What is the Beck Anxiety Inventory?
The Beck Anxiety Inventory (BAI, Beck & Steer, 1993) is another one of Beck’s well known and widely used psychological instruments. The BAI was designed “to measure symptoms of anxiety which are minimally shared with those of depression” (BAI, Beck & Steer, 1993, p. 1). On the BAI, the examinee is asked to rate 21 symptoms of anxiety on a 4-point Likert scale. On Anxiety-Scan the examinee is asked to rate 29 symptoms of anxiety on a 5-point Likert scale. On both scales, the time reference is “the past seven days, including today.”

Comparison of Item Content
As the table shows, the 29-item Anxiety-Scan on Qpass assesses all symptoms that the BAI does except for two: “face flushed” and “wobbly legs.” These items were not included on Qpass’s Anxiety-Scan because neither are DSM-IV-TR criteria for any anxiety disorder, and both were found to have relatively weak statistical relationships with the construct of anxiety in 386 outpatients (Lownsdale, 2001).

Social and relational aspects of anxiety assessed by Anxiety-Scan. The Anxiety-Scan scale on Qpass assesses the following symptoms that are related to the social and relational aspects of anxiety. Responses to these items can be very helpful to the clinician in screening for or evaluating Social Phobia, or the social-relational context of the respondent’s anxiety. However, these symptoms are not assessed by the BAI:

• Fears of being disapproved of (Qpass item 47)
• Fears of being ridiculed or laughed at (Qpass item 48)
• Fears of being rejected (Qpass item 49)
• Fears of being abandoned (Qpass item 50)

Panic attack symptoms assessed by Anxiety-Scan. Anxiety-Scan assesses all the symptoms of a panic attack, according to the DSM-IV-TR, where as the BAI does not. Specifically, the following symptoms associated with panic attack are assessed by Anxiety-Scan, but are not assessed by the BAI:

• Depersonalization (“feeling detached from the body,” Qpass item 54, which is a symptom of Depersonalization Disorder as well of a panic attack);
• Derealization (“a sense of being aliened from others” and “a sense of being in a dream or daze,” Qpass item 52)

PTSD symptoms assessed by Anxiety-Scan. Anxiety-Scan also assesses the following DSM-IV-TR criteria for Post Traumatic Stress Disorder, an anxiety disorder, which are not assessed by the BAI:

• Flashbacks (“a sense of being back in the past, experiencing a troubling event all over again,” Qpass item 55);
• Hypervigilance (“a sense of being on the lookout for danger,” Qpass item 53)

Psychometric Properties

• Internal consistency reliability. The coefficient alpha of the BAI in 160 outpatients was .92 (Beck & Steer, 1993). The coefficient alpha of Anxiety-Scan in 386 outpatients was slightly higher, at .95 (Lownsdale, 2001).
• Test-retest reliability. The test-retest reliability of the BAI in 83 outpatients administered the instrument 7 days apart was found to be .75 (Beck & Steer, 1993). In 27 outpatients administered the scale 7-14 days apart, the test-retest reliability of Anxiety-Scan was found to be .74 (Lownsdale, 2001).
• Concurrent validity. BAI scores were found to correlate with scores on the Hamilton Anxiety Rating Scale – Revised (HARS-R) at .51 in 160 outpatients (Beck & Steer, 1993). Anxiety-Scan scores were found to have a .80 correlation with scores on the BAI in a sample of 82 outpatients (Lownsdale, 2001).
• Discriminative validity. No notable difference was found between Anxiety-Scan and the BAI in their respective abilities to discriminate a group of 19 outpatients diagnosed with anxiety disorders from a group of 63 outpatients with other diagnoses (Lownsdale, 2001).

Summing Up

• Similarities. Both Qpass’s Anxiety Scan and the BAI assess self-reported anxiety symptoms over a 1-week time period, take only several minutes to complete, and have very high (+.90) internal consistency reliability. Scores on each scale were found to highly correlate at .80 (N=82; Lownsdale, 2001).
• Differences. The 29-item Anxiety-Scan is a far more comprehensive measure of anxiety than the 21-item BAI, offering 38% more self-reported data to the clinician on anxiety, which can be helpful in making a DSM-IV-TR diagnosis.