Are foster parents reliable informants of children's behaviour problems?

Clinicians and researchers primarily measure behavioural and emotional problems of children in foster care from carer-report checklists. Yet the reliability of these reports is not adequately established. The present study examines one indicator of reliability for foster parent checklist reports: interrater agreement between foster parents and teachers.

Estimates of interrater agreement of foster parent and teacher responses on the cross-informant scales of the Child Behaviour Checklist (CBCL) and the Teacher Report Form (TRF) were obtained for 47 children in long-term foster care, aged 5-11 years. The estimates included calculations of agreement for continuous measures of problem behaviour, as well as for categorical determinations of clinically significant behaviour.

Correlations of CBCL and TRF mean raw scores for the total problems (r = 0.71) and externalizing (r = 0.78) scales exceeded those described in prior studies of parent-teacher agreement, while correlation for internalizing scores (r = 0.23) was similar to that found previously. Teachers and foster parents demonstrated moderate to good agreement (kappa = 0.70-0.79) in identifying clinically significant total problems and externalizing problems, but poor agreement in identifying internalizing problems.

Discrepancies between these and prior findings are discussed. For children in long-term foster care, foster parents or teachers may be used as informants for total problems, externalizing problems, and social-attention-thought problems. The reliability of data on internalizing symptoms is less certain.