abstract
- One promising research approach in the search for clinically meaningful diagnostic categories of depression has been the application of multivariate statistical techniques such as factor analysis to the evaluation of self-report and observer rating data. The contribution of factor analysis has been limited, however, by the application of traditional mathematical criteria in the analyses which typically produce too many factors. In this study, a criterion of replicability was applied with the aid of a factor matching procedure (FACTOREP). The Beck Depression Inventory (BDI) was examined using FACTOREP and two large samples of depressed patients. The results showed that only a large general factor was present. Beyond the BDI studies, there has been debate on the validity of the endogenous versus reactive depression conceptualization. Despite the limited number of endogenous items in the BDI, the results here did not confirm the presence of two clear factors that represented endogenous and reactive groups of items.