What is the BDI-II?
The Beck Depression Inventory, 2nd Edition, (BDI-II, Beck, Steer, & Brown, 1996) is a very well-researched instrument widely used by clinicians and researchers to measure depression. The BDI-II questionnaire consists of 21 groups of statements associated with symptoms of depression. The examinee is asked to pick one statement from each group that best describes his or her experience over the past two weeks.

Comparison of Item Content
The BDI-II takes longer to complete. QPASS’s Depression-Scan is more economically worded than the BDI-II, having been formatted to the Quick-flow design. Hence, even though Depression-Scan measures 26 symptoms of depression, while the BDI-II assesses only 21 symptoms, the BDI-II takes 5-10 minutes to complete (Beck, Steet, & Brown, 1996) while Depression-Scan takes only about 2-3 minutes to complete. As Table 2.1 shows, both the BDI-II and Depression-Scan assess sadness, hopelessness, sense of failure, loss of pleasure, guilty feelings, self-dislike, self-criticism, suicidal thoughts, crying, agitation, loss of interest, indecisiveness, feeling worthless, loss of energy, change of sleep, irritability, change of appetite, loss of concentration, fatigue, and loss of libido.

However, as the table also shows, there are some few notable differences:

Depression-Scan includes the items “Difficulty laughing or feeling amused” whereas the BDI-II does not. This item was included on Depression-Scan because it was found to have a strong item-total correlation (.69) with the construct of depression in a sample of 386 outpatients (Lownsdale, 2001), and is a component of anhedonia (lack of pleasure), a major component of depression included among DSM-IV-TR criteria for Major Depressive Episode.

Depression-Scan includes the item “Feeling trapped” whereas the BDI-II does not. According to Nesse (2000), “Depression is precipitated not so much by life events that involve loss, but by those that are . . . entrapping” (p. 9). In addition to theoretical reasons for including entrapment as a symptom of depression (discussed in Chapter 6), this item was included on Depression-Scan because it was also found to have a strong item-total correlation (.69) with the construct of depression in a sample of 386 outpatients (Lownsdale, 2001).

Depression-Scan assesses “Feeling empty” whereas the BDI-II does not. The item “feeling empty” was included in Depression-Scan because it is included in the DSM-IV-TR criteria of Major Depressive Disorder, associated with “depressed mood” (p. 356), and because it was found to have a strong (.75) item-total correlation with the construct of depression in a sample of 386 outpatients (Lownsdale, 2001).

Depression-Scan assesses “Feeling helpless” but the BDI-II does not. Depression-Scan includes “feeling helpless” (item 11), even though this symptom is neither assessed on the BDI-II nor included among the criteria for Major Depressive Episode. The learned helplessness model of depression (Seligman, 1975) provides an important insight into the psychosocial causes of depression. In the sample of 386 outpatients upon which QPASS is based, “Feeling helpless” (item 11) was found to have a strong statistical relationship with the construct of depression: a .73 item-total correlation in the 40-item pilot version of Depression-Scan (Lownsdale, 2001).

BDI-II assesses “Punishment feelings” whereas Depression-Scan does not. The reason the item “Feeling punished” was excluded from Depression-Scan is because it was found to have a relatively weak statistical relationship with the construct of depression in 386 outpatients (Lownsdale, 2001). Moreover, “feeling punished” is not included as criteria for Major Depressive Disorder in the DSM-IV-TR (see p. 356).

Time Frames Compared

QPASS’s Depression-Scan collects depression symptomatology over a 1-week time span, whereas the BDI-II collects symptom data over a 2-week time span. The 1-week time referent for QPASS scales was selected primarily to decrease inaccurate reporting of symptoms, which is explained in more detail in Chapter 5. As Hamilton (1970) and Derogatis (1994) have both noted, the ability to recall one’s subjective emotional experience fades with time. This is especially true for individuals with mental disorders, who, in their distress, can be expected to easily forget how they felt more than a week ago. Therefore, making the time referent longer than 7 days, as the BDI-II does, may increase the risk of inaccurate, distorted self-reports on negative emotions

The BDI-II collects depression symptomatology over a 2-week time span in order to correspond with the 2-week DSM-IV criteria for the diagnosis of Major Depressive Disorder. However, the earlier version of the instrument, the Beck Depression Inventory (BDI), collected self-reported data on depressive symptoms over a 1-week period. The extended time frame of the BDI-II places it at risk of being less accurate than the BDI. As Dozois, Dobson, and Ahnberg (1998) observe, the 2-week time frame of BDI-II makes it “less sensitive to temporal fluctuations in the depressive experience than the BDI” (p. 88).

Psychometric Properties

Reliability. Reliability depends on the internal consistency of a measure, which reflects the degree to which the items comprising the measure assess the same construct (Anastasi, 1982). A standard method for estimating reliability in tests with multiple-scored items is calculating Cronbach's alpha, sometimes referred to as coefficient alpha (Cortina, 1993; Cronbach, 1951). In a group of 500 psychiatric outpatients, the coefficient alpha of the BDI-II was found to be .93 (Beck, Steer, & Brown, 1996). The coefficient alpha for Depression-Scan in 386 mental health outpatients was found to be .95 (Lownsdale, 2001). The test-retest reliability of the BDI-II in 26 outpatients administered the instrument 7 days apart was found to be .93 (Beck, Steer, & Brown, 1996). The test-retest reliability of Depression-Scan was found to be .79 (Lownsdale, 2001) in 27 outpatients administered the scale 7-14 days apart.

Concurrent validity. Scores on the BDI-II correlated with scores on the Revised Hamilton Psychiatric Rating Scale for Depression (HRSD-R) at .71 in 87 outpatients (Beck, Steer, & Brown, 1996). Depression-Scan scores correlated with scores on the BDI-II at .83 in a sample of 82 mental health outpatients (Lownsdale, 2001).

Discriminative validity. Discriminative validity refers to the ability of an instrument to discriminate between groups that are expected to differ on a construct (Foster & Cone, 1995). As reported in Chapter 8 of this manual, the two instruments performed equally well in discriminating between 41 depressed and 41 non-depressed outpatients (Lownsdale, 2001).

Summing Up

Similiarities. The BDI-II and Depression-Scan have similar item content that is highly consistent with DSM-IV-TR criteria for Major Depressive Disorder. Internal consistency reliability coefficients for the two instruments are very high (+.90), and the correlation between scores on Depression-Scan scores and the BDI-II is high (.83; N=82; Lownsdale, 2001). According to one study (Lownsdale, 2001), evidence suggests no difference between the two instruments in discriminating between depressed and non-depressed patients.

Differences. The 26-item Depression-Scan is a more comprehensive measure of depression than the 21-item BDI-II, and takes less time to complete. Depression-Scan assesses the following symptoms which the BDI-II does not: “difficulty laughing,” “feeling trapped,” feeling helpless” and “feeling empty,” that latter of which is included in the DSM-IV-TR criteria for Major Depressive Disorder (see p.356). Although the BDI-II assesses punishment feelings, this is not a symptom included in the DSM-IV-TR criteria for Major Depressive Disorder. The BDI-II’s 2-week time reference increases its risk for error in self-report than Depression-Scan, which has a 1-week time reference.