Is the Autism-Spectrum Quotient a Valid Measure of Traits Associated with the Autism Spectrum? A Rasch Validation in Adults with and Without Autism Spectrum Disorders

Abstract

The Autism-Spectrum Quotient (AQ) is among the most widely used scales assessing autistic traits in the general population. However, some aspects of the AQ are questionable. To test its scale properties, the AQ was translated into Swedish, and data were collected from 349 adults, 130 with autism spectrum disorder (ASD) and 219 without ASD, and analysed with Rasch. Several scale properties of the AQ were satisfactory but it did not meet the criterion of a unidimensional measure of autistic traits. The Rasch analysis showed that the 50-item AQ could be reduced to a 12-item subset with little loss of explanatory power, with the potential to efficiently measure the degree to which adults with and without ASD show autistic traits.

Autistic traits in psychotic disorders: prevalence, familial risk, and impact on social functioning

“…First, despite their psychometric strengths and common use, self-report questionnaires such as AQ-50 and SFS have notable shortcomings. For example, the AQ-50 is not equivalent to an objective assessment of autistic traits, potentially affected by subgroup bias (Agelink van Rentergem, Lever, & Geurts, 2019) and has a debatable factor structure (Lundqvist & Lindner, 2017). Additional observations, interviews, or informant reports are required to confirm the presence of traits and/or a clinical diagnosis of ASD in PD.…”

The Glasgow Sensory Questionnaire: Validation of a French Language Version and Refinement of Sensory Profiles of People with High Autism-Spectrum Quotient

“…The AQ is made of 50 items assessing social skills, attention switching, attention to details, communication and imagination. As the AQ has adequate levels of specificity and sensitivity, it allows targeting individuals with high AQ who would also meet the criteria for ASD diagnosis in large sample data (Baron-Cohen et al 2001;Lundqvist and Lindner 2017;Sonié et al 2013). Limitations of the AQ include the relatively low internal consistency of its subscores, as well as results…”

“…showing that the AQ does not meet all criteria for a unidimensional scale of autistic traits (Lundqvist and Lindner 2017).…”

Introduction to the Special Issue on Open Science Practices: Badges of Honor

“…The differing results between the two measures are likely due to subtle differences in the ways they were designed, including in the creators' definition of the exact construct they respectively measure. Whereas the AQ variants are often administered in clinical populations (Woodbury Smith, Robinson, & BaronCohen, 2005) and were designed to measure autistic traits which typically manifest in autistic individuals (Kanne, Wang, & Christ, 2012;Lundqvist & Lindner, 2017), the BAPQ was designed to measure traits that are qualitatively similar but nonetheless more mild, that may occur in nonautistic relatives of autistic individuals (Hurley et al, 2007).…”