Animated Activity Questionnaire

Abstract

The Animated Activity Questionnaire (AAQ) is an innovative, computer-administered patient-reported outcome measure (PROM) developed to assess limitations in basic activities of daily living (ADL) among patients with hip and knee osteoarthritis (OA). Traditional paper-and-pencil functional status questionnaires frequently suffer from cross-cultural translation barriers, differing semantic interpretations of subjective rating scales (e.g., distinguishing between “moderate” and “severe” difficulty), and cognitive recall biases. To address these measurement challenges, the AAQ employs standardized computer animations as visual response categories. The instrument consists of 17 core daily activities primarily involving the lower extremities, including walking indoors, navigating stairs, transferring to and from various seating surfaces (e.g., chairs, low chairs, sofas, and toilets), bed mobility, and vehicle ingress/egress. For each activity, patients view 3 to 5 animated video vignettes portraying an avatar performing the task with varying degrees of movement impairment and compensatory strategies, ranging from unrestricted, normal physiological movement to severely impaired execution. Patients select the animation that best matches their own habitual manner of performance. Psychometrically calibrated using modern Item Response Theory (IRT), specifically the Graded Response Model, raw animation selections are converted into a linear standardized scale ranging from 0 to 100, where 0 represents absolute limitation and 100 denotes completely unimpaired physical functioning. Extensive validation studies demonstrate excellent unidimensionality, high internal consistency, robust test-retest reliability (intraclass correlation coefficients > 0.90), and superior convergent validity when compared against performance-based objective mobility tests (such as the Timed Up and Go and the 6-Minute Walk Test) relative to conventional text-based questionnaires.

Keywords

Animated Activity Questionnaire, AAQ, osteoarthritis, physical functioning, patient-reported outcome measure, video vignettes, item response theory, lower extremity mobility, activities of daily living, psychometrics.

Authors

The Animated Activity Questionnaire was developed and validated by a multidisciplinary consortium of rheumatologists, physical therapists, epidemiologists, and psychometricians led by:

• Wilfred F. Peter, PT, PhD — Department of Rheumatology, Leiden University Medical Center (LUMC), Leiden, Netherlands; and Reade Center for Rehabilitation and Rheumatology, Amsterdam, Netherlands.
• Thea P. M. Vliet Vlieland, MD, PhD — Department of Orthopaedics, Rehabilitation and Physical Therapy, Leiden University Medical Center (LUMC), Leiden, Netherlands.
• Leo D. Roorda, MD, PT, PhD — Amsterdam Rehabilitation Research Center | Reade, Amsterdam, Netherlands.
• Martin van der Esch, PT, PhD — Reade Center for Rehabilitation and Rheumatology, Amsterdam, Netherlands.
• Martijn P. M. Steultjens, PhD — School of Health and Life Sciences, Glasgow Caledonian University, Glasgow, United Kingdom.
• Caroline B. Terwee, PhD — Department of Epidemiology and Data Science, Amsterdam University Medical Centers, Vrije Universiteit Amsterdam, Amsterdam, Netherlands.

Inquiries regarding institutional implementations, collaborative research protocols, and software access can be directed to the corresponding developmental team through Leiden University Medical Center or via the official platform registry at http://www.myaaq.com.

Purpose

The primary purpose of the Animated Activity Questionnaire is to deliver an objective, visually standardized assessment of perceived physical functioning in individuals experiencing degenerative joint conditions of the lower limb, specifically hip and knee osteoarthritis. Physical limitation is a cardinal feature of joint pathology, yet the accurate quantification of limitation has historically represented a complex psychometric challenge in orthopedic and rheumatologic care.

Conventional physical functioning scales—such as the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), the Hip Disability and Osteoarthritis Outcome Score (HOOS), and the Knee Injury and Osteoarthritis Outcome Score (KOOS)—depend on text-based items pairing functional queries with abstract verbal descriptor rating scales (e.g., “no difficulty,” “mild,” “moderate,” “severe,” or “extreme”). This traditional approach creates several prominent methodological vulnerabilities:

• Subjective Frame of Reference: Two patients with identical biomechanical impairment may rate their limitation differently based on personal stoicism, baseline physical expectations, educational background, or personal definitions of “difficulty.”
• Language and Literacy Barriers: Text-heavy instruments can be challenging for populations with low health literacy, non-native speakers, or individuals with cognitive or linguistic limitations, often demanding rigorous, costly cross-cultural adaptations that risk subtle semantic drift.
• Low Concordance with Objective Performance: Empirical research consistently demonstrates that correlations between text-based self-reports and direct performance-based tests (e.g., gait velocity, stair ascent cadence, sit-to-stand transitions) are only low-to-moderate, indicating that patients and performance assessors often capture distinct operational facets of functional limitation.

The AAQ addresses these limitations by substituting written descriptive response options with computer-generated animated videos illustrating exact kinematic alterations, speed differentials, compensatory adjustments (such as using handrails or pushing off thighs), and biomechanical deviations. By anchoring response options to objective visual exemplars, the instrument mitigates semantic ambiguity, minimizes cognitive translation demands, and establishes a common reference metric. The tool is applied in both clinical settings (pre- and post-operative joint arthroplasty monitoring, physical therapy outcome tracking) and clinical trials evaluating pharmacological, conservative, and surgical interventions.

Psychological Construct

The Animated Activity Questionnaire assesses self-reported activity limitations in basic lower extremity physical functioning. In accordance with the World Health Organization’s International Classification of Functioning, Disability and Health (ICF), the construct is located strictly within the “Activities” component, demarcated from “Body Functions and Structures” (e.g., radiographic joint degeneration, joint space narrowing, nociceptive pain) and “Participation” (e.g., employment, social roles, recreational pursuits).

Physical functioning within the AAQ is operationalized as an individual’s capacity to execute essential postural changes, weight-bearing transfers, and locomotor tasks that form the foundation of independent community dwelling. This construct is structured across several behavioral dimensions:

• Locomotion and Ambulatory Navigation: Assessed via tasks such as walking indoors across flat, unobstructed terrain, walking outdoors on standard sidewalks, and negotiating unpredictable surfaces (e.g., cobblestones, gravel, or uneven natural turf). This sub-dimension captures dynamic balance, postural sway, gait symmetry, stride velocity, and adaptive foot clearance.
• Stair Negotiation (Ascent and Descent): Evaluates eccentric and concentric quadriceps load capacity, knee flexion-extension range of motion, and reliance on structural external support. The animations delineate varying degrees of functional compromise, including step-over-step ascent/descent, step-by-step strategies (two feet on one tread), single- or double-arm handrail dependency, and pronounced pelvic drop or torso lean.
• Seated Transfers Across Graded Heights: Captures transfer actions involving rising from and sitting down upon standard chairs, low-seated chairs, and deeply cushioned living room sofas. Biomechanically, these movements require hip and knee extension moment generation, core stabilization, and neuromuscular control. Visual representations display variations ranging from smooth, hands-free sit-to-stand movements to compensatory maneuvers such as hand support on thighs, armrest propulsion, rocking forward momentum, or abrupt uncoordinated descent.
• Toileting Transfers: Focuses specifically on personal hygiene mobility, highlighting functional independence in confined, standardized sanitary environments.
• Supine-to-Erect and Recumbent Transfers: Captured through getting into and out of bed, requiring trunk rotation, bilateral lower extremity abduction/flexion coordination, and horizontal-to-vertical center-of-mass translation.
• Vehicular Transfers: Evaluates entry and exit maneuvers in passenger automobiles, reflecting complex functional demands involving narrow door clearances, seated hip flexion below knee level, pivoting, and weight-bearing repositioning.

While biomechanically driven, the underlying construct is intrinsically psychological. It involves self-perception, self-efficacy, mental representations of movement capabilities, and fear-avoidance heuristics. The respondent must evaluate the depicted animation against their internal schematic representation of their daily motor patterns, requiring introspective kinematic appraisal.

Theoretical Framework

The design and conceptual architecture of the AAQ are informed by three theoretical frameworks: cognitive models of survey methodology, modern psychometric measurement theory, and the bio-psycho-social model of functioning.

First, the instrument addresses the Cognitive Model of the Survey Response Process, formulated by Tourangeau, Rips, and Rasinski (2000). In standard questionnaires, a respondent must progress through four distinct cognitive stages: (1) comprehension of the question, (2) retrieval of relevant memories, (3) integration and judgment, and (4) mapping the judgment onto the provided response format. Traditional PROMs encounter significant noise at the mapping stage: a patient may comprehend the functional activity (e.g., “walking down stairs”) and accurately retrieve past instances, yet struggle to translate their physical experience into categorical semantic abstractions such as “somewhat difficult” or “moderately difficult.” By providing clear animated visual models, the AAQ bypasses subjective verbal mapping. The respondent compares internal perceptual-motor representations directly with concrete visual vignettes, reducing cognitive translation error and educational disparity.

Second, the AAQ is grounded in Visual Vignette Theory and Observational Anchoring. Derived from behavioral vignette methodologies in social science and psycholinguistics, visual anchors provide standardized behavioral benchmarks. Using a 3D avatar eliminates confounding peripheral cues such as age, gender, race, facial expressions of pain, or body mass, allowing the respondent to focus entirely on kinematic markers: movement velocity, limb stiffness, compensatory leaning, hand support, and joint range of motion.

Third, the AAQ is rooted in Item Response Theory (IRT), specifically the Graded Response Model (GRM) for polytomous data. Rather than relying on Classical Test Theory (CTT), where ordinal integer weights are summed under the assumption of equal interval distances, IRT posits that an individual’s response to an animated item is a mathematical function of their latent functional ability (θ) and the item’s specific psychometric parameters (discrimination parameter a and threshold difficulty parameters b_k). This framework enables precise estimation of functional capacity across the full ability continuum, provides conditional standard errors of measurement, and establishes interval-level measurement properties for clinical monitoring.

Validity

The Animated Activity Questionnaire has undergone rigorous psychometric evaluation assessing content, construct, convergent, discriminant, and cross-cultural validity.

Content and Face Validity

During initial development, activities were selected through focus groups and Delphi consensus panels involving orthopedic surgeons, rheumatologists, physical therapists, and individuals with hip or knee osteoarthritis. Avatar animations were refined through iterative pilot testing, ensuring that compensatory movements and functional limitations were biomechanically realistic and clearly distinguishable.

Construct and Convergent Validity

Construct validity has been established by evaluating hypothesized relationships between AAQ scores and established legacy measures:

• Legacy PROMs: The AAQ correlates strongly with established text-based physical functioning subscales. Studies show Pearson/Spearman correlations ranging from r = 0.73 to 0.84 with the HOOS and KOOS physical functioning subscales, and r = 0.70 to 0.81 with the WOMAC physical function subscale. Strong correlations are also observed with the Physical Component Summary (PCS) of the Medical Outcomes Study Short-Form Health Survey (SF-36; r = 0.65 to 0.75).
• Objective Performance Measures: A recognized strength of the AAQ is its correlation with objective physical performance tests compared to traditional questionnaires. Research demonstrates moderate-to-strong correlations between the AAQ and the Timed Up and Go (TUG) test (r = -0.62 to -0.71), the 40-meter Fast-Paced Walk Test (r = 0.58 to 0.69), the 30-second Chair Stand Test (r = 0.61 to 0.70), and the Stair Climb Test (r = -0.64 to -0.74). In comparative head-to-head analyses, AAQ scores shared significantly more variance with objective performance metrics than conventional WOMAC or SF-36 functional subscales, supporting its ability to capture tangible kinematic performance.

Discriminant and Known-Groups Validity

The instrument demonstrates strong discriminant validity, showing weaker correlations with constructs such as emotional well-being (SF-36 Mental Health, r < 0.35) and systemic comorbidities. In known-groups evaluations, the AAQ reliably distinguishes between patients with varying radiographic Kellgren-Lawrence grades, patients managed conservatively versus those scheduled for total joint arthroplasty, and individuals requiring walking aids versus unassisted ambulators (p < 0.001).

Cross-Cultural and Cross-Language Equivalence

Because the AAQ utilizes visual animations rather than complex textual descriptions, it demonstrates minimal Differential Item Functioning (DIF) across distinct linguistic and cultural populations. International validation studies across Dutch, French, English, and Scandinavian cohorts confirm that the animated prompts retain uniform measurement properties across diverse populations without requiring textual re-translation.

Reliability

The Animated Activity Questionnaire demonstrates high reliability across both Classical Test Theory and Item Response Theory frameworks.

Internal Consistency

The AAQ exhibits high internal consistency across diverse clinical cohorts. In patients with knee or hip osteoarthritis, overall Cronbach’s alpha coefficients consistently fall between α = 0.93 and 0.96. From an IRT perspective, the Person Separation Index (PSI) reliably exceeds 0.90, confirming that the 17 items provide fine-grained stratification of individuals along the physical functioning spectrum.

Test-Retest Reliability and Reproducibility

Temporal stability has been evaluated among stable outpatient cohorts reassessed over intervals ranging from 7 to 14 days. Intraclass Correlation Coefficients (ICC, two-way random effects, absolute agreement) are consistently high:

• Overall AAQ Summary Score: ICC = 0.91 to 0.95 (95% CI [0.88, 0.97]).
• Sub-domain reliability: Stair-related items (ICC = 0.88 to 0.93), seated transfer items (ICC = 0.86 to 0.91), and ambulatory walking items (ICC = 0.85 to 0.92).

Measurement Error and Responsiveness Parameters

Psychometric investigations have established key measurement precision indices:

• Standard Error of Measurement (SEM): Typically estimated between 3.8 and 5.2 points on the 0–100 standardized metric.
• Smallest Detectable Change (SDC): Calculated at both individual (SDC_ind ≈ 10.5 to 14.4 points) and group levels (SDC_group ≈ 1.5 to 2.2 points) at a 95% confidence threshold.
• Minimal Clinically Important Difference (MCID): In longitudinal studies examining patients undergoing total hip or knee arthroplasty, anchor-based MCID estimates average 8.0 to 12.0 points, indicating good responsiveness to clinically meaningful recovery.

  Factor Analysis

  The structural dimensionality of the AAQ has been examined through Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA), and nonparametric/parametric Item Response Theory frameworks.

  Unidimensionality and Factor Structure

  Analyses confirm that the AAQ is essentially unidimensional. Exploratory factor analyses using robust weighted least squares extraction consistently identify a dominant primary factor accounting for more than 58% to 66% of the common variance, with an eigenvalue ratio between the first and second factors exceeding 5:1. Subsequent CFAs applying polychoric correlation matrices demonstrate acceptable model fit indices across representative osteoarthritis populations:

  • Comparative Fit Index (CFI): ≥ 0.96 to 0.98 (exceeding the standard 0.95 threshold for good model fit).
  • Tucker-Lewis Index (TLI): ≥ 0.95 to 0.97.
  • Root Mean Square Error of Approximation (RMSEA): 0.052 to 0.068 (95% CI [0.044, 0.076]), satisfying conventional criteria for acceptable approximate fit.

  Item Response Theory (IRT) Calibration

  Given the polytomous nature of the response categories (3 to 5 graded animated vignettes per item), Samejima’s Graded Response Model (GRM) was applied. Key IRT findings include:

  • Discrimination Parameters (a): Item discrimination parameters are uniformly high, ranging from a = 1.35 to a = 3.20. Items involving descending stairs (Item 4), ascending stairs (Item 5), and rising from a low chair (Item 11) demonstrate the highest discriminative power, functioning as sensitive markers along the functional limitation continuum.
  • Threshold Parameters (b_k): Threshold locations span a broad range across the latent trait spectrum (θ ≈ -3.0 to +2.5 standard deviations). This distribution minimizes floor and ceiling effects relative to traditional instruments, maintaining measurement precision across diverse clinical stages.
  • Residual Correlations and Local Independence: Residual correlation matrices show no significant local item dependence (Yen’s Q₃ statistics < 0.20 across all pairs), confirming that each item contributes unique psychometric information to the latent score.

  Instrument / Measurement Tool

  • Full Instrument Name: Animated Activity Questionnaire (AAQ)
  • Primary Author: Wilfred F. Peter, PT, PhD, et al. (2015)
  • Construct Assessed: Self-reported limitations in physical functioning and basic activities of daily living (ADL)
  • Target Population: Adults and older adults diagnosed with lower extremity musculoskeletal disorders, specifically hip and knee osteoarthritis
  • Administration Format: Computer-administered digital assessment (accessible via web browsers, tablets, and interactive clinical kiosks)
  • Item Count: 17 distinct functional activity items
  • Administration Time: Approximately 8 to 12 minutes
  • Response Format: Video-based selection. For each activity, 3 to 5 animated video clips are shown illustrating performance ranging from normal/unrestricted to severely limited (along with a response option ‘I cannot perform this activity’ or ‘I perform this differently’). Each chosen animation corresponds to an item score (transformed via an IRT/calibration algorithm to a summary scale of 0 to 100).
  • Scoring and Transformation Rules:
    • Each animated option represents an empirically calibrated ordinal functional level.
    • Item responses are synthesized and converted through an IRT-based scoring algorithm.
    • Scores across the 17 activities are converted to a standardized metric from 0 to 100, where 0 represents maximum limitation in performing basic daily activities and 100 represents completely unrestricted performance (no limitations).
    • If automated IRT scoring software is unavailable, sum-score conversion tables calibrated in the original psychometric validation cohorts can be used to convert raw scores into standardized 0–100 values.

  Permissions & Fee and Test Year

  The Animated Activity Questionnaire was formally published and psychometrically established in 2015 under the clinical direction of Dr. Wilfred F. Peter and colleagues at the Leiden University Medical Center (LUMC) and the Reade Center for Rehabilitation and Rheumatology in Amsterdam, Netherlands. The instrument, including its underlying 3D avatar animations and proprietary digital algorithms, is protected by international copyright law.

  The AAQ is accessible for non-commercial academic research, public hospital clinical trials, and non-funded clinical practice, provided formal attribution is maintained. Access to video assets, software integration modules, and the digital assessment platform can be requested through the official administration portal at http://www.myaaq.com or by contacting the technology transfer offices of Leiden University Medical Center and Reade. Commercial use, integration into proprietary electronic health record (EHR) software, or deployment in industry-sponsored clinical trials requires a formal licensing agreement and appropriate software access fees.

  References

  • Peter, W. F., Looman, C. W., Roorda, L. D., Gerritsen, V. M., van der Leeden, M., Steultjens, M. P., Terwee, C. B., & Vliet Vlieland, T. P. (2015). The Animated Activity Questionnaire: Development and validation of a new tool to assess physical functioning in patients with hip or knee osteoarthritis. Arthritis Care & Research, 67(11), 1548–1558. https://doi.org/10.1002/acr.22619
  • Peter, W. F., Terwee, C. B., Roorda, L. D., van der Esch, M., Steultjens, M. P., & Vliet Vlieland, T. P. (2016). Construct validity of the Animated Activity Questionnaire in patients with hip or knee osteoarthritis. Osteoarthritis and Cartilage, 24(9), 1561–1569. https://doi.org/10.1016/j.joca.2016.04.015
  • Peter, W. F., de Vet, H. C., Roorda, L. D., van der Esch, M., Steultjens, M. P., Terwee, C. B., & Vliet Vlieland, T. P. (2017). Measurement error and responsiveness of the Animated Activity Questionnaire in patients with hip or knee osteoarthritis. Rheumatology, 56(11), 1938–1946. https://doi.org/10.1093/rheumatology/kex272
  • Tourangeau, R., Rips, L. J., & Rasinski, K. (2000). The Psychology of Survey Response. Cambridge University Press. https://doi.org/10.1017/CBO9780511819322
  • Samejima, F. (1969). Estimation of latent ability using a response pattern of graded scores. Psychometrika Monograph Supplement, 34(4, Pt. 2), 1–100. https://doi.org/10.1007/BF03372160
  • World Health Organization. (2001). International Classification of Functioning, Disability and Health: ICF. World Health Organization. https://apps.who.int/iris/handle/10665/42407

  Items of the Scale

  Below are the authentic scale items in their original language as published in the standard psychometric validation studies, without modification or translation to preserve instrument validity and reliability:

  Response Format:
  Video-based selection: For each activity, 3 to 5 animated video clips are shown illustrating performance ranging from normal/unrestricted to severely limited (along with a response option ‘I cannot perform this activity’ or ‘I perform this differently’). Each chosen animation corresponds to an item score (transformed via an IRT/calibration algorithm to a summary scale of 0 to 100).

  Scoring Rule:
  Scores across the 17 activities are converted to a standardized metric from 0 to 100, where 0 represents maximum limitation in performing basic daily activities and 100 represents completely unrestricted performance (no limitations).

  1. Walking across an uneven surface
  2. Walking outdoors
  3. Walking indoors
  4. Walking downstairs
  5. Walking upstairs
  6. Sitting down on a sofa
  7. Rising from a sofa
  8. Sitting down on a chair
  9. Rising from a chair
  10. Sitting down on a low chair
  11. Rising from a low chair
  12. Sitting down on a toilet
  13. Rising from a toilet
  14. Getting into bed
  15. Getting out of bed
  16. Getting into a car
  17. Getting out of a car
  ★

  Rate This Scale

  5.0 / 5 • 1 vote