Abstract

The Child Behavior Checklist (CBCL) is an extensively standardized parent-completed checklist of competencies and behavior problems of children and adolescents. Clinicians and researchers frequently assume that the published scale scores for the CBCL nonclinical sample are stable even across demographically heterogeneous populations. The present study, a school-based postal questionnaire survey, was designed to compare the CBCL nonclinical sample with a different community sample collected in the U.S. The parents of 530 children, 6 to 10 years of age (73% of the eligible sample), attending one public school system in northern New Jersey were recruited. Mean total behavior problem scores for both sexes in the school sample were dramatically higher than the CBCL nonclinical sample even after removing clinically referred cases from the analyses. Additionally, in contrast to the manual, marked race/ethnicity effects were found in the male subsample. These results, in conjunction with those from other studies, raise serious questions about the common practice of using the CBCL norms as a yardstick for sample comparisons.