hypothesis, affiliation with deviant peers has been found to mediate the link between sensation seeking and marijuana use (Hampson et al., 2008 and Yanovitzky, 2005). As a moderator, sensation seeking may also play a role in social influence, affecting how an individual responds to peer influence. High sensation seekers may be more responsive to the immediate rewards of peer interaction and approval and thus more vulnerable to deviant social influence. Consistent with moderating relations between personality and contextual risk, behavior genetic research has found evidence for gene × peer deviance interaction effects, whereby genetic risks on substance use are exacerbated among adolescents with deviant peers (Harden, Hill, Turkheimer, & Emery, 2008). Although the specific genetic vulnerabilities underlying these effects are unknown, other research has shown that sensation seeking is a heritable personality trait (Koopmans, Boomsma, Heath, & van Doornen, 1995) that partly accounts for heritable variation in adolescent delinquency (Harden et al., 2012). These findings suggest the effects of peer groups on delinquent behavior may be intensified when genetic risk for delinquency—including risk conferred by high sensation-seeking—is present. Finally, the negative effects of sensation seeking on adolescent delinquency may wane in protective environments. Parental monitoring, defined by Dishion and McMahon (1998, p. 61) as “parenting behaviors involving attention to and tracking of the child’s whereabouts, activities, and adaptations,” is a protective factor that may mitigate the deleterious effects of various risks on adolescent behavior (Hill and Tyson, 2009 and Lac and Crano, 2009). From the perspective of social control theory (Hirschi, 1969), antisocial behavior is prevented by bonds to conventional society, including parents. Parental monitoring, by both constraining certain behaviors and by communicating awareness and caring about adolescents’ activities, may function as a key mechanism of social control (Longmore, Manning, & Giordano, 2013). Specifically, parental monitoring may buffer the negative effects of high sensation seeking by preventing adolescents’ affiliation with deviant peers and by limiting the influence of those peers ( Kiesner et al., 2010 and Steinberg et al., 1994). In a large sample of adolescents, lower levels of peer deviance mediated the protective effect of parental monitoring on alcohol use ( Kim & Neff, 2010). Moreover, a study with late adolescents found that the relation between peer influence and drinking behavior was moderated by parental monitoring ( Wood, Read, Mitchell, & Brand, 2004). Finally, molecular genetics research has found evidence of a gene × parental monitoring interaction, whereby genetic risks for externalizing behavior decrease under high levels of parental monitoring ( Dick et al., 2009 and Dick et al., 2011). 1.1. Goals of the current study Building off previous research, we test five hypotheses in this study. First, high levels of sensation seeking and peer deviance and low levels of parental monitoring will independently predict adolescent delinquency. Second, peer deviance will partially mediate the effect of sensation seeking on delinquent behavior, such that adolescents high in sensation seeking will select deviant peer groups and, in turn, increase risk for delinquency. Third, peer deviance will also moderate the association between sensation seeking and delinquency, such that adolescents high in sensation seeking will be more vulnerable to the influence of deviant peers. Fourth, peer deviance will mediate the protective effect of parental monitoring on delinquent behavior, such that high levels of parental monitoring will prevent adolescents from affiliating with deviant peers and, in turn, prevent exposure to contextual risk for delinquency. Fifth, the protective effects of parental monitoring will be highest for youth high in both intra- and inter-personal risk. Therefore, we hypothesize that parental monitoring will moderate the combined influence that sensation seeking and peer deviance has on delinquency, such that a three-way interaction between sensation seeking, peer deviance and parental monitoring will be observed. Although many of the individual pieces of this model have been tested in previous research, this study is the first to test a comprehensive model that includes a three-way interaction between sensation seeking, peer deviance and parental monitoring.

3. Results Table 1 summarizes descriptive and reliability statistics and zero-order correlations among study variables. Results from the full model are illustrated in Fig. 1. Note that because focal predictors were standardized, the main effects can be interpreted as population-average effects, i.e., effects holding all moderators at their mean levels. In support of our first hypothesis, sensation seeking (b = .347, 95% CI = .264, .430, p < .001), peer deviance (b = .375, 95% CI = .305, .445, p < .001) and parental monitoring (b = -.115, 95% CI = −.195, −.035, p < .01) had significant main effects on adolescent delinquency, even after controlling for age, gender and parental education. Second, high sensation seeking adolescents (b = .194, 95% CI = .111, .278, p < .001) and adolescents with lower levels of parental monitoring (b = −.334, 95% CI = −.425, −.215, p < .001) reported higher levels of peer deviance. Tests of indirect effects (see Table 2) indicated that affiliation with deviant peers partially mediated the association between sensation seeking and delinquency (b = .073 95% CI = .039, .106, p < .001), as well as parental monitoring and delinquency (b = −.125, 95% CI = −.168, −.083, p < .001). That is, these results indicate that delinquency increases by approximately .07 standard deviations for every standard deviation increase in sensation seeking via the effect of sensation seeking on increased peer deviance. Likewise, adolescent delinquency is predicted to decrease by approximately .13 standard deviations for every standard deviation increase in parental monitoring via the effect of parental monitoring on reduced peer deviance. Table 1. Zero-order correlations, descriptive & reliability statistics. N = 470 α M (SD) R Sex SS PD PM PE DEL Age 15.545 (1.185) 13.57–17.99 −.033 .127∗ .137∗ −.203∗∗ .047 .262∗∗ Sex 0.470 (0.500) 0 = M, 1 = F −.044 −.145∗ .151∗ −.046 −.164∗ SS .718 3.179 (0.693) 1.13–5.00 .318∗∗ −.270∗∗ −.165∗ .457∗∗ PD .867 1.732 (0.334) 1.05–2.91 −.415∗∗ −.233∗∗ .495∗∗ PM .801 2.624 (0.286) 1.07–3.00 .153∗ −.409∗∗ PE .808 17.247 (2.879) 6.00–22.0 −.086 DEL .877 6.432 (7.303) 0.00–50.0 Note: Descriptive statistics for untransformed variables & correlations for transformed variables are reported. α = Cronbach’s alpha. M = mean. (SD) = standard deviation. R = range. SS = sensation seeking. PD = peer deviance. PM = parental monitoring. PE = average of parental education. DEL = delinquent behavior. ∗p(two-tailed) < .01. ∗∗p(two-tailed) < .001. Table options Table 2. Total, direct & indirect effects on delinquency. Total Direct Indirect B S.E. B S.E. B S.E. Sensation seeking .420∗∗ (.044) .347∗∗ (.042) .073∗∗ (.017) Parental monitoring −.240∗∗ (.040) −.115∗ (.041) −.125∗∗ (.022) Note: Mediator variable = peer deviance. B = unstandardized path coefficients for standardized variables. S.E. = standard error. ∗p(two-tailed) < .01. ∗∗p(two-tailed) < .001. Table options In support of our moderation hypotheses, there were significant two-way interactions between sensation seeking and peer deviance (b = .122, 95% CI = .066, .179, p < .001), sensation seeking and parental monitoring (b = −.068, 95% CI = −.120, −.016, p < .01), and peer deviance and parental monitoring (b = −.072, 95% CI = −.115, −.030, p < .01). There was also a significant three-way interaction between sensation seeking, peer deviance and parental monitoring (b = −.123, 95% CI = −.182, −.064, p < .001), such that the association between sensation seeking and delinquency was magnified among adolescents who socialized with deviant peers and who were low in parental monitoring. See Fig. 2 for a plot of the simple slopes from the three-way interaction. Sensation seeking×peer deviance×parental monitoring interaction on delinquency. ... Fig. 2. Sensation seeking × peer deviance × parental monitoring interaction on delinquency. Note: Simple slopes calculated from parameters shown in Fig. 1. Predicted delinquency displayed for low (−1σ), average and high (+1σ) peer deviance. Panels present sensation seeking × peer deviance interaction across high (+1σ), average and low (−1σ) parental monitoring. Figure options Note that, controlling for all other predictors, older adolescents were monitored less by parents (b = −.165, 95% CI = −.241, −.090, p < .001), and reported higher levels of sensation seeking (b = .111, 95% CI = .029, .193, p < .01) and delinquent behavior (b = .173, 95% CI = .108, .257, p < .001). Boys reported more delinquent behavior than girls (b = −.197, 95% CI = −.358, −.037, p < .05) and were also more likely to report deviant peer affiliation (b = −.191, 95% CI = −.380, −.001, p < .05). Conversely, girls were more likely to be monitored by their parents (b = .290, 95% CI = .116, .464, p < .01). In sum, covariates and mediating and moderating pathways among focal predictors accounted for more than a third of the variance in adolescents’ self-report delinquent behavior (R2 = .35).