Re-examining the Concordance Between Child Abuse Reported by Youth and Casefile Records

Studying child abuse dimensions (type, severity, and frequency) provides fine-grained information for understanding abuse sequelae. However, different abuse measurement methods often exhibit low agreement with one another. Retrospective self-reports and official records show questionable concordance in measuring abuse types. To date, no research has explored concordance of abuse severity and frequency. The current study examined concordance of official case records and self-reports of child abuse type, severity, and frequency. We also explored how concordance may depend on confirmed versus alleged case reports, developmental timing, and geographical location.

We analyzed data from 884 youth in the Longitudinal Study of Child Abuse and Neglect (LONGSCAN). Data were collected from five U.S. sites.

Case records were collected every two years, starting at age 4, and were coded with the Modified Maltreatment Classification System for abuse type (physical, sexual, and emotional abuse), severity (degree of harm), and frequency. Self-reports were completed at age 12 using a computer-audio interview developed to mirror record coding.

Sexual abuse had the highest concordance compared to other abuse types. Sexual abuse severity had the highest agreement at Midwest and Eastern sites, with particularly high concordance at Midwest sites for elementary-aged exposure. The majority of other concordances indicated only poor to slight agreement. Notable variation existed among site locations and alleged/confirmed reports.

In general, concordance of childhood abuse measures is questionable, with the exception of particular dimensions of sexual abuse. Parsing these dimensions provides information that can inform future research methodology.