A meta‐analytic review of findings from national samples on psychological correlates of child sexual abuse

Abstract

In response to the availability of a growing literature on the psychological correlates of child sexual abuse (CSA), numerous researchers have conducted literature reviews of these correlates. These reviewers have generally reported that CSA is associated with a wide variety of adjustment problems, and many have additionally implied or concluded that, in the population of persons with CSA experiences, (a) CSA causes psychological harm, (b) this harm is pervasive, (c) this harm is intense, and (d) boys and girls experience CSA equivalently. However, with few exceptions, these reviewers have included in their reviews mostly studies using clinical and legal samples; these samples cannot be assumed to be representative of the general population. To evaluate the implications and conclusions of these reviewers, we conducted a literature review of seven studies using national probability samples, which are more appropriate for making population inferences. We found that, contrary to the implications and conclusions contained in previous literature reviews that were focused on biased samples, in the general population, CSA is not associated with pervasive harm and that harm, when it occurs, is not typically intense. Further, CSA experiences for males and females are not equivalent; a substantially lower proportion of males reports negative effects. Finally, we found that conclusions about a causal link between CSA and later psychological maladjustment in the general population cannot safely be made because of the reliable presence of confounding variables. We concluded by cautioning that analysis at the population level does not characterize individual cases: When CSA is accompanied by factors such as force or close familial ties, it has the potential to produce significant harm.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 221 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.