هسته عاطفی زورگویی: شواهد بیشتر برای نقش صفات سنگدلانه - عاری از احساس و همدلی

Abstract The aim of this paper was to further explore the role of callous–unemotional traits (CU) and empathy in bullying within 529 middle school children. We tried to advance limits of previous studies by accounting for a measure of victimization and considering age-related effects. Our results indicated that in younger students (mean age = 11 years and 8 months) the uncaring dimension of CU traits were positively related to bullying, but this association was completely mediated by a lack of affective empathy; in older students (mean age = 13 years and 8 months) the callous dimension of CU traits was directly related to bullying, and empathy was not associated when taking into account CU traits. The impact of CU traits on bullying in the transition from late childhood to incoming adolescence is discussed.

1. Introduction 1.1. Bullying and emotions Bullying is a subtype of proactive aggression defined as a relation in which aggressive behaviors are repetitively carried out towards a victim unable to defend herself/himself (Olweus, 1993 and Salmivalli, 2010). Harassment displayed in bullying refers to a wide range of misdeeds (e.g., physical contact, attack to personal properties, verbal abuse, social exclusion and impairment of victim’s relationships), representing both a serious risk factor for children’s psychological well-being and a major problem for scholastic institutions in many countries as well as in Italy (Gini, 2008 and Menesini et al., 2012). To date, many psychological correlates of bullying have been identified; nevertheless, a number of questions have still to be clarified, especially in relation to bullies’ affective dimensions. According to several studies on social-cognitive models, many children performing bullying behaviors were able to accurately process social information and had more advanced skills in theory of mind compared to reactive aggressive peers, but the formers used their skills in order to manipulate social environment and to achieve self-oriented goals (e.g., obtaining material advantages and/or a powerful position within peer group; Arsenio and Lemerise, 2001, Crick and Dodge, 1999, Salmivalli, 2010, Sutton et al., 1999a, Sutton et al., 1999b and Sutton et al., 2001). Moreover, many bullies (especially males) performed their actions without experiencing aversive vicarious emotions and without evaluating the emotional impact on victims, showing an impaired ability to share others’ affective states. Several researches have provided considerable evidence that bullies had significantly lower affective empathy than peers who did not bully (Caravita et al., 2009, Jolliffe and Farrington, 2006, Jolliffe and Farrington, 2011 and Stavrinides et al., 2010). 1.2. Bullying and callous–unemotional traits Some recent studies considered the role of callous–unemotional traits (CU, i.e., lack of guilt, lack of empathy, poor affect, use of others for personal gain) on bullying (Fanti et al., 2009, Muñoz et al., 2011, Pardini et al., 2012 and Viding et al., 2009). These personality traits designate a subtype of childhood-onset severe antisocial behaviors that are more likely to persist into adolescence and adulthood (Frick and White, 2008 and Pardini and Frick, 2013). The timely interest for the assessment of CU traits in juvenile samples resulted in the development of the Inventory of Callous–Unemotional Traits (ICU; Frick, 2004), and suggested the presence of three specific CU dimensions: callousness (i.e., lack of empathy, guilt and remorse for misdeeds), uncaring (i.e., lack of care about ones performance in tasks and for the feelings of other people), and unemotional (i.e., deficient emotional affect). Both the callousness and the uncaring dimensions were found to be related with antisocial, aggressive, and delinquent behaviors, while the uncaring and the unemotional scales showed the strongest negative association with empathy (Essau et al., 2006, Kimonis et al., 2008 and Roose et al., 2010). As for bullying, a general score of CU traits was positively related to direct bullying in British preadolescents (Viding et al., 2009) and to bullying behaviors in American preadolescent girls with high conduct disorder symptoms (Pardini et al., 2012). Other studies examined the role of different components of CU traits (i.e. callousness, uncaring and unemotional) in bullying. Fanti et al. (2009) found in 12–18 years old Greek Cypriot students exhibiting pure bullying scored higher on the uncaring dimension compared to those exibiting low bullying, pure victimization, or a combination of bullying and victimization. A recent contribution of Muñoz and colleagues (2011) investigated the additive role of CU traits and other affective deficits in exploring bullying behaviors within a sample of British students aged 11–12 years. Results suggested that both cognitive and affective empathy were associated to bullying, but these associations did not emerge when the uncaring dimension of CU traits was taken into account. Authors concluded that the personality trait related to not caring about others was more important than empathic disposition in predicting bullying behaviors. 1.3. The present study The present study was developed in order to further explore the role played by empathy and CU traits dimensions in their association with bullying, addressing some limitations of previous studies. First, given evidence of co-occurrence in the involvement in bullying and victimization (4–6% of the children are both bullies and victims; Salmivalli, 2010), we aimed to determine which characteristics were uniquely associated with bullying, controlling for the level of victimization. Moreover, the transition from late childhood into adolescence is a period of rapid biological, psychological and relational changes, that seem to have implications for the development of personality traits (Soto, John, Gosling, & Potter, 2011). Considering a life-span developmental perspective, we emphasized the importance of age-related changes in the considered variables, and we aimed to explore whether the impact of CU traits on bullying could be different at ending childhood and incoming adolescence. As suggested by Pardini et al. (2012), it is possible that CU traits become more pronounced during adolescence as these personality characteristics become more stable and solidified. Considering that, we aimed to advance past research comparing middle school students in grade 6th and in grade 8th; middle school in Italy consists of three grades (6th to 8th) and matches with the transition from late childhood into incoming adolescence: grade 6th is usually attended by students aged 11–12, and grade 8th by students aged 13–14. Based on reported literature, we predicted that bullying would be more likely to show stronger and positive association with CU traits (especially callousness and uncaring), rather than with empathy. Moreover, we expected that this association would be more prominent in 8th grade students compared to 6th grade ones, due to the fact that CU traits in older children might be more stable in their personality structure, and able to play a stronger role in determining bullying behaviors.

3. Results 3.1. Gender and grade differences of study variables Three 2 × 2 (gender x grade) MANOVAs were performed to explore gender and grade differences for bullying scales, ICU subscales and empathy scores. As for bullying and victimization, main effects of gender (Pillai’s Trace = .08; F (2, 524) = 23.74; η2 = .08; p < .001), grade (Pillai’s Trace = .10; F (2, 524) = 29.26; η2 = .10; p < .001) and gender × grade (Pillai’s Trace = .04; F (2, 524) = 11.33; η2 = .04; p < .001) emerged; in order to qualify the above mentioned interaction term, the MANOVA was re-performed splitting by grade. Main effects of gender emerged in both younger (Pillai’s Trace = .02; F (2, 269) = 3.31; η2 = .02; p < .05) and older students (Pillai’s Trace = .17; F (2, 254) = 26.81; η2 = .17; p < .001): boys in grade 6th presented higher scores on bullying (M = −.91, SD = .13) than females (M = −.94, SD = .09), F (1, 271) = 6.53; η2 = .02; p < .05; similar evidence emerged for students in grade 8th (boys (M = −.80, SD = .14), girls (M = −.90, SD = .11), F (1, 256) = 36.26; η2 = .12; p < .001). No gender differences emerged for victimization (boys (M = −.83, SD = .17), girls (M = −.86, SD = .15), F (1, 271) = 2.41; η2 = .01; p = .12 for younger students; boys (M = −.85, SD = .15), girls (M = −.82, SD = .14), F (1, 271) = 1.56; η2 = .01; p = .21 for older students. Considering the three ICU subscales, main effects of gender (Pillai’s Trace = .09; F (3, 522) = 17.74; η2 = .09; p < .001) and grade (Pillai’s Trace = .02; F (3, 522) = 3.60; η2 = .02; p < .05) emerged: boys showed higher scores than girls on all scales (callousness: boys (M = .65, SD = .42), girls (M = .43, SD = .36), F (1, 527) = 39.18; η2 = .06; p < .001; uncaring: boys (M = .87, SD = .50), girls (M = .64, SD = .40), F (1,527) = 34.54; η2 = .06; p < .001; unemotional: boys (M = 1.52, SD = .60), girls (M = 1.30, SD = .61), F (1,527) = 16.51; η2 = .03; p < .001). No grade differences emerged for callousness (younger students (M = .56, SD = .41), older students (M = .50, SD = .39), F (1,527) = 2.95; η2 = .01; p = .09), uncaring (younger students (M = .72, SD = .49), older students (M = .78, SD = .44), F (1,527) = 2.71; η2 = .01; p = .10) and unemotional (younger students (M = 1.41, SD = .64), older students (M = 1.39, SD = .59), F (1,527) = .15; η2 = .00; p = .70). As for affective and cognitive empathy, main effects of gender (Pillai’s Trace = .24; F (2, 524) = 83.16; η2 = .24; p < .001) and grade (Pillai’s Trace = .01; F (2, 524) = 3.67; η2 = .01; p < .05) emerged: girls showed higher scores than boys on affective empathy (boys (M = 2.11, SD = .58), girls (M = 2.72, SD = .53), F (1, 528) = 159.24; η2 = .23; p < .001), but no gender difference was found for cognitive empathy (boys (M = 2.70, SD = .61), girls (M = 2.80, SD = .53), F (1, 528) = 3.71; η2 = .01; p = .06); students enrolled in grade 8th showed higher level of cognitive empathy than students in grade 6th (younger students (M = 2.69, SD = .60), older students (M = 2.82, SD = .53), F (1, 528) = 7.13; η2 = .01; p < .01) whereas no grade difference emerged considering affective empathy (younger students (M = 2.41, SD = .66), older students (M = 2.45, SD = .60), F (1, 528) = .24; η2 = .00; p = .62). 3.2. Correlation analyses The zero-order correlations coefficients are reported in Table 2. In younger students bullying was strongly and positively related to victimization (r = .50, p < .001); moreover, bullying was positively related to all ICU subscales (callousness: r = .20, p < .001; uncaring: r = .23, p < .001; unemotional: r = .13; p < .05), and negatively to affective empathy (r = −.23, p < .001). In older students we found that bullying was positively correlated with victimization (r = .37, p < .001) and both callousness (r = .40, p < .001) and uncaring (r = .35, p < .001) dimensions, and it was negatively correlated with both affective (r = −.23, p < .001) and cognitive (r = −.13, p < .05) empathy. Table 2. Correlations (Pearson’s r) between study variables (6th grade is over main diagonal). 1 2 3 4 5 6 7 1 - Bullying − .50⁎⁎⁎ .20⁎⁎⁎ .23⁎⁎⁎ .13∗ −.23⁎⁎⁎ −.05 2 - Victimization .37⁎⁎⁎ − .14⁎ .10 .06 −.01 −.02 3 - ICU callousness .40⁎⁎⁎ .10 − .45⁎⁎⁎ .33⁎⁎⁎ −.33⁎⁎⁎ −.09 4 - ICU uncaring .35⁎⁎⁎ .15⁎ .57⁎⁎⁎ − .31⁎⁎⁎ −.37⁎⁎⁎ −.23⁎⁎⁎ 5 - ICU unemotional −.01 −.01 .31⁎⁎⁎ .35⁎⁎⁎ – −.27⁎⁎⁎ −.10 6 - Affective empathy −.23⁎⁎⁎ .05 −.37⁎⁎⁎ −.40⁎⁎⁎ −.27⁎⁎⁎ – .38⁎⁎⁎ 7 - Cognitive empathy −.13⁎ −.01 −.16⁎ −.14⁎ −.15⁎ .32⁎⁎⁎ – ∗∗p < .01. ⁎ p < .05. ⁎⁎⁎ p < .001. Table options 3.3. Results of regression analyses Two hierarchical regressions for grade 6th and grade 8th were separately performed. Bullying scale was regressed onto gender and victimization on step 1. According to Muñoz and colleagues (2011), the ICU dimensions that were previously found to be correlated with bullying were entered on step 2, and similarly we considered empathy on step 3. As for younger students (Table 3), victimization was found positively predictive of bullying (β = .49, p < .001); with the addition of ICU subscales, uncaring turned out to be positively related to bullying (β = .13, p < .05). The final step showed a negative association with affective empathy (β = −.18, p < .01), while uncaring reduced to non significance. Table 3. Regression analysis for bullying in grade 6th. B SE β ΔR2 Step 1 .26⁎⁎⁎ Gender .02 .01 .10 Victimization .32 .04 .49⁎⁎⁎ Step 2 .03⁎ Gender .01 .01 .05 Victimization .31 .04 .47⁎⁎⁎ ICU callousness .02 .02 .06 ICU uncaring .03 .01 .13⁎ ICU unemotional .01 .01 .03 Step 3 .02⁎⁎ Gender .00 .01 −.01 Victimization .32 .03 .48⁎⁎⁎ ICU callousness .01 .02 .03 ICU uncaring .02 .01 .10 ICU unemotional .00 .01 .02 Affective empathy −.03 .01 −.18⁎⁎ Bullying model: R2 = .31, F(6,270) = 19.78, p < .001. Step1: ΔR2 = .26, ΔF(2,268) = 49.90, p < .001. Step2: ΔR2 = .03, ΔF(3,265) = 3.63, p < .05. Step3: ΔR2 = .02, ΔF(1,264) = 8.26, p < .01. ⁎ p < .05. ⁎⁎ p < .01. ⁎⁎⁎ p < .001. Table options As for older students (Table 4), male gender (β = .38, p < .001) and victimization (β = .40, p < .001) were found positively predictive of bullying. The addition of ICU subscales revealed the positive contribution of callousness (β = .23, p < .001) in predicting bullying. Step 3 resulted to be not significant, adding no variance. Table 4. Regression analysis for bullying in grade 8th. B SE β ΔR2 Step 1 .28⁎⁎⁎ Gender .10 .01 .38⁎⁎⁎ Victimization .36 .05 .40⁎⁎⁎ Step 2 .08⁎⁎⁎ Gender .08 .01 .30⁎⁎⁎ Victimization .32 .05 .35⁎⁎⁎ ICU callousness .08 .02 .23⁎⁎⁎ ICU uncaring .03 .02 .10 Step 3 .00 Gender .09 .02 .32⁎⁎⁎ Victimization .31 .05 .35⁎⁎⁎ ICU callousness .08 .02 .22⁎⁎⁎ ICU uncaring .03 .02 .11 Affective empathy .01 .01 .06 Cognitive empathy −.02 .01 −.08 Bullying model: R2 = .36, F(6,256) = 23.80, p < .001. Step1: ΔR2 = .28, ΔF(2,254) = 49.59, p < .001. Step2: ΔR2 = .08, ΔF(2,252) = 14.93, p < .001. Step3: ΔR2 = .00, ΔF(2,250) = 1.28, p = .28. ∗p < .05. ∗∗p < .01. ⁎⁎⁎ p < .001. Table options 3.4. Further analyses Considering results of hierarchical regression in younger students, we wondered whether affective empathy played a complete mediation effect between uncaring and bullying. A mediator can be defined as a third variable that explains how or why two other variables are related (Fairchild & McQuillin, 2010); to investigate it, we followed the procedure recommended by Holmbeck (1997). Results are reported in Fig. 1: the “uncaring predicting affective empathy” path (β = −.21, p < .001) and the “affective empathy predicting bullying” path (β = −.18, p < .01) were both negative and significant; the “uncaring predicting bullying” path (β = .13, p < .05) was positive and significant, but decreased to non-significance after the introduction of empathy. The Sobel test (2.24, p < .05) indicated that affective empathy was a complete mediator between uncaring and bullying. Mediation model: affective empathy mediates the relationship between uncaring ... Fig. 1. Mediation model: affective empathy mediates the relationship between uncaring and bullying in grade 6th. The model was controlled for gender, victimization, callousness and unemotional.