Overview
The original Autism-Spectrum Quotient (AQ; Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley, 2001) is an individually-administered questionnaire designed to identify the presence of the characteristics of autism spectrum disorder (ASD) in adults via self-report. Since that time, additional versions of the AQ have been developed, including the school-age adolescent version (AQ-Adol; Baron-Cohen, Hoekstra, Knickmeyer, & Wheelwright, 2006) for ages 9 years, 8 months to 15 years, 4 months, and the children’s version (AQ-Child; Auyeung, Baron-Cohen, Wheelwright, & Allison, 2008) for ages 4 years to 11 years. The AQ contains 50 items that describe five areas associated with ASD: social skills, attention switching, attention to detail, communication, and imagination. For the Child and Adolescent versions, parents rate each item on a range from “definitely agree” to “definitely disagree.” Cut-off scores vary depending on the version (Adult: 32; Adolescent: 30; Child: 76). All versions of the AQ have been translated into multiple languages. The instrument is not commercially-available but is available on the researchers’ website, listed below the Summary table.
Summary
Age: 18 years - 100 years
Time to Administer: Approximately 20 minutes
Method of Administration: Individually-administered self-report
Points are summed to yield a comparison to a cut-off score that indicates clinically-significant levels of autistic traits; higher scores indicate more autistic traits.
Subscales: Five areas are assessed: Social Skill, Attention Switching, Attention to Detail, Communication, Imagination
Screening/Diagnosis: S
Autism Related Research
Gomez, Stavropoulos, & Vance (2019)
Age Range: 4-11 years
Sample Size: 404
Topics Addressed:
Psychometric properties of AQ-Child
Outcome:Gomez, Stavropoulos, & Vance (2019)
Some items (29, 30, 49) did not adequately discriminate between those with and without ASD. Exploratory factor analysis with the remaining 47 items supported a four-factor model (mind reading, attention to details, social skills, imagination). Internal consistency reliability for all four factors were acceptable (above .70), further supporting a four-factor model. In the four-factor model, 32 items had salient loadings (> .45).
Conclusions: removal of some items from the AQ-Child may result in a better measure for this age group.
Ashwood et al. (2016)
Age Range: Adults
Sample Size: 476
Topics Addressed:
Predictive validity of AQ
Outcome:Ashwood et al. (2016)
73% received a clinical diagnosis of ASD, and AQ scores did not significantly predict receipt of a diagnosis. AQ scores provided high sensitivity of 0.77 and positive predictive value of 0.76, but specificity was low at 0.29 and negative predictive value was 0.36. Thus, 64% of those who scored below the AQ cut-off were false negatives. Results also suggested that comorbid generalized anxiety disorder may “mimic” ASD and inflate AQ scores, leading to false positives.
Conclusion: in this sample, the AQ utility for screening referrals was limited in this sample of adults.
Auyeung, Baron-Cohen, Wheelwright, & Allison (2008)
Age Range: 4-11 years
Sample Size: 1,765
Topics Addressed:
Adaptation of AQ to a child version; investigation of psychometric properties
Outcome:Auyeung, Baron-Cohen, Wheelwright, & Allison (2008)
Excellent test-retest reliability and high alpha coefficients for the questionnaire as a whole, as for each of the five subscales. Factor analysis supported four factors which were highly correlated with four of the original subscales. All four factors showed high reliability coefficients. Age was not associated with score.
Conclusion: the AQ-Child is a strong measure that is able to discriminate the presence of autistic characteristics in a clinical sample.
Baron-Cohen, Hoekstra, Knickmeyer, & Wheelwright (2006)
Age Range: 9:8-15:4 years
Sample Size: 181
Topics Addressed:
Adaptation of AQ to an adolescent version
Outcome:Baron-Cohen, Hoekstra, Knickmeyer, & Wheelwright (2006)
~90% of participants with ASD scored 30+ vs. none of the controls; boys scored higher than girls. Age was not associated with score.
Conclusion: the Adolescent version of the AQ is useful for identifying where an adolescent is situated on the continuum from autism to normality.