Flow in screen-based gamified virtual reality sports: development and validation of a behaviorally anchored proxy...

Adolescents with autism spectrum disorder (ASD) face substantial barriers to physical activity participation because of motor coordination difficulties, sensory hypersensitivity, and motivational challenges. Screen-based gamified virtual reality (VR) sports have emerged as a promising alternative, with flow identified as a key psychological mechanism underlying their effects. However, alexithymia and atypical interoceptive processing in adolescents with ASD render conventional self-report flow scales inappropriate for this population. This study aimed to develop and validate a behaviorally anchored rating scale (BARS)-format proxy assessment tool, VR-Flow-BARS, that lets observers evaluate the flow-related behaviors of adolescents with ASD during VR sports participation.

Eighteen preliminary items were developed from flow theory and the theoretical framework underlying the Flow State Scale-2. An eight-member expert panel of specialists in adapted physical education, ASD education, and VR sports evaluated the items against three criteria: content validity, observability, and appropriateness for ASD. The items were refined through cognitive interviews (n = 8) and pilot testing (n = 30). Exploratory factor analysis (EFA) using principal axis factoring with direct oblimin rotation was conducted on Sample 1 (n = 100), followed by confirmatory factor analysis (CFA) on Sample 2 (n = 200). Reliability was assessed using Cronbach's α, composite reliability (CR), and the intraclass correlation coefficient (ICC) (2,1). Validity was examined using the average variance extracted (AVE), the Fornell-Larcker criterion, and the heterotrait-monotrait (HTMT) ratio.

The expert review yielded an item-level content validity index of 1.00 and a scale-level content validity index of 1.00 across all three criteria for the 15 retained items; the factor "sensorimotor flow expression" (with three items) was deleted for conceptual overlap and limited observability. EFA identified a five-factor structure accounting for 68.394% of the total variance. CFA demonstrated acceptable model fit (χ ²/df = 2.000, comparative fit index = 0.969, Tucker-Lewis index = 0.959, root mean square error of approximation = 0.071), with standardized factor loadings ranging from 0.763 to 0.940. Internal consistency was high (Cronbach's α = 0.864-0.946; CR = 0.868-0.946). Inter-rater reliability (ICC) ranged from 0.729 to 0.910 (indicating moderate to excellent reliability). AVE values ranged from 0.687 to 0.854, satisfying the convergent validity criterion of 0.50, and HTMT values for all factor pairs were below 0.90 at the point-estimate level; however, the 95% bootstrap confidence interval for the FR-CEP pair [0.851, 0.939] extended beyond 0.90, indicating borderline discriminant validity for this factor pair.

To our knowledge, VR-Flow-BARS tool is the first BARS-format proxy assessment instrument developed to evaluate flow-related behaviors in adolescents with ASD during screen-based gamified VR sports. It demonstrates an acceptable factor structure, high internal consistency, adequate inter-rater reliability, and internal structural validity. This scale may serve as a practical tool for the real-time monitoring of immersive behaviors and individualized intervention design in VR-based physical activity programs for adolescents with ASD who have mild-to-moderate educational support needs and are able to engage in VR tasks with routine prompting. Future research should address external criterion-based validation, test-retest reliability, and applicability across a broader range of ASD severity levels.