The relationship of child and adolescent sexual abuse with symptomatology: A meta-analytic and narrative review

The purpose of this dissertation was to provide a bridging between narrative and statistical information regarding the effects of sexual abuse on children and adolescents by conducting both a narrative and meta-analytic review. A theoretical model was presented to provide a base upon which both the narrative and statistical data could be meaningfully interpreted.;At this time a meta-analysis has not been conducted specifically addressing the effects of sexual abuse on children and adolescents. This meta-analysis examines 75 studies published between 1985 and 1996 which addressed sexual abuse effects in children and adolescents (ages 1-18). Results indicated that sexually abused children and adolescents demonstrated significantly more total psychopathology than both community and clinical non-sexually abused control groups (ES =.60). In other words, the average sexually abused child and/or adolescent who was sexually abused was 73% more likely to experience detrimental psychological effects than the average non-sexually abused child. Sexually abused children also demonstrated greater depression, anxiety, and sexually acting-out than non-sexually abused children. As is consistent with narrative reviews it appears than the largest specific effect score is present when sexual problems are the domain of interest. Regarding methodology, it is important to note that the source of the control group contributed to the degree of symptomatology asserted for sexually abused children. When utilizing community controls there tended to be a greater level of symptomatology than when a study utilized a clinical sample. Narrative reviews tended to conclude that sexually abused samples demonstrate less symptomatology than clinical samples. This was not found to be the case in this meta-analysis. Though the effect size was smaller when sexually abused groups were compared against clinical controls as opposed to community controls, it remained significant. The meta-analysis also found that the relationship between source of informant was not straightforward and tended to interact with the source of the control group. In studies utilizing a clinical control group, parent-report measures actually reported lower effect sizes than self-report and projective measures. However in studies utilizing community control groups parent-report measures were significantly greater than both self-report and projective measures. Thus, sexually abused samples when compared against clinical groups self-reported similar effects as when compared against community groups. However, when compared against clinical samples, parents reported effect sizes more similar to these clinical samples than did sexually abused children themselves.