Validity of the Beck Depression Inventory
The Beck Depression Inventory (BDI) is a widely used instrument in the measuring of depression and its severity. BDI is a simple instrument: there are twenty-one multiple-choice items; the respondent answers according to his or her feelings as remembered over the past two weeks. The BDI was developed in 1961 for use in care situations so that patients who may have a mood disorder could be detected, but was not originally designed to make a diagnosis. The instrument was revised in accordance to changes made in DSM IV (APA, 1994) for Major Depressive Disorder — the items now match more closely the list of symptoms in the DSM — and is now used primarily to measure the severity of depression in patients diagnosed with depression, though other uses (including more purely diagnostic purposes) remain prominent as well. However, some critics still believe that the BDI, or any self-report instrument, is insufficient to diagnose the disorder of depression. (Coyne, 1994) It should be noted that BDI is often used to screen, or as part of a diagnosis, rather than being used as a singular diagnostic tool.
[...] 905) The literature on BDI tends to confirm an understanding of the limitations on the validity of the test. Some of these limitations are intrinsic to the test itself, to the likely subjects of the instrument (i.e., individuals with depression), to the nature of self-reporting as a process itself. The "genre" of studies generally depends on checking the results of the BDI against the results of some other test, which begs the question of the validity of these tests in general. [...]
[...] Contradictory interpretations are possible: not all depression tests showed signs of susceptibility to fake-good responding, suggesting that the issue is psychometric, but the inverse relation between this susceptibility and a test's validity can be interpreted as evidence that, the better a test measures depression, the more likely it is to detect on a fundamental link between depression and response style. (p. 282) The BDI-II retains a fair amount of validity despite the loss of fidelity at the extreme low end of the scale and the high "face validity" of the instrument, which tends to make respondent manipulation fairly easy. [...]
[...] The Beck Depression Inventory for Primary Care (BDI-PC), for example, is a seven-item version of the instrument that concentrates on only the cognitive subscale. (Steer et al, 1999) Lambert et al (2006) describe the literature that demonstrates the validity of the instrument under a wide variety of circumstances: The BDI displays an acceptable level of internal consistency with psychiatric patients (Cronbach's alpha = .86) and with nonpsychiatric patients (Cronbach's alpha = .81; Beck et al., 1988). Cronbach's alpha for the current study sample was .87. [...]