خودتنظیمی عاطفی، طرد همسالان و رفتار ضد اجتماعی: ارتباط تکاملی از اوایل کودکی تا اوایل نوجوانی

Abstract This study examined relations among emotional self-regulation, peer rejection, and antisocial behavior in a sample of 122 boys from low-income families who participated in a summer camp and were followed longitudinally from early childhood to early adolescence. Emotional self-regulation strategies were coded in early childhood from a waiting task, measures of peer rejection were collected during middle childhood at the summer camp, and reports of antisocial behavior were obtained during early adolescence. Structural equation modeling was utilized to examine longitudinal relations among these constructs, with results supporting a negative association between use of active distraction and peer rejection and a positive association between peer rejection and antisocial behavior. Furthermore, an indirect effect of active distraction on antisocial behavior was found through peer rejection. Thus, adaptive self-regulation strategy use in early childhood demonstrated direct longitudinal relations with peer rejection and an indirect association with antisocial behavior in early adolescence. Results have implications for early prevention and intervention efforts to foster adaptive self-regulation of emotion and reduce risk for later social problems and delinquency.

. Introduction Aggression and other forms of overt externalizing symptoms reach their peak between ages two and three. However, a small minority of children continue to show high levels of disruptive behavior problems across childhood (Shaw, Gilliom, Ingoldsby, & Nagin, 2003), and early conduct problems are associated with delinquency and mental health problems in adolescence (Moffitt et al., 2002 and Shaw and Gross, in press). Due in part to the personal, economic, and social toll that delinquency takes on individuals and society, predictors of antisocial behavior have received extensive examination. Perspectives on early-starting conduct problems have emphasized the interplay of child temperamental factors and context in the emergence and maintenance of these problems across childhood (Campbell et al., 2000 and Shaw et al., 2000). In line with a focus on individual and social mechanisms in the persistence of early-starting conduct problems from early childhood to early adolescence, the present study examined emotional self-regulation and peer rejection as precursors to early adolescent antisocial behavior. The present study was also informed by a developmental psychopathology perspective on sensitive periods in development, which emphasizes that each stage of child development presents key tasks and challenges. A developmental psychopathology perspective suggests that deviations from normative processes at earlier stages of development increase the likelihood of psychopathology later in development (Sroufe, 1997). In early childhood, one key challenge is the attainment of self-regulation of emotion (Kopp, 1989), and difficulty with adaptive emotional self-regulation may portend later social and behavioral problems (e.g., Gilliom, Shaw, Beck, Schonberg, & Lukon, 2002). In middle childhood, peer inclusion becomes a critical element of positive adaptation (Rose-Krasnor, 1997), and peer rejection indicates poor social adaptation and risk for antisocial behavior (e.g., Laird, Jordan, Dodge, Pettit, & Bates, 2001). Previous research has not simultaneously examined the influences of emotional self-regulation and peer rejection during developmental periods when normative deviations may be particularly salient for later adaptation. In the present study, we examined whether self-regulation strategies in early childhood predicted peer rejection in middle childhood and how these constructs were associated with antisocial behavior in early adolescence. Longitudinal data allowed examination of the central hypothesis that peer rejection would account for indirect relations between emotional self-regulation strategies in early childhood and antisocial behavior in early adolescence. 1.1. Emotional self-regulation in early childhood Emotion regulation is a multi-faceted construct without a single, widely-accepted definition (Cole, Martin, & Dennis, 2004). At a broad level, emotion-related regulation can include attentional, cognitive, or behavioral attempts to manage internal states or the external expression of emotion (Eisenberg, Spinrad, & Smith, 2004). Because emotion regulation is a multi-faceted construct, numerous approaches exist to examine emotion regulatory processes, ranging from studies of the reflexive regulation of distress in infancy to the analysis of emotion dynamics in interpersonal interactions. An increasingly common approach to investigate emotion regulation involves structured observation of “self-initiated attempts to modulate negative emotion” (Cole et al., 2004, p. 325). This method to examine emotion regulation is particularly relevant during early childhood because it reflects the developmental importance and rapid growth of multiple levels of self-regulatory competence during the toddler and preschool periods (Calkins & Fox, 2002). For example, self-regulatory attempts to focus, shift, or inhibit attention, behavior, and emotion increase during the toddler period when aggressive outbursts peak and caregivers are frequently needed to calm distressed toddlers (Kochanska et al., 2000, Kopp, 1989 and Shaw et al., 2000). As children move into the preschool years, a growing proficiency with effortful control promotes increased use of adaptive self-regulation strategies and fewer behavioral problems (Kochanska et al., 2000). When confronted with a distressing event, young preschoolers are often able to actively distract themselves from distressing stimuli or focus on more pleasant aspects of the situation (Denham, 1998). However, children who are unable to master adaptive strategies for emotional self-regulation during the preschool period demonstrate numerous problematic outcomes, including impaired social competence and externalizing problems (e.g., Denham et al., 2003 and Gilliom et al., 2002). In line with an operational definition of emotion regulation that focuses on self-regulatory attempts to manage negative emotion, the present study examined behavioral strategies for regulating emotion in the context of a frustrating situation. Previous research has focused on specific self-regulation strategies that may be more or less adaptive in the immediate context and in relation to later adaptation (e.g., Silk, Shaw, Skuban, Oland, & Kovacs, 2006). For example, the ability to utilize self-regulatory strategies to delay gratification during a waiting task in preschool predicted social and academic competence during adolescence (Mischel, Shoda, & Peake, 1988). Specific emotional self-regulatory strategies may have unique implications for externalizing behaviors, including early-starting conduct problems that emerge in childhood (Calkins and Howse, 2004 and Dishion and Patterson, 2006). In an earlier report using data from the present study, a tendency to focus on the desired object and less use of self-regulatory distraction during a frustrating waiting task at age 3.5 were associated with teacher reports of externalizing problems 3 years later (Gilliom et al., 2002). No prior research has examined specific self-regulation strategies in early childhood as predictors of antisocial behavior during early adolescence. However, a broad observational composite of self-control that included ratings of emotional reactivity and regulation at age 3 was related to antisocial behavior in adolescence and differentiated life-course-persistent from adolescent-limited antisocial behavior in the Dunedin study (Caspi et al., 1995 and Moffitt and Caspi, 2001). 1.2. Peer rejection in middle childhood The role that school-age peer relationships play in pathways from emotional self-regulation in early childhood to antisocial behavior in adolescence also has yet to be elucidated. In particular, characteristics of peer relationships may account for indirect relations between earlier emotional self-regulation and later antisocial behavior, a hypothesis that is best examined with a longitudinal study. Youngsters typically develop their first friendships during the toddler and preschool periods, but inclusion by peers takes on heightened significance during middle childhood as children increasingly make social comparisons based on shared feelings, values, and loyalty (Rose-Krasnor, 1997 and Rubin et al., 2006). Thus, a substantial portion of the empirical research on the developmental salience of peer rejection has been conducted during middle childhood. Middle childhood peer rejection predicts concurrent and later antisocial behavior (for a review, see Dodge, Coie, & Lynam, 2006), even after accounting for the effect of earlier externalizing problems (e.g., Laird et al., 2001). Because peer rejection is a consistent risk factor for negative behavioral outcomes, there is an extensive literature examining its precursors. Children's adaptive self-regulation of emotion often occurs concomitantly with positive social adaptation and also serves as a buffer against peer rejection (Halberstadt et al., 2001 and Hubbard and Dearing, 2004). Children who use more adaptive emotional self-regulation strategies in distressing situations are more likely to master the social skills necessary for effective social relationships (e.g., Fabes & Eisenberg, 1992). Conversely, children who have difficulty managing their negative emotions are more likely to become disruptive in social interactions, leading to lower acceptance and more rejection by peers (Maszk, Eisenberg, & Guthrie, 1999). In support of these predictions, concurrent relations exist between constructive forms of emotional self-regulation and sociometric status, and between less adaptive emotional self-regulation and problematic peer relations during preschool, especially among boys (Eisenberg et al., 1993). Longitudinal studies have also confirmed relations between emotional self-regulation and peer relations over short periods (e.g., Maszk et al., 1999), but these associations have not yet been examined across a span of several years from early to middle childhood. 1.3. The present study The present study was designed to investigate relations among emotional self-regulation strategies, peer rejection, and antisocial behavior in a sample of low-income boys who participated in a summer camp as part of a larger longitudinal study. The selection of emotional self-regulation and peer rejection as central constructs was guided by (a) their potential importance in the persistence of externalizing behavior problems from early childhood to early adolescence and (b) the developmental salience of these constructs in early childhood and middle childhood, respectively. We were particularly interested in the role that peer rejection plays in longitudinal relations between emotional self-regulation and antisocial behavior. Over a long period of study, deviation from developmental norms of early childhood would not necessarily directly predict maladjustment in early adolescence; rather, it might set the stage for developmental deviation leading to further maladjustment (Sroufe & Rutter, 1984). Thus, peer rejection was examined as a construct that could account for the indirect association between emotional self-regulation and antisocial behavior. Specifically, we examined whether specific emotional self-regulation strategies in early childhood were associated with rejection by peers during the summer camp in middle childhood. In turn, we examined whether peer rejection at the summer camp predicted antisocial behavior during early adolescence. To examine potential indirect relations between emotional self-regulation strategies and antisocial behavior, we determined whether the indirect pathway from emotional self-regulation to antisocial behavior through peer rejection met statistical criteria for an indirect effect. Given the long time-span between the early childhood and early adolescent assessments (7.5 to 8.5 years), the early childhood self-regulation strategies were not necessarily expected to be directly related to early adolescent antisocial behavior. However, we also examined whether direct relations were evident between these constructs. Early childhood behavior problems were controlled for in analyses examining relations among emotional self-regulation, peer rejection, and antisocial behavior to rule out the possibility that early-starting conduct problems accounted for relations among these constructs.

. Results 3.1. Analysis plan Study hypotheses were examined with structural equation models. Following inspection of means, standard deviations, and intercorrelations, structural equation models were evaluated in AMOS to examine direct and indirect relations among study constructs. Early childhood externalizing problems and emotional self-regulation strategies (active distraction and task focus) were examined as individual manifest indicators, and the multiple indicators of peer rejection (sociometric nominations and sociometric ratings) and antisocial behavior (self-, parent-, and teacher-reports) were used to create latent constructs. Overall model fit was examined with multiple fit indices including Chi-square, RMSEA, CFI, and TLI, and individual paths in the model were tested for statistical significance. Lastly, the Sobel (1982) test was used to evaluate the hypothesized indirect relation between emotional self-regulation strategies and antisocial behavior through boys' rejection by peers. 3.2. Descriptive statistics Table 1 presents descriptive statistics for the 122 boys included in the present study. The mean of 10.43 for active distraction indicates that this self-regulation strategy occurred in over half of the 18 coding intervals. The ranges for active distraction and task focus were 1 to 18 intervals and 0 to 17 intervals respectively, and the range and standard deviation for each strategy supports variability in strategy use for this sample. For the variables presented in Table 1, there were no significant group differences when comparing non-minority (i.e. European American; n = 57) to ethnic minority boys (n = 65), with p > .05 for all t-tests. Thus, ethnicity was not included in subsequent analyses. Table 1. Means and standard deviations for study variables. Variable Mean SD Active distraction 10.43 5.13 Task focus 4.75 4.61 Externalizing problems: CBCL raw score at 42 months 26.37 12.14 Peer rejection: Nominations .04 1.73 Peer rejection: Ratings .82 .46 Antisocial behavior: SRD sum score 3.43 2.74 Antisocial behavior: CBCL delinquent behavior sum score 2.18 2.30 Antisocial behavior: TRF delinquent behavior sum score 2.43 2.66 Note. CBCL = Child Behavior Checklist, SRD = Self-Reported Delinquency, TRF=Teacher Report Form. Table options Table 2 presents intercorrelations between the variables. Not surprisingly, the self-regulation strategies were negatively correlated with each other. Neither regulation strategy was significantly correlated with concurrent externalizing problems. Active distraction was negatively correlated with rejection nominations and ratings, and task focus was positively correlated with rejection nominations but not ratings. Neither regulation strategy was correlated with the indicators of antisocial behavior, except for marginally significant correlations with maternal reports of boys' antisocial behavior. There was a significant positive correlation between the nomination and rating indicators of peer rejection. Both indicators of peer rejection were positively correlated with maternal and teacher reports of antisocial behavior, and rejection ratings were positively correlated with self-reports of antisocial behavior. Robust bivariate correlations also existed between the indicators of antisocial behavior. Table 2. Intercorrelations among variables. Variable 1 2 3 4 5 6 7 1. Active distraction 2. Task Focus −.50⁎⁎ 3. Externalizing (42 months) .00 .02 4. Peer rejection: nominations −.26⁎⁎ .19⁎ .14 5. Peer rejection: ratings −.25⁎⁎ .13 .24⁎⁎ .74⁎⁎ 6. Antisocial behavior: SRD −.08 .10 .20⁎ .17a .18⁎ 7. Antisocial behavior: CBCL −.15a .16a .36⁎⁎ .28⁎⁎ .29⁎⁎ .38⁎⁎ 8. Antisocial behavior: TRF −.13 .07 .21⁎ .47⁎⁎ .47⁎⁎ .51⁎⁎ .56⁎⁎ Note. CBCL = Child Behavior Checklist, SRD = Self-Reported Delinquency, TRF = Teacher Report Form. ap < .10, *p < .05, **p < .01. Table options 3.3. Model estimation Structural equation models were examined with AMOS 5.0 (Arbuckle, 2003). All 122 boys had peer rejection and emotional self-regulation data, but a small percentage of the boys were missing a portion of the antisocial behavior data. Specifically, 2%, 2%, and 21% of the 122 boys were missing antisocial behavior data from self report, maternal report, and teacher report, respectively. These missing data were estimated using full information maximum likelihood procedures in AMOS. The model presented in Fig. 1 was created to examine developmental predictors of antisocial behavior. This model included two early childhood self-regulation strategies, active distraction and task focus, and maternal reports of externalizing problems in early childhood as exogenous variables. The model also included two measures collected during the camp, sociometric rejection nominations and sociometric rejection ratings, to create a peer rejection latent construct. Lastly, the model included self, maternal, and teacher reports of antisocial behavior to create an antisocial behavior latent construct. Model of relations among emotional self-regulation strategies, peer rejection, ... Fig. 1. Model of relations among emotional self-regulation strategies, peer rejection, and antisocial behavior. Note. Standardized path coefficients and loadings are presented in the figure. χ2(15) = 17.73, RMSEA = .04, CFI = .99, TLI = .97. f = fixed path. ⁎p < .05, ⁎⁎p < .01. Figure options A path was included from each early childhood exogenous variable to the peer rejection construct to test our hypothesis that each self-regulation strategy in early childhood would predict peer rejection while simultaneously accounting for the other strategy and early childhood externalizing problems. Paths were also included from early childhood externalizing problems and the peer rejection construct to the antisocial behavior construct to test the hypothesis that peer rejection would predict a key indicator of early adolescent maladjustment while accounting for stability in behavior problems from early childhood to early adolescence. Given the small and non-significant bivariate correlations between the self-regulation strategies and measures of antisocial behavior, direct paths were not included from the self-regulation strategies to the antisocial behavior construct. Model fit was tested with multiple indices. The Chi-square goodness of fit index tests exact model fit, and a nonsignificant Chi-square value supports model fit. There are also a number of relative fit indices. The Root Mean Square Error of Approximation (RMSEA) is one such measure of relative fit, and RMSEA values below .06 support good model fit (Hu & Bentler, 1999). Two other statistics, the Comparative Fit Index (CFI) and the Tucker–Lewis Index (TLI) measure the absolute fit of the model in comparison to the absolute fit of a null (baseline) model, and values above .95 for the CFI and TLI indicate good model fit (Hu & Bentler, 1999). The model demonstrated excellent model fit, with χ2(15) = 17.73, RMSEA = .04, CFI = .99, and TLI = .97. The standardized coefficients for paths specified in the model are presented in Fig. 1. Based on the factor loadings, the model supported peer nominations and ratings as elements of a peer rejection latent construct and self, maternal, and teacher reports as elements of an antisocial behavior latent construct. The model also supported direct paths from active distraction and early childhood externalizing problems to peer rejection and direct paths from early childhood externalizing problems and peer rejection to antisocial behavior. However, the direct path from task focus to the peer rejection construct was nonsignificant. To more closely examine the developmental relations between early childhood emotional self-regulation and antisocial behavior in early adolescence, we evaluated the indirect effect of active distraction on the antisocial behavior construct with the Sobel (1982) test (see Sobel, 1987 for a description of this test using coefficients from structural equation models). This method gives a significance statistic for the indirect effect of a predictor on an outcome through an intermediate variable. The Sobel procedure uses the unstandardized path coefficients and their standard errors for the direct effect of the predictor (active distraction) on the intermediate variable (peer rejection) and the direct effect of the intermediate variable (peer rejection) on the dependent variable (antisocial behavior) to test the significance of the indirect effect. The results of the Sobel test indicated that active distraction had a significant indirect relation with antisocial behavior, z = − 2.14, p < .05. In addition, the Sobel test indicated that early childhood externalizing problems had a significant indirect relation with antisocial behavior through peer rejection (z = 2.28, p < .05) in addition to the direct effect between these constructs.