Scales, diagnoses, and child psychopathology: I. CBCL and DISC relationships

To clarify the relationship between scales and structured diagnostic interview diagnoses, the authors used a two-stage screening method to study 201 military families with one or more children ages 5 to 17.

Parents and children were interviewed with the Diagnostic Interview Schedule for Children (DISC 2.1); parents also completed the Child Behavior Checklist (CBCL) while the children completed other self-report symptom scales.

Results indicate only a modest ability of scales to discriminate among discrete DISC-derived DSM-III-R diagnoses. Inclusion of diagnostic information from both parents and children resulted in more diagnoses than from either informant alone, and the additional diagnoses consisted mostly of internalizing disorders contributed by child-derived DISC information. In general, correlations were larger between scales and diagnoses within the same informant (regardless of diagnostic construct) than across informants (but within the same diagnostic construct). Child self-report measures tended to outperform the CBCL as screeners against the overall "caseness" criterion on the DISC. However, child self-report scales were relatively nonspecific and showed little ability to selectively identify internalizing disorders such as anxiety and/or depression. Compared with single informant diagnoses, combined-informant diagnoses were generally superior in demonstrating broader relationships to both parent and child symptom scales.

Additional research is needed in order to build careful crosswalks between the various approaches to assessing childhood psychopathology, to decide on optimal rules for combining information to establish diagnoses, and to validate the currently available assessment alternatives.