Autism Spectrum Quotient (AQ) – Self-administered

1. Abstract

The Autism Spectrum Quotient (AQ) – Self-administered is a widely recognized psychometric screening instrument developed by Simon Baron-Cohen and colleagues at the Autism Research Centre (ARC), University of Cambridge. Designed for adults aged 16 years and older with average or above-average intellectual functioning, the AQ quantifies the degree of autistic traits across five distinct phenotypic domains: social skills, attention switching, attention to detail, communication, and imagination. Comprising 50 self-report items evaluated on a 4-point response scale ranging from “definitely agree” to “definitely disagree,” the instrument was originally scored dichotomously (0 or 1 point per trait endorsement; theoretical range 0–50), though continuous 4-point Likert scoring (range 50–200) is frequently utilized in contemporary behavioral genetics and dimensional psychopathology research. In normative and clinical validation samples, an operational cutoff score of 32 correctly identified 79.3% of adults with Asperger syndrome or high-functioning autism, demonstrating high sensitivity (0.77–0.90) and specificity (0.80–0.98) relative to general population controls. The AQ exhibits sound internal consistency across dimensions (α ranging from 0.63 to 0.77 for subscales, and 0.79 to 0.88 for the total score), robust test-retest stability (r = 0.70–0.85 over 2 to 6 weeks), and solid construct validity corroborated by inverse correlations with the Empathy Quotient (EQ) and convergent correlations with the Systemizing Quotient (SQ). While not a standalone diagnostic instrument, the AQ represents a foundational tool for cognitive phenotype profiling, epidemiological screening, neurobiological subtyping, and continuous trait quantification across both clinical and neurotypical adult populations.

2. Keywords

Autism Spectrum Quotient, AQ, autism spectrum disorder, screening instrument, autistic traits, social communication, attention switching, systemizing, theory of mind, psychometrics, adult autism assessment, broad autism phenotype

3. Authors

The Autism Spectrum Quotient was developed by an interdisciplinary research team led by Professor Sir Simon Baron-Cohen at the Autism Research Centre (ARC), Department of Psychiatry, University of Cambridge, United Kingdom. Key co-developers and contributing psychometricians include:

• Simon Baron-Cohen, PhD, FBA, FMedSci: Professor of Developmental Psychopathology, Director of the Autism Research Centre, Fellow of Trinity College, Cambridge.
• Sally Wheelwright, MA: Senior Psychometrician and Researcher, Autism Research Centre, Department of Psychiatry, University of Cambridge.
• Richard Skinner, PhD: Clinical Researcher and Psychometrician, Autism Research Centre, University of Cambridge.
• Jacqueline Martin: Research Collaborator and Clinical Coordinator, Autism Research Centre.
• Emma Clubley: Clinical Researcher and Data Analyst, Autism Research Centre.

Institutional Contact: Autism Research Centre, Department of Psychiatry, University of Cambridge, Douglas House, 18b Trumpington Road, Cambridge CB2 8AH, United Kingdom. Official Resource Repository: Autism Research Centre Tests Portal.

4. Purpose

The primary purpose of the Autism Spectrum Quotient (AQ) is to provide a brief, standardized, and self-administered dimensional measurement of cognitive, behavioral, and perceptual traits associated with the autism spectrum in individuals aged 16 years and older who possess average or above-average intellectual quotient (IQ ≥ 85). Historically, diagnostic frameworks such as the DSM-IV and ICD-10 conceptualized conditions such as Asperger syndrome, childhood autism, and pervasive developmental disorder not otherwise specified (PDD-NOS) as discrete diagnostic categories. Baron-Cohen and colleagues developed the AQ to test the empirical proposition that autism represents an extreme manifestation of continuous traits distributed across the entire general population.

From a clinical perspective, the AQ serves as a triage and screening mechanism in adult mental health services, tertiary neuropsychiatric clinics, and general practice. Adults presenting with late-onset functional difficulties, interpersonal friction, chronic social anxiety, or atypical executive functioning frequently go undiagnosed through childhood. The AQ provides clinicians with an objective, standardized metric to determine whether a comprehensive diagnostic evaluation—typically involving the Autism Diagnostic Observation Schedule (ADOS-2) and developmental history via the Autism Diagnostic Interview-Revised (ADI-R)—is warranted. Using the recommended empirical cutoff score of 26 or 32, the AQ discriminates between clinical and non-clinical populations with exceptional precision, thereby preventing unnecessary overburdening of specialized neurodevelopmental diagnostic pathways.

In scientific and research settings, the AQ serves multiple pivotal functions. In molecular genetics, neuroimaging, and cognitive neuroscience, researchers require dimensional indices to examine genetic loci, neural activation patterns (e.g., in the fusiform face area, amygdala, and temporoparietal junction), and cognitive markers that scale continuously with autistic traits in non-clinical cohorts. Furthermore, the AQ facilitates the characterization of the Broad Autism Phenotype (BAP) in unaffected first-degree relatives of autistic probands, allowing behavioral geneticists to model heritability and endophenotypic transmission. Finally, the scale allows cross-disciplinary comparisons, demonstrating documented variations across academic disciplines (e.g., elevated systemizing-related scores in mathematics, engineering, and computer science compared to humanities and social sciences).

5. Psychological Construct

The psychological construct evaluated by the AQ is the dimensional autistic phenotype, operationalized through five interrelated yet psychometrically dissociable subdomains. Each domain targets core cognitive, social-pragmatic, or sensory-perceptual characteristics historically defined in clinical accounts of autism and Asperger syndrome:

5.1. Social Skills (10 items: 1, 11, 13, 15, 22, 36, 44, 45, 47, 48)

This subscale assesses an individual’s intuitive preference for social interaction, comfort in interpersonal environments, ability to form friendships, and diplomatic social acumen. Higher trait scores reflect marked introversion, preference for solitary over communal pursuits (e.g., Item 13: “I would rather go to a library than to a party”), difficulties initiating and maintaining social relationships (e.g., Item 22: “I find it hard to make new friends”), and reduced inclination toward social diplomacy (e.g., Item 48: “I am a good diplomat” [reverse scored]).

5.2. Attention Switching / Tolerance of Change (10 items: 2, 4, 10, 16, 25, 32, 34, 37, 43, 46)

This dimension examines executive cognitive flexibility, propensity for repetitive behaviors, insistence on sameness, and cognitive set-shifting ability. Endorsement of autistic traits in this domain indicates intense monotropic absorption (e.g., Item 4: “I frequently get so strongly absorbed in one thing that I lose sight of other things”), cognitive rigidity regarding daily schedules (e.g., Item 25: “It does not upset me if my daily routine is disturbed” [reverse scored]), executive difficulties managing parallel streams of information (e.g., Item 10), and acute anxiety when encountering novel or unplanned circumstances (e.g., Item 46: “New situations make me anxious”).

5.3. Attention to Detail (10 items: 5, 6, 9, 12, 19, 23, 28, 29, 30, 49)

This subscale captures sensory hyper-reactivity, local vs. global perceptual processing biases, and systemizing inclinations toward structured alphanumeric data. Grounded in the concept of Weak Central Coherence, high scorers exhibit heightened visual or auditory acuity for minute environmental stimuli (e.g., Item 5: “I often notice small sounds when others do not”; Item 12: “I tend to notice details that others do not”), a preoccupation with cataloging and numerical information (e.g., Item 6, Item 9, Item 19), and a tendency to prioritize local feature extraction over overarching thematic gestalt (e.g., Item 28).

5.4. Communication (10 items: 7, 17, 18, 26, 27, 31, 33, 35, 38, 39)

The communication dimension evaluates pragmatic language competence, discourse reciprocity, conversational turn-taking, and the interpretation of figurative or non-literal expressions. Individuals with higher autistic traits exhibit breakdowns in social-pragmatic fluency, difficulties inferring unstated social intent (e.g., Item 27: “I find it easy to ‘read between the lines’ when someone is talking to me” [reverse scored]), challenges regulating monologic speech on specialized topics (e.g., Item 18, Item 39), and pragmatic misunderstandings regarding verbal politeness norms (e.g., Item 7: “Other people frequently tell me that what I’ve said is impolite, even though I think it is polite”).

5.5. Imagination (10 items: 3, 8, 14, 20, 21, 24, 40, 41, 42, 50)

The imagination subscale probes generative pretend play, mentalistic perspective-taking, narrative comprehension, and aesthetic preferences. Grounded in mentalizing and empathizing frameworks, elevated autistic scores indicate difficulties attributing internal mental states to fictional characters or real-life peers (e.g., Item 20: “When I’m reading a story, I find it difficult to work out the characters’ intentions”; Item 42: “I find it difficult to imagine what it would be like to be someone else”), historical avoidance of collaborative pretend play in childhood (e.g., Item 40, Item 50), and a relative preference for literal, factual categorization over abstract narrative fiction (e.g., Item 21, Item 41).

6. Theoretical Framework

The development and interpretation of the Autism Spectrum Quotient are anchored in three major theoretical frameworks formulated within contemporary cognitive psychopathology and neuropsychology:

6.1. Empathizing-Systemizing (E-S) Theory and the Extreme Male Brain Theory

Formulated by Simon Baron-Cohen (2002, 2004), the Empathizing-Systemizing (E-S) theory posits that psychological phenotypes vary along two fundamental cognitive dimensions: empathizing (the drive to identify, predict, and emotionally respond to mental states) and systemizing (the drive to analyze, construct, and predict lawful, rule-governed inanimate systems). The AQ assesses manifestations of reduced empathic drives (reflected in the Social Skills, Communication, and Imagination subscales) juxtaposed with enhanced systemizing drives (reflected in the Attention to Detail and Attention Switching subscales). In its extended form, the Extreme Male Brain (EMB) theory hypothesizes that autism represents a hyper-masculinized cognitive profile characterized by extreme systemizing alongside severe empathizing deficits. This theoretical framework accurately predicted that neurotypical males would score moderately higher on the AQ than neurotypical females, while autistic individuals of both sexes would exhibit extreme scores skewed in the hyper-systemizing direction.

6.2. Theory of Mind (Mindreading / Mentalizing) Deficit Hypothesis

The cognitive core of autism has long been conceptualized as an impairment in Theory of Mind (ToM)—the capacity to attribute mental states (beliefs, desires, intentions, emotions) to self and others to explain and predict behavior (Baron-Cohen, Leslie, & Frith, 1985). Within the AQ, items addressing pragmatic conversational reciprocity, irony detection, indirect requests, and character mental state attributions (Items 7, 20, 27, 31, 35, 36, 42, 45) directly translate ToM performance paradigms into self-reflective psychometric inquiries. A failure to spontaneously engage mentalizing algorithms results in difficulties reading micro-expressions, identifying conversational boredom, or resolving social cues.

6.3. Weak Central Coherence (WCC) and Enhanced Perceptual Functioning (EPF)

Proposed by Uta Frith and elaborated by Laurent Mottron, the Weak Central Coherence account posits a perceptual processing bias toward local, piecemeal detail over contextual gestalt. Rather than viewing this as purely a deficit, the AQ recognizes that weak central coherence confers cognitive strengths in specific domains, such as pattern recognition, error identification, proofreading, and alphanumeric recall. Items measuring heightened perceptual acuity for subtle sounds, visual patterns, and serial information (Items 5, 6, 9, 12, 19, 23, 28) capture this distinct perceptual cognitive style.

7. Validity

The psychometric validity of the AQ has been extensively corroborated across diverse international populations, clinical cohorts, and cross-sectional designs.

7.1. Criterion and Known-Groups Validity

In the seminal validation investigation by Baron-Cohen et al. (2001), the AQ was administered to a clinical sample of adults diagnosed with Asperger syndrome or high-functioning autism (HFA; n = 58) alongside non-clinical controls (n = 174 random general population; n = 840 Cambridge University undergraduates). The clinical group obtained a mean score of 35.8 (SD = 6.5), significantly exceeding the general population control mean (M = 16.4, SD = 6.3; t = 20.3, p < .0001, Cohen’s d > 3.0). Furthermore, 79.3% of the clinical group scored ≥ 32, whereas only 2.3% of general population controls reached this threshold.

In a clinical validation study conducted by Woodbury-Smith et al. (2005) among adults presenting consecutively to a diagnostic clinic, an AQ cutoff of 26 yielded an optimal balance of sensitivity (0.95), specificity (0.52), positive predictive value (0.84), and negative predictive value (0.78), establishing the scale’s robust utility as a primary clinical triage mechanism.

7.2. Convergent and Discriminant Validity

Convergent validity is substantiated by strong correlations between the AQ and other validated instruments assessing social communication and broad autism phenotype characteristics:

• Negative correlations with the Empathy Quotient (EQ): r = −0.56 to −0.67 (Baron-Cohen & Wheelwright, 2004), confirming that lower empathizing capacity scales reliably with elevated autistic traits.
• Moderate positive correlations with the Systemizing Quotient (SQ): r = 0.28 to 0.41, supporting the EMB theoretical axis.
• Strong convergent correlation with the Broad Autism Phenotype Questionnaire (BAPQ): r = 0.68 to 0.76 (Hurley et al., 2007).
• Discriminant validity: The AQ differentiates autistic features from generalized neuroticism and primary depressive disorders. While moderate correlations exist with social anxiety measures (e.g., Social Phobia Inventory; r ≈ 0.45–0.55), factor analyses demonstrate that the communication and detail-orientation facets of the AQ remain psychometrically distinct from affective anxiety symptom clusters.

8. Reliability

The AQ demonstrates strong psychometric reliability across internal consistency metrics, split-half analyses, and longitudinal test-retest investigations.

8.1. Internal Consistency

In the original Baron-Cohen et al. (2001) investigation, the total AQ scale demonstrated excellent internal consistency, with a Cronbach’s alpha of α = 0.82 across the combined cohort. Within individual subdomains, Cronbach’s alpha coefficients were moderate to satisfactory:

• Social Skills: α = 0.70
• Attention Switching: α = 0.67
• Attention to Detail: α = 0.63
• Communication: α = 0.65
• Imagination: α = 0.65

When evaluated using continuous Likert scoring (1–4) rather than binary scoring (0–1), internal reliability improves substantially due to preserved item variance. Austin (2005) reported a total scale Cronbach’s alpha of α = 0.86, with subscale alphas ranging from 0.65 (Imagination) to 0.77 (Social Skills). Subsequent cross-cultural validations (e.g., Dutch version by Hoekstra et al., 2008; Japanese version by Wakabayashi et al., 2006) reported total scale internal consistency estimates of α = 0.79 to 0.88.

8.2. Test-Retest Stability

Test-retest reliability has been documented over various time horizons, demonstrating high stability consistent with trait-based personality and cognitive endophenotypes:

• Baron-Cohen et al. (2001) evaluated a subsample of n = 17 autistic participants retested over an interval of 2 to 3 weeks, yielding a test-retest correlation of r = 0.70 (p < .002).
• In a larger non-clinical student sample evaluated over a 2-month interval, test-retest reliability reached r = 0.85 (p < .001).
• Intraclass correlation coefficients (ICC) across language adaptations range consistently between 0.78 and 0.89, establishing that AQ scores remain temporally stable in the absence of targeted psychological or behavioral interventions.

9. Factor Analysis

The latent structural composition of the AQ has been the subject of extensive empirical examination via exploratory (EFA) and confirmatory factor analysis (CFA), leading to refined psychometric models.

9.1. Original Theoretical Five-Factor Model

Baron-Cohen et al. (2001) conceptualized the AQ as a theoretical five-factor structure matching the five hypothesized domains: Social Skills, Attention Switching, Attention to Detail, Communication, and Imagination. However, subsequent structural equation modeling investigations in both general and clinical populations revealed that the original five factors exhibited variable empirical orthogonality, with several items exhibiting cross-loadings, particularly between Communication and Social Skills.

9.2. Hoekstra et al. (2008) Hierarchical and 2-Factor / 5-Factor Solutions

In a definitive psychometric modeling study involving large non-clinical (N = 1,003) and clinical samples, Hoekstra et al. (2008) demonstrated that a revised five-factor model omitting problematic, low-loading items yielded superior model fit (Root Mean Square Error of Approximation [RMSEA] = 0.046; Comparative Fit Index [CFI] = 0.902). More critically, second-order hierarchical modeling identified two higher-order foundational dimensions:

1. Social Behavior: Encompassing Social Skills, Communication, Attention Switching, and Imagination. This superordinate factor reflects the core socio-communicative impairments central to the diagnostic criteria of autism.
2. Attention to Detail: Primarily composed of local perceptual acuity, numeric fascination, and pattern recognition. This second factor operates largely independently of social-communicative features, accounting for cognitive strengths and sensory profiling.

9.3. Austin (2005) and Other Factor Analytic Variants

Austin (2005) conducted principal components analysis with varimax rotation on a 4-point scoring distribution, identifying a robust three-factor solution representing (1) Social Interaction and Communication, (2) Fascination with Numbers/Patterns, and (3) Detail/Change Tolerance. Item loadings on primary latent constructs generally range from 0.40 to 0.74, with minimal residual error when assessed using polychoric correlation matrices tailored for ordinal Likert responses.

10. Instrument / Measurement Tool

• Instrument Name: Autism Spectrum Quotient (AQ) – Self-administered
• Target Population: Adolescents and adults aged 16 years and older with normal to superior cognitive functioning (IQ ≥ 85)
• Administration Format: Self-report questionnaire; available in paper-and-pencil, computer-based, or mobile digital formats
• Completion Time: Approximately 8 to 12 minutes
• Number of Items: 50 statements
• Item Formats: First-person declarative behavioral and cognitive statements
• Response Scale: 4-point ordinal forced-choice format:
  • Definitely agree
  • Slightly agree
  • Slightly disagree
  • Definitely disagree
• Subscale Item Distribution:
  • Social Skills (10 items): 1, 11, 13, 15, 22, 36, 44, 45, 47, 48
  • Attention Switching (10 items): 2, 4, 10, 16, 25, 32, 34, 37, 43, 46
  • Attention to Detail (10 items): 5, 6, 9, 12, 19, 23, 28, 29, 30, 49
  • Communication (10 items): 7, 17, 18, 26, 27, 31, 33, 35, 38, 39
  • Imagination (10 items): 3, 8, 14, 20, 21, 24, 40, 41, 42, 50
• Scoring Methodology:
  • Original Binary Scoring (0–50): Endorsement of an autistic trait yields 1 point, regardless of whether “definitely” or “slightly” was selected; non-endorsement yields 0 points.
    • Positively keyed items (1 point for “Definitely agree” or “Slightly agree”): 2, 4, 5, 6, 7, 9, 12, 13, 16, 18, 19, 20, 21, 22, 23, 26, 33, 35, 39, 41, 42, 43, 45, 46.
    • Reverse keyed items (1 point for “Definitely disagree” or “Slightly disagree”): 1, 3, 8, 10, 11, 14, 15, 17, 24, 25, 27, 28, 29, 30, 31, 32, 34, 36, 37, 38, 40, 44, 47, 48, 49, 50.
  • Continuous Likert Scoring (50–200): Items are scored 1, 2, 3, 4 (or 0, 1, 2, 3) along the trait gradient to preserve full variance for parametric psychometric analyses.
• Clinical Interpretation Cutoffs (Binary Scoring):
  • 0 – 10: Significantly below average autistic traits.
  • 11 – 21: Average/normative range (typical general population mean ≈ 16.4).
  • 22 – 25: Mildly elevated autistic traits; common in STEM specialists.
  • 26 – 31: Clinically elevated autistic traits (screening threshold suggested by Woodbury-Smith et al., 2005; warrants clinical diagnostic referral).
  • 32 – 50: Highly indicative of Asperger syndrome or high-functioning autism (original Baron-Cohen cutoff; 79.3% of clinical adult autism sample ≥ 32).

11. Permissions & Fee and Test Year

The Autism Spectrum Quotient was first published in 2001 by Simon Baron-Cohen, Sally Wheelwright, Richard Skinner, Jacqueline Martin, and Emma Clubley. The instrument is copyrighted by the Autism Research Centre (ARC), University of Cambridge.

Licensing and Accessibility: In accordance with the Autism Research Centre’s open-science mission, the AQ is released as an open-access psychometric instrument for clinical, academic, educational, and non-commercial research applications. No licensing fees or royalties are required to administer, score, or reprint the measure for scientific or clinical purposes, provided appropriate academic attribution is maintained. Commercial applications, programmatic digital integrations, or inclusion within proprietary software platforms require explicit written authorization from the Autism Research Centre. The instrument, translation guidelines, and diagnostic resources can be accessed via the official ARC portal: www.autismresearchcentre.com.

12. References

Austin, E. J. (2005). Personality, intelligence and the Autism-Spectrum Quotient (AQ): A study with an Edinburgh student sample. Personality and Individual Differences, 39(8), 1471–1481. https://doi.org/10.1016/j.paid.2005.06.005

Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in Cognitive Sciences, 6(6), 248–254. https://doi.org/10.1016/S1364-6613(02)01904-6

Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a “theory of mind”? Cognition, 21(1), 37–46. https://doi.org/10.1016/0010-0277(85)90022-8

Baron-Cohen, S., & Wheelwright, S. (2004). The Empathy Quotient: An investigation of adults with Asperger syndrome or high functioning autism, and normal sex differences. Journal of Autism and Developmental Disorders, 34(2), 163–175. https://doi.org/10.1023/B:JADD.0000022607.19833.00

Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The Autism-Spectrum Quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31(1), 5–17. https://doi.org/10.1023/A:1005653411471

Hoekstra, R. A., Bartels, M., Cath, D. C., & Boomsma, D. I. (2008). Factor structure, reliability and criterion validity of the Autism-Spectrum Quotient (AQ): A study in Dutch population and patient samples. Journal of Autism and Developmental Disorders, 38(8), 1555–1566. https://doi.org/10.1007/s10803-008-0538-x

Hurley, R. S., Losh, M., Parlier, M., Reznick, J. S., & Piven, J. (2007). The broad autism phenotype questionnaire. Journal of Autism and Developmental Disorders, 37(9), 1679–1690. https://doi.org/10.1007/s10803-006-0299-3

Sucksmith, E., Allison, C., Baron-Cohen, S., Chakrabarti, B., & Hoekstra, R. A. (2013). Empathy and emotion recognition in people with autism, first-degree relatives, and controls. Neuropsychologia, 51(1), 98–105. https://doi.org/10.1016/j.neuropsychologia.2012.11.013

Wakabayashi, A., Baron-Cohen, S., Wheelwright, S., & Tojo, Y. (2006). The Autism-Spectrum Quotient (AQ) in Japan: A cross-cultural comparison. Journal of Autism and Developmental Disorders, 36(2), 263–270. https://doi.org/10.1007/s10803-005-0061-3

Woodbury-Smith, M. R., Robinson, J., Wheelwright, S., & Baron-Cohen, S. (2005). Screening adults for Asperger syndrome using the AQ: A preliminary study of its diagnostic validity in clinical practice. Journal of Autism and Developmental Disorders, 35(3), 331–335. https://doi.org/10.1007/s10803-005-3300-7

13. Items of the Scale (Questionnaire)