The Children's Depression Inventory: A systematic evaluation of psychometric properties.

Community Mental Health Center, Yorkville, New YorkThis article presents data from a series of studies designed to determine thepsychometric properties of the Children's Depression Inventory (GDI). A total of294 school children and 269 children who were psychiatric inpatients served assubjects. Reliability studies were conducted with both normal and clinicalpopulations. Good internal consistency was documented, but test-retest data werevariable across populations and test-retest intervals. Validity studies indicate thatthe GDI can distinguish children with general emotional distress from normalschool children. However, differences between GDI scores of depressed (bysymptom checklists from the Diagnostic and Statistical Manual of MentalDisorders, DSM-1II; American Psychiatric Association, 1980) and nondepressedchildren were not significantly different. There was a good correspondencebetween the GDI and self-report measures of self-concept. Our data suggestedthat the GDI measures a multidimensional construct that overlaps with otherchildhood disorders, particularly anxiety. Findings are reviewed and discussedrelative to data presented by Kovacs, Kazdin, and others. Although the GDI maystill be the best researched instrument available to measure depression from thechild's viewpoint, more work is needed before it can be interpreted withconfidence in clinical and research settings.