Examining the Effectiveness of Psychological Treatments and Interventions for Child Maltreatment: A Meta-analysis

Child maltreatment, consisting of physical abuse, sexual abuse, neglect and general maltreatment, is a prevalent problem in the United States. Children who experience child maltreatment experience a variety of sequelae both in childhood (Tyler, Allison, & Winsler, 2006) and adulthood (Putnam, 2003). Previous meta-analytic studies (e.g., Hetzel-Riggin et al., 2007; Skowron & Reinemann, 2005) have demonstrated the effectiveness of interventions for victims of various forms of child maltreatment, but utilized strict inclusion and exclusion criteria, excluding several studies that did not use control groups, therefore limiting the scope of current research. The purpose of the current meta-analysis was to expand upon the current literature base by examining the overall effectiveness of psychological treatments for child maltreatment using meta-analytic techniques and to examine possible moderating factors. Studies examining treatments or interventions for child maltreatment were collected from online computer searches and examination of review articles. In order to be considered for the meta-analysis the interventions had to target, at least in part, child symptom reduction and also needed to provide sufficient statistical information from which to derive an effect size. The final sample consisted of 49 studies published between 1981 and 2009. A coding sheet was created and each study was coded on sample variables (e.g., age, type of abuse) treatment variables (e.g., treatment modality, setting) and outcome variables (e.g., outcome target, measures). Each relevant statistical comparison (i.e., pre-post test, treatment versus control group or alternate treatment) was converted into an effect size. In general, effect sizes were calculated using Rosenthal and DiMatteo\u27s (2001) suggested procedures. Effect sizes, in the form of correlation coefficients, were determined by converting relevant statistical information from the results of the study (e.g., t statistics, F statistics, chi squares) using single degree of freedom tests, resulting in 837 effect sizes. Results demonstrated an overall mean effect size of r = .299 for treatments for child maltreatment, which according to Cohen (1988) is a medium effect size, and demonstrates practical significance in intervention research (Lipsey, 1998). Examination of moderating factors determined that parental and/or caregiver involvement did not significantly improve the effectiveness of treatments but parents and/or caregivers also benefited from treatment. Behaviorally oriented interventions were more effective at decreasing internalizing and externalizing symptoms than internalizing, externalizing and sexualized symptoms. In terms of treatment modality, multimodal treatments were not found to be more effective than single modality treatments and individual therapy was found to be the least effective treatment modality. The current study aimed to expand upon the work of Skowron and Reinemann (2005) by broadening the inclusion criteria to include studies that had previously been excluded (e.g., studies that did not use a control group) and studies published between 2000 and 2010. Overall, the current meta-analysis determined that treatments for child maltreatment demonstrate medium effect sizes and are effective in reducing children and parent and/or caregiver symptoms or improving parent-child relationships. Through determining effect sizes, meaningful comparisons of the effects of treatments for child maltreatment determined in this meta-analysis can be used as a resource for clinicians to make informed decisions in creating appropriate treatment plans or making referrals for children who have been abused or neglected