Autism Symptom Dimensions Questionnaire (ASDQ)

Abstract

The Autism Symptom Dimensions Questionnaire (ASDQ) is a psychometrically validated, open-source informant-report instrument engineered by Thomas W. Frazier and colleagues (2023) to assess core symptomatology of autism spectrum disorder (ASD) across children and adolescents aged 2 to 17 years. Conceived directly to align with the diagnostic architecture of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), the ASDQ remediates long-standing operational limitations observed in traditional screening instruments, such as floor and ceiling effects, proprietary paywalls, and poor differentiation among distinct clinical sub-dimensions. The instrument comprises 39 items evaluated on an electronic 5-point Likert frequency scale ranging from 1 (“never”) to 5 (“very often”), accompanied by a non-scorable “Not Applicable or Not Able to Rate” option. Psychometric analyses demonstrate an overarching general bifactor structure encompassing an overarching ASD dimension and nine correlated yet distinct lower-order symptom factors: Basic Social Communication, Affiliation, Perspective Taking, and Peer Relationships within the social communication domain; and Repetitive Motor Behavior, Sensory Interests, Insistence on Sameness, Sensory Sensitivities, and Restricted Interests within the restricted and repetitive behavior domain. The general ASD factor exhibits exceptional model-based composite reliability (ω = 0.97; Cronbach’s α = 0.95), with lower-order subscale reliabilities demonstrating strong internal consistency (ω ≥ 0.85; α ≥ 0.75). Conditional item response theory (IRT) reliability confirms robust measurement precision (≥ 0.90) across a wide latent trait spectrum (θ = −2.1 to +4.0). Furthermore, strict measurement invariance across child sex, age groups, race, and ethnicity substantiates its cross-demographic utility in clinical characterization, epidemiologic surveillance, and longitudinal intervention monitoring.

Keywords

Autism Symptom Dimensions Questionnaire, ASDQ, Autism Spectrum Disorder, DSM-5, Psychometrics, Bifactor Model, Social Communication, Restricted and Repetitive Behaviors, Informant Report, Measurement Invariance

Authors

The Autism Symptom Dimensions Questionnaire was developed through a collaborative multicenter initiative led by academic clinical psychologists, child developmental pediatricians, and clinical neuroscientists:

• Thomas W. Frazier, Ph.D. (ORCID: 0000-0002-6951-2667) — Department of Psychology, John Carroll University, University Heights, Ohio, United States. Correspondence: [email protected].
• Anastasia Dimitropoulos, Ph.D. — Department of Psychological Sciences, Case Western Reserve University, Cleveland, Ohio, United States.
• Leonard Abbeduto, Ph.D. — Department of Psychiatry and Behavioral Sciences and the MIND Institute, University of California, Davis, Sacramento, California, United States.
• Melissa Armstrong-Brine, Ph.D. — MetroHealth Autism Assessment Clinic, MetroHealth Medical Center, Cleveland, Ohio, United States.
• Shanna Kralovic, D.O. — Division of Developmental Behavioral Pediatrics and Psychology, University Hospitals Rainbow Babies & Children’s Hospital / Case Medical Center, Cleveland, Ohio, United States.
• Andy Shih, Ph.D. — Autism Speaks, New York, New York, United States.
• Antonio Y. Hardan, M.D. — Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California, United States.
• Eric A. Youngstrom, Ph.D. — Department of Psychology and Neuroscience, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States.
• Mirko Uljarević, M.D., Ph.D. — Melbourne School of Psychological Sciences, Faculty of Medicine, Dentistry, and Health Sciences, University of Melbourne, Melbourne, Victoria, Australia.
• Quadrant Biosciences – As You Are Team — Institutional collaborating clinical research consortium.

Purpose

The primary clinical and scientific purpose of the ASDQ is to provide a comprehensive, open-access, psychometrically robust measurement framework tailored directly to the core symptom domains of autism spectrum disorder as conceptualized by the American Psychiatric Association in DSM-5. Historically, autism diagnostic characterization and symptom tracking have relied heavily upon instruments designed during the DSM-IV or ICD-10 eras, such as the Social Communication Questionnaire (SCQ) or the Social Responsiveness Scale (SRS). Although foundational, these legacy instruments exhibit several notable operational limitations in contemporary clinical and translational research:

• Diagnostic Realignment: Many legacy questionnaires conflate generalized developmental delay, severe intellectual disability, or broader externalizing and internalizing behavioral problems with core autism symptomatology, hindering precise mapping onto DSM-5 Criteria A (persistent deficits in social communication and social interaction) and Criteria B (restricted, repetitive patterns of behavior, interests, or activities).
• Granularity and Structural Parsimony: Traditional screeners frequently aggregate heterogeneous presentations into a unitary score or two broad subscales, masking clinically critical variation across fine-grained phenotypic dimensions (e.g., distinguishing sensory hyper-reactivity from stereotypic motor mannerisms or cognitive insistence on sameness).
• Floor and Ceiling Psychometric Restrictions: Standard scales often lack conditional measurement precision at the extreme poles of symptom severity, creating significant measurement error when assessing minimally verbal children with high support needs or highly cognitively capable adolescents with subtle social deficits.
• Open Science and Accessibility: Widespread clinical and research utilization of established rating scales is frequently constrained by substantial per-protocol licensing fees, proprietary scoring algorithms, and paper-and-pencil delivery constraints that impede scalable digital phenotyping and decentralized clinical trials.

The ASDQ was explicitly engineered to overcome these barriers. Originally drafted as an exploratory 33-item inventory, the scale was iteratively expanded to 39 items to fully capture empirically supported sub-constructs established across recent quantitative structural reviews. By utilizing caregiver and informant observations across naturalistic home, community, and educational environments, the ASDQ functions both as a granular phenotypic profiling instrument in clinical neurodevelopmental evaluations and as an efficient, sensitive outcome measure for pharmacological, behavioral, and psychoeducational clinical trials.

Psychological Construct

The ASDQ conceptualizes autism spectrum disorder not as an undifferentiated, unitary entity, but as a hierarchical, multifaceted neurodevelopmental construct. Consistent with contemporary structural equation modeling and bifactor psychometrics, the instrument captures an overarching general ASD liability dimension while simultaneously decomposing observable behavioral manifestations into nine distinct, lower-order symptom dimensions nested within two broad DSM-5 domains.

Domain 1: Social Communication and Interaction

This domain captures reciprocal socio-emotional processes, communicative adaptations, and interpersonal engagement strategies across four granular dimensions:

• Basic Social Communication: Encompasses fundamental nonverbal and verbal signaling mechanisms necessary for reciprocal engagement. Behaviors include coordinating eye contact during conversational turns, utilizing communicative and emphatic gestures (pointing, waving, nodding), matching facial expressions to emotional contexts, and communicating inner affective states with clarity.
• Affiliation: Evaluates intrinsic motivation for interpersonal proximity, social orientation, and belonging. It measures whether the individual actively prefers companionship over isolation, seeks emotional and physical closeness with family and peers, initiates spontaneous unprompted interaction, and demonstrates that ongoing interpersonal relationships carry personal value.
• Perspective Taking: Reflects cognitive and affective Theory of Mind (ToM) capacities. This subscale measures the child’s ability to infer what others are thinking or feeling, read subtle social cues (e.g., changes in vocal cadence or postural stance), understand that behavioral expectations shift according to specific contexts (e.g., volume control in formal versus playful settings), and avoid overly literal interpretations of idioms, sarcasm, and playful teasing.
• Peer Relationships: Focuses specifically on egalitarian, reciprocal interactions outside the structured caregiver-child dynamic. Indicators assess engagement in reciprocal back-and-forth play (such as cooperative rule-based games), sharing mutual enjoyment, participating in reciprocal conversational exchanges concerning topics beyond personal fixations, and sustaining meaningful reciprocal friendships with two or more peers.

Domain 2: Restricted and Repetitive Behaviors (RRBs)

This domain quantifies non-functional routines, sensory processing alterations, and focused motoric patterns across five specialized dimensions:

• Repetitive Motor Behavior: Captures classic stereotyped and repetitive motor mannerisms, including hand flapping, finger flicking, rocking, spinning, whole-body posturing, and purposeless, repetitive object manipulations (e.g., wheel spinning, repetitive light-switch toggling, or lining objects in strict geometric sequences).
• Sensory Interests: Reflects sensory-seeking behaviors and intense sensory fascinations. It assesses excessive preoccupation with visual patterns, movement dynamics (e.g., staring intently at rotating ceiling fans or cascading water), fixation on tactile textures, unusual olfactory exploration, or an idiosyncratic preference for interacting exclusively with isolated component parts of objects rather than their functional wholes.
• Insistence on Sameness: Quantifies behavioral, cognitive, and environmental rigidity. Items evaluate distress provoked by unexpected disruptions to daily routines, marked difficulty handling environmental transitions, rigid adherence to idiosyncratic behavioral rituals (e.g., specific transit routes or rigid food ordering), and severe cognitive inflexibility when challenged to change plans.
• Sensory Sensitivities: Measures sensory hyper-reactivity across sensory modalities. Core indicators capture aversive responses or behavioral meltdown triggers resulting from loud auditory stimuli, intense lighting, specific tactile fabrics, crowded or bustling sensory environments, or distinct olfactory and taste sensations.
• Restricted Interests: Quantifies circumscribed, highly focused preoccupations that are abnormal in intensity or narrowness of focus. Manifestations include repetitive verbal perseveration on circumscribed subjects, obsessive engagement with idiosyncratic topics that peers find uninteresting, excessive memorization of factual lists or technical manuals, and an inability to disengage attention toward alternative tasks.

Theoretical Framework

The theoretical architecture of the ASDQ is anchored in contemporary developmental psychopathology, structural quantitative genetics, and cognitive neuroscience models of neurodevelopmental divergence.

The Hierarchical / Bifactor Model of Neurodevelopment

For decades, autism assessment oscillated between two competing conceptual paradigms: a strictly categorical model viewing ASD as a bounded disease entity (typified by earlier iterations of the DSM and ICD), and a unidimensional continuous model viewing autism as a solitary trait spectrum running through the general population into severe clinical manifestations. The ASDQ integrates these paradigms through a bifactor psychometric model. In this framework, common variance across all autistic behaviors is driven by a broad, general neurodevelopmental liability factor, while significant, clinically meaningful unique variance is partitioned into specific lower-order dimensions. This architecture recognizes that while individuals share an overarching clinical autism phenotype, two individuals with identical general severity may exhibit radically different lower-order configurations (e.g., one manifesting intense sensory hyper-reactivity alongside intact affiliation, and another displaying profound peer isolation with marked motor stereotypies).

Social Motivation and Predictive Processing Formulations

The specific subscales of the ASDQ reflect modern neurocognitive theories:

• The Social Motivation Theory of Autism: Posits that primary deficits in reward circuitry (e.g., ventral striatum and orbitofrontal cortex) attenuate the salience of social stimuli early in life, cascading into diminished social attention, reduced peer interactions, and underdeveloped Theory of Mind. The ASDQ captures this pathway by isolating “Affiliation” and “Peer Relationships” from basic mechanics of communication.
• Predictive Processing and Bayesian Brain Accounts: Conceptualize autism as an imbalance in the weighting of sensory prediction errors relative to prior expectations. Hyper-precise prediction errors generate overwhelming sensory environments (captured by the “Sensory Sensitivities” and “Sensory Interests” subscales) and drive secondary compensatory adaptations wherein the individual enforces strict environmental predictability to reduce perceptual chaos (captured by the “Insistence on Sameness” subscale).

Validity

The psychometric validation of the ASDQ was conducted across rigorous empirical investigations involving large cohorts of caregivers of children with verified ASD, neurotypical controls, and clinical contrast groups with non-autism developmental conditions (e.g., intellectual disability, ADHD, language disorders).

Convergent and Criterion Validity

The ASDQ demonstrated robust convergent validity when benchmarked against established, gold-standard diagnostic and dimensional screening instruments. ASDQ total scores correlated strongly with the Social Communication Questionnaire (SCQ; r > 0.70) and total behavioral symptom scores on the Strengths and Difficulties Questionnaire (SDQ). Furthermore, total scores differentiated sharply between children with formal clinical ASD diagnoses and neurotypical counterparts, demonstrating powerful diagnostic discrimination.

Discriminant and Divergent Validity

Discriminant validity was established through systematic comparisons against non-autism behavioral constructs. ASDQ scores demonstrated moderate correlations (ranging from r = 0.39 to 0.50) with SDQ internalizing and externalizing subscales, global executive functioning ratings, and adaptive behavioral scales. This moderate magnitude confirms that while secondary psychiatric and cognitive challenges frequently co-occur with autism, the ASDQ specifically captures unique variance attributable to core autistic phenotypes rather than generalized childhood emotional or behavioral disturbance.

Predictive and Diagnostic Accuracy

Receiver operating characteristic (ROC) analyses examining the predictive validity of the ASDQ yielded strong classification metrics. The scale demonstrated high sensitivity and specificity in discriminating children with ASD from neurotypical youth (areas under the curve [AUC] exceeding 0.90) as well as significant discriminative accuracy when distinguishing ASD from non-autistic neurodevelopmental and psychiatric conditions in clinically enriched, at-risk samples.

Reliability

The ASDQ exhibits exceptional reliability across classical test theory indices, modern item response theory parameters, and structural equation modeling evaluations.

Internal Consistency Reliability

Model-based composite reliability, assessed via McDonald’s omega (ω), demonstrated extraordinary precision for the overarching general ASD factor (ω = 0.97). Each of the nine lower-order specific symptom dimensions exceeded rigorous standards for research and clinical subscale interpretation, with all omega coefficients satisfying ω ≥ 0.85. Evaluated via Cronbach’s alpha (α), total scale internal consistency was α = 0.95, while all nine individual subscales yielded alpha coefficients ranging from α = 0.75 to α = 0.91, confirming uniform item-scale coherence across the instrument.

Conditional Measurement Precision (IRT)

Applying modern latent trait modeling, conditional standard errors of measurement were computed across the continuum of autistic severity (θ). The ASDQ total scale maintained high conditional reliability (≥ 0.90) across an expansive latent trait spectrum spanning from θ = −2.1 standard deviations below the mean up to θ = +4.0 standard deviations above the mean. This wide reliability window indicates that the instrument retains measurement precision across mild, moderate, and profound levels of clinical presentation without succumbing to the ceiling truncation common in brief screening measures.

Factor Analysis

The internal latent structure of the 39-item ASDQ was rigorously tested using advanced structural equation modeling, comparing standard confirmatory factor analysis (CFA) with exploratory structural equation modeling (ESEM) bifactor architectures.

Bifactor ESEM Architecture

Empirical analyses demonstrated that an a priori hypothesized nine specific symptom factor plus one general bifactor ESEM solution provided optimal fit to the data, outperforming both unidimensional and traditional correlated-trait specifications. While standard orthogonal bifactor CFA models (Model 9b) yielded adequate fit, the bifactor ESEM approach accounted more accurately for minor, clinically expected item cross-loadings across complex social-communicative behaviors. Within this structural framework, the general ASD factor accounted for approximately 42% of the common variance, confirming that while a strong general autistic liability exists, the remaining 58% of common variance is distributed across the nine specific lower-order dimensions.

Measurement Invariance

A critical psychometric strength of the ASDQ is its demonstrated measurement equivalence across demographic subpopulations. Multi-group confirmatory factor analysis established that the nine-factor bifactor model achieved strict measurement invariance across:

• Biological Sex: Factor loadings, item thresholds, and unique residual variances operated equivalently across male and female children, confirming that sex differences in ASDQ scores reflect genuine phenotypic divergence rather than measurement artifact or reporting bias.
• Developmental Age: Invariance was sustained across preschool (2–5 years), school-age (6–12 years), and adolescent (13–17 years) cohorts.
• Race and Ethnicity: Strict invariance criteria held across diverse racial and ethnic groups, validating the tool for multicultural clinical practice and nationwide research registries.

Instrument / Measurement Tool

• Instrument Name: Autism Symptom Dimensions Questionnaire (ASDQ)
• Instrument Type: Standardized Informant-Report Rating Scale / Behavioral Questionnaire
• Authors: Thomas W. Frazier, Ph.D., Anastasia Dimitropoulos, Ph.D., Leonard Abbeduto, Ph.D., Melissa Armstrong-Brine, Ph.D., Shanna Kralovic, D.O., Andy Shih, Ph.D., Antonio Y. Hardan, M.D., Eric A. Youngstrom, Ph.D., Mirko Uljarević, M.D., Ph.D., and Quadrant Biosciences – As You Are Team
• Publication Year: 2023
• Target Population: Children and adolescents aged 2 to 17 years exhibiting or evaluated for social communication difficulties or repetitive behaviors
• Respondent / Informant: Parents, biological mothers, primary legal guardians, or knowledgeable caregivers with longitudinal observation of the child
• Item Count: 39 behavioral items
• Administration Format: Electronic administration (computer, tablet, or web interface); paper-and-pencil compatible
• Response Scale: 5-point Likert frequency scale: 1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, 5 = Very Often. An additional non-scorable categorical response option (“Not Applicable or Not Able to Rate”) is provided for behaviors outside the child’s developmental capacity or unobserved by the caregiver.
• Domain and Subscale Structure:
  • Domain A: Social Communication and Interaction (17 Items)
    • Basic Social Communication
    • Affiliation
    • Perspective Taking
    • Peer Relationships
  • Domain B: Restricted and Repetitive Behaviors (22 Items)
    • Repetitive Motor Behavior
    • Sensory Interests
    • Insistence on Sameness
    • Sensory Sensitivities
    • Restricted Interests
• Scoring Approach: Items are keyed such that higher scores reflect greater autistic symptomatology and behavioral frequency (pro-social items are reverse-scored according to standard scoring protocols). Raw sum scores can be calculated for each of the nine lower-order symptom subscales, the two DSM-5 overarching domains, and an omnibus total ASD severity score. Standardized T-scores and percentile ranks derived from normative validation samples are utilized for clinical characterization.

Permissions & Fee and Test Year

• Test Year: 2023
• Copyright: © 2021–2023 Thomas W. Frazier, Ph.D.
• Licensing: Released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.
• Permissions: Free for public use in clinical practice, non-commercial scientific research, academic education, and health system monitoring without royalty fees, provided appropriate academic attribution is cited.
• Commercial Use: Commercial redistribution or inclusion in proprietary software suites requires explicit authorization from the primary copyright holder.
• Digital Repository / Open Science Framework: Instrument materials, supplemental data, and scoring scripts are publicly archived at OSF (Open Science Framework: osf.io/pyq8r/).

References

• American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). American Psychiatric Publishing. https://doi.org/10.1176/appi.books.9780890425596
• Frazier, T. W., Dimitropoulos, A., Abbeduto, L., Armstrong-Brine, M., Kralovic, S., Shih, A., Hardan, A. Y., Youngstrom, E. A., Uljarević, M., & Quadrant Biosciences – As You Are Team. (2023). The Autism Symptom Dimensions Questionnaire: Development and psychometric evaluation of a new, open-source measure of autism symptomatology. Developmental Medicine & Child Neurology, 65(8), 1081–1092. https://doi.org/10.1111/dmcn.15497
• Goodman, R., Meltzer, H., & Bailey, V. (1998). The Strengths and Difficulties Questionnaire: A pilot study on the validity of the self-report version. European Child & Adolescent Psychiatry, 7(3), 125–130. https://doi.org/10.1007/s007870050057
• Rutter, M., Bailey, A., & Lord, C. (2003). The Social Communication Questionnaire: Manual. Western Psychological Services.
• Uljarević, M., Frazier, T. W., Phillips, J. M., Jo, B., Littlefield, S. C., & Hardan, A. Y. (2020). Mapping the fine-grained phenotypic profile of restricted and repetitive behaviors in autism: A bifactor approach. Molecular Autism, 11(1), 1–13. https://doi.org/10.1186/s13229-020-00358-0

13. Items of the Scale (Questionnaire)