Classifying college students as depressed or nondepressed using the Beck Depression Inventory: an empirical analysis.

Abstract

Two studies were conducted to assess the consistency of classification of college students as depressed or not depressed. In the first study, classification of students at three depression levels (not depressed, mildly depressed, moderately depressed) was compared using Beck Depression Inventory (BDI) scores obtained on two occasions separated by either a 1-day or a 1-week interval. Whereas the overall test-retest reliability coefficients were acceptable, particularly when both administrations of the BDI occurred on the same day, the consistency of classification of subjects into the mild and moderate levels of depression was poor. In an attempt to determine the factors contributing to the poor classification consistency, a second study compared the effect of exposure to the BDI on subsequent BDI responses. Subjects who completed the BDI on two occasions showed significantly lower BDI scores compared to subjects exposed to an alternate depression inventory (the Zung Self-Rating Depression Scale) or no prior questionnaire. In addition, a large percentage of subjects classified as moderately depressed following the initial BDI administration failed to be classified within the moderate categories on retesting the same day. Subjects tested using the Zung Depression Scale maintained the same depression classification when retested on the BDI. Item analyses failed to reveal significantly different amounts of change among individual BDI items.