اثربخشی مداخلات و برنامه های خشم مبتنی بر مدرسه: فرا تجزیه و تحلیل

Twenty peer-reviewed journal articles that described outcomes of interventions that took place in school settings and either focused on anger or included anger as a dependent variable were meta-analyzed. No differences in outcomes were found for group comparisons by school setting, special education status, entrance criteria, or treatment agents. Although 60% of articles discussed its importance, only two articles actually measured treatment integrity. Across outcomes, the weighted mean effect size of the interventions post treatment was determined to be .31. The largest effects were found for anger and externalizing behaviors, internalizing, and social skills, with mean effect sizes of .54, .43, and .34 respectively. Weighted mean effect sizes for follow-up studies were also calculated, but given the small number of studies that reported follow-up effects, those must be interpreted with caution. The results of this meta-analysis are discussed as they relate to research, practice, and intervention with children.

School violence in the United States is a complex and escalating problem. Due to growing public safety concerns, a variety of approaches has been utilized to counter this trend. These approaches have ranged from early intervention programs designed to decrease the risk of children engaging in aggressive and violent behavior (e.g., Webster-Stratton & Reid, 2003), to skills-based intervention programs in schools (e.g., Deffenbacher, Lynch, Oetting, & Kemper, 1996), to “zero-tolerance” policies in schools for weapons and drugs, as well as for violent behavior (Skiba & Peterson, 2000 and Yell & Rozalski, 2000). One approach to decreasing school violence that has an extensive history has been to target anger as a precursor of aggressive behavior (Fryxell & Smith, 2000). Anger is a strong feeling of displeasure that includes a sense of antagonism (Merriam-Webster Inc., 1984). Anger may be both a significant concern in schools as well as an important mediator of concerns such as aggressive behavior. Anger can be an unpleasant experience for the angry person as well as those at whom the anger is directed, and can lead to negative consequences for those whose anger leads them to engage in inappropriate behaviors. Aside from the uncomfortable experience of anger at the time that it occurs, it is associated with a large number of negative outcomes for children, adolescents, and adults. These include, but are not limited to physical damage to others, themselves, and their environments; poor quality interpersonal relationships; poor quality work, school, and social experiences; anxiety and depression; drug abuse; and health problems such as hypertension and cardiovascular disease (Deffenbacher et al., 1996). Clearly, learning for all students may be in jeopardy when students' anger and its expression as aggression prevent them from attending to instruction, and when anger expressed as aggression and violence prevents access to instruction due to removal from the classroom, detention, or expulsion (Feshbach, 1983). Given the long-term consequences and the seriousness of the negative outcomes associated with anger, researchers and clinicians have designed and described a variety of intervention programs in the literature to combat it. Most programs tend to be multi-component in nature, and the large majority of anger treatment outcome studies use a cognitive-behavioral approach (CBT, Beck & Fernandez, 1998). These approaches have generally focused on Novaco's (1975) adaptation of stress inoculation training (SIT, Meichenbaum, 1975), which was originally designed for treating anxiety (Beck and Fernandez). The basic idea of training is that participants learn to identify triggers of the anger response, followed by rehearsing self-statements that will allow them to think about the situation in a way that is less likely to induce anger than their usual response. Relaxation training follows, giving participants the opportunity to learn a response that is incompatible with anger and aggression. Practice, role-play, and imagery are common characteristics of the final phase of intervention (Fernandez & Beck, 2001, Golden & Consorte, 1982 and Spirito et al., 1981). Meta-analysis of CBT studies in a variety of treatment settings such as clinics, schools, residential treatment programs, and community centers shows effects in the moderate range (Beck & Fernandez, 1998 and Sukhodolsky et al., 2004). A meta-analysis of CBT in the treatment of anger with mostly adult samples found positive effects for treatment, with a mean effect size (Cohen's d) of .70, in which effect sizes were collapsed to yield one effect size per construct per study and weighted according to sample size (Beck & Fernandez). In a meta-analysis of treatment outcome studies for CBT with children and adolescents, the overall mean effect size (Cohen's d) was .67 (Sukhodolsky, et al.). This meta-analysis further separated the effects of skills training, problem solving, affective education, and multimodal interventions, finding Cohen's ds of .79, .67, .36, and .74, respectively (Sukhodolsky, et al.). Sukhodolsky and colleagues found that duration of the treatment was not related to the effect size for the study. In addition, skills training and multimodal treatments were more effective in reducing measures of aggressive behavior and improving measures of social skills, while problem-solving treatments showed greater effects for reducing subjective reports of anger by participants (Sukhodolsky et al.). In this meta-analysis, despite reports of heterogeneity of variance among effect sizes, the studies' effect sizes were not weighted prior to being analyzed (Sukhodolsky et al., p. 257). Social skills training (SST) is another common approach to anger control treatment for children (Stern & Fodor, 1989). This type of training has become more widespread in response to literature suggesting that children who display angry behavior do not have the requisite skills to succeed in social situations (e.g., Coie & Dodge, 1998 and Dodge & Price, 1994), which leads to frustration, acting out, and/or aggressive behaviors. Direct training of social skills is a logical approach to increasing social competence by decreasing the number of negative social interactions children experience. Standard social skills curricula are derived from the social learning theory, social information-processing, and cognitive-behavioral therapy literatures (Frey, Hirschstein, & Guzzo, 2000). Social skills training commonly includes the promotion of skill acquisition, enhancement of skill performance, the reduction or removal of interfering behaviors, and the increase of generalization of skills across settings and over time (Gresham, 1988). Social skills training has been shown to be effective in reducing a variety of problem behaviors in a variety of settings (e.g., Lewis et al., 1998 and Lochman et al., 1984). Although SST is often a part of anger management programs, reviews of SST's effectiveness have not yet addressed its impact on anger and anger-related behaviors specifically (Ang & Hughes, 2002, Nowicki, 2003, Quinn et al., 1999 and Swanson & Malone, 1992). Consequently, it is difficult to assess its effectiveness as an anger-management intervention. The extant literature has yet to be synthesized in a manner that addresses several important questions with respect to anger-related interventions. Are these interventions effective when implemented in school settings? What components of treatment are most effective? Do effect sizes for treatment vary as a function of the characteristics of the participants, study, or of the constructs measured? The meta-analytic studies that have been published in this broad area of the literature (Beck & Fernandez, 1998 and Sukhodolsky et al., 2004) do not uniformly use the procedures that Lipsey and Wilson (1993) recommend for such analyses: i.e., mean effect sizes weighted by the inverse of the variance, corrected for small n and heterogeneity of variance among effect sizes. This meta-analysis extends this literature in several ways. First, it provides a quantitative synthesis of the results of interventions specifically for anger related interventions using objectively verifiable data and appropriate meta-analytic techniques. This kind of analysis is less subject to bias than qualitative reviews ( Cooper & Hedges, 1994). Second, it focuses on interventions that have been implemented with students in schools. Third, it compares the effectiveness of programs that use cognitive behavior therapy, social skills training, and other approaches to remediating students' problems. Fourth, it compares the effectiveness of interventions based upon study characteristics.