شخصیت و ارتباط عصب روانشناختی رفتار زورگویی

The psychological and neuropsychological correlates of bullying behavior were examined in a group of 41 middle school students (age range 11–15 years) and group-matched controls. The students were identified as bullies by school administrators, their teachers, and self-ratings. Parents of children in both groups completed the Coolidge Personality and Neuropsychological Inventory, a 200-item, DSM-IV-TR aligned, parent-as-respondent, standardized measure. It was found that bullying behavior was associated more with DSM-IV-TR Axis I diagnoses of conduct disorder, oppositional defiant disorder, attention-deficit/hyperactivity disorder, and depressive disorder than in matched controls. Bullying behavior was also correlated more with Axis II diagnoses of passive–aggressive, histrionic, paranoid, and dependent personality disorders than in matched controls. Bullying behavior was also more correlated with measures of neuropsychological dysfunction and executive function deficits. An implication of these findings is that traditional short-term psychotherapeutic interventions for bullying behavior may be of limited value given the complex nature of the associated psychopathology.

In a variety of forms, school violence pervades American society. Various methods of research have been undertaken in an effort to pinpoint the etiology of school violence, particularly bullying behavior in schools. Definitions of bullying behavior incorporate such factors as an evident power differential, physical and/or verbal abuse, and severity and duration of abuse (Atlas & Pepler, 1998). Olweus (1991) defines bullying and victimization as the exposure of an individual, repeatedly and over time, to negative actions on the part of one or more others. Bullying can take physical forms, such as hitting, pushing, kicking, or punching, and/or verbal forms, exemplified in threatening, teasing, taunting, and name calling.

The validity of the diagnosis of the two groups (bullying and controls) was established by performing t tests for independent samples on the sum of the Weinhold's teacher's measure and student's measure of bullying behavior. The mean teacher's rating of bullying behavior for the bullying group (M=33.2, SD=14.9) was significantly higher than the control groups' mean (M=5.6, SD=4.6), Welch's t (44.9)=11.06, p<0.0005. The mean student's self-ratings of bullying behavior for the bullying group (M=14.2, SD=10.2) was also significantly higher than the control groups' mean (M=5.4, SD=7.0), Welch's t (67.3)=4.43, p<0.0005. Together, these results show that the referred bullies were rated by their teachers and self-rated higher on bullying scales than the controls. Thus, these findings help to establish the validity of the bullying diagnosis. 3.1. Clinical (Axis I) scales A multivariate analysis of variance (MANOVA) was performed on the five Axis I scales of the CPNI for the main effect of group (bully and controls). The MANOVA was significant, approximate F(5,76)=4.01, p=0.003. Post hoc univariate analyses of variance (ANOVA) with a modified Bonferroni correction ( Holm, 1979) revealed that the Conduct Disorder, Oppositional Defiant Disorder, ADHD, and Depressive Disorder scales produced a significant main effect for the bullying diagnosis. One scale (Overanxious Disorder of Childhood) was not significant (see Table 1). All three Axis I hypotheses were supported. The bullying group produced significantly higher means than the controls for Conduct Disorder, Oppositional Disorder, and ADHD scales, and all had large effect sizes. Inspection of the bullying group revealed that 46%, 49%, and 51% of the individuals were clinically elevated (T scores ⩾ 60) for the Conduct Disorder, Oppositional Disorder, and ADHD scales, respectively. Although not hypothesized, the Depressive Disorder scale was significantly and clinically elevated (T score ⩾60) in the bullying group, and it also had a large effect size. Approximately 49% of the children in the bullying group exhibited a clinical elevation on the Depressive Disorder scale. Table 1. Means, T scores, t values, and correlation of effect size for bullies and non-bullies on the CPNI T scores t Sig. r∗∗ Bullies (SD) Non-bullies (SD) Axis I ADHD 60.4 (10.9) 51.9 (10.9) 3.64 0.001∗ 0.38 Conduct Dis. 59.9 (15.4) 49.6 (10.5) 3.53 0.001∗ 0.37 Oppos. Defiant Dis. 60.6 (13.0) 51.9 (11.6) 3.18 0.002∗ 0.33 Major Depress. Dis. 62.4 (15.6) 54.3 (13.1) 2.55 0.013∗ 0.27 Overanxious Dis. 50.1 (9.8) 51.4 (11.4) −0.59 0.558 0.06 Axis II Passive–aggressive PD 60.9 (13.7) 50.1 (12.0) 3.80 0.001∗ 0.42 Paranoid PD 60.1 (10.9) 54.5 (11.1) 2.30 0.024 0.26 Histrionic PD 58.4 (13.4) 51.9 (12.7) 2.25 0.027 0.28 Dependent PD 54.0 (11.7) 49.1 (10.3) 2.01 0.048 0.25 Borderline PD 58.8 (13.1) 53.8 (11.7) 1.84 0.069 0.22 Depressive PD 55.5 (12.6) 51.9 (10.1) 1.41 0.163 0.19 Schizoid PD 56.2 (11.7) 53.7 (11.5) 0.96 0.338 0.15 Narcissistic PD 53.3 (12.6) 51.1 (10.5) 0.89 0.379 0.15 Schizotypal PD 56.2 (11.7) 52.0 (11.6) 0.79 0.432 0.14 Avoidant PD 53.1 (13.4) 53.7 (13.5) −0.19 0.847 0.00 Obs.–comp. PD 51.8 (9.7) 52.1 (9.6) −0.14 0.890 0.02 Neuropsychological scales Execut. Funct. Def. 61.9 (12.7) 52.2 (11.3) 3.61 0.001∗ 0.41 Neuropsych. Dysf. 63.1 (12.2) 54.0 (12.2) 3.39 0.001∗ 0.39 Mild Neurocog. Dis. 62.5 (12.2) 53.6 (11.8) 3.37 0.001∗ 0.39 Other clinical scales Dangerousness 60.9 (13.6) 51.8 (12.9) 3.13 0.002∗ 0.32 Aggression 57.6 (12.2) 50.0 (11.3) 2.93 0.004∗ 0.33 Disinhibition 55.8 (11.8) 49.4 (9.8) 2.66 0.009∗ 0.30 Emotional Lability 58.1 (11.7) 516 (11.5) 2.52 0.014∗ 0.31 Note: * Significant according to modified Bonferroni correction. ** r=correlation of effect size; small=0.100, medium=0.243, large=0.371. Table options 3.2. Personality disorder (Axis II) scales A MANOVA was also performed on the 11 Axis II personality disorder scales. The MANOVA was again significant, approximate F(11,70)=2.58, p=0.008. Post hoc ANOVA's (with the modified Bonferroni correction) revealed that only the Passive–Aggressive scale was significant. The bullying group was significantly and clinically elevated for the Passive–Aggressive scale with a large effect size. The bullying group was also elevated compared to the control group on the Histrionic, Paranoid, and Dependent scales, and each had a large effect size but the difference did not reach statistical significance with the modified correction (see Table 1). It was hypothesized, in general, that some personality disorders would be more prevalent in the bullying group compared to controls, and this hypothesis was partially supported. However, the finding that the bullying group was elevated (compared to controls) with a large effect size on the Histrionic scale was somewhat surprising. Therefore, a post hoc t test item analysis was conducted for the eight items on the scale. Interestingly, the item representing Criterion 7 in DSM-IV-TR “is suggestible, i.e., easily influenced by others or circumstances” produced the highest t value of the eight items, and it had a large effect size. 3.3. Neuropsychological scales The third hypothesis that the bullying group would also have greater neuropsychological behavioral dysfunction was also supported. A MANOVA was performed on the three neuropsychological scales of the CPNI, and it was again significant, approximate F(3,78)=4.49, p=0.006. Post hoc ANOVA's (with modified Bonferroni) revealed that the Executive Function Deficits, the General Neuropsychological Dysfunction, and Mild Neurocognitive scales produced significant group main effects. The bullying group was significantly and clinically elevated on all three scales with large effect sizes (see Table 1). Post hoc t test analyses were also performed on the three subscales of the Executive Function Deficits scale of the CPNI between the bullying and control groups. The three subscales, derived through previous factor analyses, measure (a) decision-making, planning, and organizational problems, (b) metacognitive dysfunctions such as problems with learning, reading, memory, and concentration, and (c) social misjudgments including poor interpersonal decision-making and choices. All three subscales were significantly higher and clinically elevated for the bullying group. 3.4. Other clinical scales The fourth hypothesis that the bullying group would be elevated on the Dangerousness, Aggression, Emotional Lability, and Disinhibition scales of the CPNI was also supported. A MANOVA was performed on the four clinical scales, and it was significant, F(4,77)=2.99, p=0.024. Post hoc ANOVA's revealed that the bullying group was significantly elevated on all four scales with large effect sizes (see Table 1).