MEASURING DEPRESSION FOR SWEDISH NONCLINICAL ADOLESCENTS - FACTORIAL VALIDITY AND EQUIVALENCE OF THE BECK DEPRESSION INVENTORY ACROSS GENDER

Building upon earlier confirmatory factor analytic research that valid ated a 2nd-order 4-factor structure of the Beck Depression Inventory f or Canadian and Swedish nonclinical adolescents, the purposes of the p resent study were twofold: (a) to test the factorial structure of the Swedish version of the instrument separately for males (n = 559) and f emales (n = 537), and (b) to test for its measurement and structural e quivalence across gender. Except for three minor discrepancies involvi ng correlated errors, the hypothesized model best described the data f or both males and females. Nonetheless, the imposition of equality con straints across sex revealed differences bearing on the measurement of six items, and latent construct relations involving the Somatic Eleme nts factor. Four gender differences in factorial structure replicated findings for Canadian adolescents. Results have important implications for researchers, clinicians, and others whose interests focus on nonc linical adolescent depression.