(GDI), the most widely used self-report measure of depression for children, has been studied, but we still know relatively little about how developmental differences relate to CD1 structure. Here, we evaluated whether the GDI factor structure differed in large samples of clinic-referred children versus adolescents. Our results indicated real but modest differences. Both groups produced 5 first-order factors, but the composition of the factors differed somewhat for children versus adolescents. Both groups also produced a 2nd-order, general depression factor, although items loading on the factor differed somewhat for the 2 groups: For children but not adolescents, several externalizing behavior items (e.g., / never do what I am told} loaded on the general factor: in contrast, for adolescents but not children, several vegetative items (e.g.. / am tired all the time} loaded on the general factor. Overall, the child-factor pattern appeared to involve fewer CD1 items than did the adolescent pattern.

No abstract