Reliability and validity of a Spanish version of the Beck Depression Inventory-II in a sample of the general population of the Canary Islands

This paper analyzes the reliability and validity of Sanz, Navarro, & Vázquez’s Spanish version of the Beck Depression Inventory-II (BDI-II) within a sample of 373 adults of which two thirds were women. Half of them were students and half workers. On the basis of the exploratory and confirmatory factor analysis and the Schmid-Leiman’s transformation, we can affirm that the BDI-II measures a general dimension of depression composed by two highly related factors. Findings related to internal consistency were high: = 0.90, 0.87, and 0.83 for the total scale and factors 1 and 2 respectively. Pearson’s correlations with other self-report measures of depression were high; likewise, the correlations with self-report measures of anxiety and stress were high too. Therefore, the convergent and discriminant validity of the BDI-II is called into question. We conclude that the Spanish adaptation of the BDI-II can be recommended as a suitable screening instrument for depressive symptomatology. Nevertheless, it is clear that the BDI-II has discriminant validity problems, specially, with instruments that measure anxiety.