A Meta-analysis of the Relationship between Nonoffending Parental Support and Abuse Sequelae in Sexually Abused Children

Purpose: Professionals have long believed that parental support is related to better adjustment in their sexually abused children. This relationship has been reported in numerous studies and provides the basis for intervening therapeutically with nonoffending parents as well as some removals of children from their homes. Yet, in a recent analysis by the author, the relationship between parental support and child symptoms emerged for only 2 of 17 outcome scales. The author found several other studies in which null relationships were as or more frequent than significant relationships. The purpose of this meta-analysis was to examine in a rigorous manner whether parental support and child outcomes are or are not related across studies.

Methods: Studies were collected for the meta-analysis using various search engines. The table of contents of certain child maltreatment journals were also reviewed, and a request for studies was placed on a child maltreatment listserv. Criteria for including studies in the meta-analysis were that the paper had to empirically measure postdisclosure parental support and abuse sequelae in children and their relationship between, and had to be accessible as a published paper, paper in press, or a technical report. Thirty-one studies met these criteria. Pearson's Product Moment Correlation, transformed to a Fisher's Z, was used as the effect size. Potential moderators of effect size were also coded and included.

Results: A series of analyses determined effect sizes for the relationship between parental support and different outcomes, including total psychopathology, total behaviors, internalizing (CBCL), externalizing (CBCL), anger and hostility, depression, anxiety, posttraumatic symptoms, dissociation, sexual problems, social competence, and self concept. Although the relationships between parent support and total psychopathology, depression, externalizing, and self concept were significant, effect sizes were uniformly small. Effect sizes, stated as correlations, ranged between .000 and .184 for all but total psychopathology, which was .335 for the four studies in which it was measured. Eleven correlations were below .100. These findings suggest that the relationship between parental support and child outcomes has been greatly overstated.

Implications: The implications of this meta-analysis are many, especially because the findings are the opposite of what has been argued convincingly in the empirical literature for over 20 years. Methodologically, the measurement of parental support is in disarray, with 26 different measures capturing 15 different dimensions in the 31 studies in this meta-analysis. Construct validity must be established by rigorous psychometric analysis for parental support, including a process of determining the important dimensions of parental support. There is also a need to determine the actual effect of professional's assessment of parental support on how systems intervene. For example, some research suggests that the strongest factor related to whether children are removed from their home is parental support (Everson et al., 1989). Given the findings of the meta-analysis, professionals need to reconsider all interventions triggered wholly or in part by their assessment of parental support to determine whether the interventions are still warranted or need to be modified.