Attribution Questionnaire

Development and history

Surveys indicate that people hold deeply embedded beliefs about mental illness, often attributing behaviors to the individual's control. Attribution theory suggests that such perceptions can lead to stigmatizing responses like blame, refusal to help, anger, and fear. A 2003 study by Corrigan, Markowitz, Watson, Rowan, and Kubiak examined the validity of the AQ, analyzing common areas of stigma. The researchers created the AQ based on attribution theory and tested it on a sample of U.S. community college students, using a vignette manipulated to describe Harry as either non-violent or violent.

The original version had 21 items grouped into 6 subscales: personal responsibility, pity, anger, fear, helping/avoiding behavior, and coercion/segregation. After the study, Corrigan developed a 27-item version with 9 subscales, adding dangerousness and splitting coercion, segregation, avoidance, and helping into separate subscales. This AQ-27 is the most current version. Its subscales include blame (mental illness is controllable and the person is responsible), anger (irritation due to blame), pity (sympathy), help (willingness to assist), and dangerousness (perception of threat).

Versions

The AQ-9 is a shorter version with 9 items, each representing one of the AQ-27 subscales, selected for strongest factor loadings. The r-AQ, based on the AQ-9, was modified for children aged 10–18 using simpler vocabulary and a vignette of a younger individual. Its first eight items measure stigma constructs, and the last asks whether the child would seek mental health treatment if needed. The children's Attribution Questionnaire (AQ-8-C) includes 8 items for each stigma except coercion, with simplified vignette and items tested as reliable for this age group.

Psychometric properties

The AQ has been used extensively in research across various age ranges, but normative data for calibrating scores are not available to the general population. Reliability testing is still needed, as the scale is relatively new and requires more studies to confirm stable and consistent results. Validity assessments, particularly discriminative validity for screening, are used to evaluate whether the tool measures what it intends, though further validation is ongoing.

Limitations

As a self-report measure, the AQ is susceptible to social desirability bias, where respondents may underreport stigmatizing attitudes. It relies on self-report rather than behavioral observation, so answers may not reflect actual behaviors. Most studies have been conducted in U.S. populations, limiting generalizability, and validation has primarily involved volunteer samples of college and adolescent students. The nine subscales may miss key components of stigma, and additional subscales might be needed as research progresses.

Did You Know?

• The r-AQ includes a revised Level of Contact Report to assess familiarity with mental illness.
• An Italian version of the AQ exists, but it is not widely translated into other languages.