Designing an AI-VR Storytelling Animation for Early Reading Stimulation in Early Childhood Education: A Needs Analysis and Design Validation Study

Abstract

Background: Early reading ability is a fundamental literacy competency for children aged 5–6 years, yet early childhood education (ECE) settings in Indonesia remain predominantly reliant on conventional, non-adaptive instructional media. Objective: This study aimed to (1) identify the learning needs and technology readiness of PAUD teachers regarding early reading media, and (2) design a conceptual AI-VR storytelling animation media at Technology Readiness Level (TRL) 1–3. Methods: A Research and Development (R&D) approach using the ADDIE model was applied, bounded at TRL 1–3 (Analysis and Design phases). An online, self-administered needs-analysis questionnaire (40 Likert items across five sub-scales, plus open-ended items) and a storyboard response questionnaire (7 Likert items plus open-ended items) were distributed to PAUD teachers via a shared digital form; 167 teachers completed both instruments (97.6% female; 74.3% from Central Sulawesi, 22.2% from South Sulawesi, 3.6% unclassifiable location). A five-episode AI-VR storyboard was rated by three expert validators (two media experts, one ECE curriculum expert) against an 8-item rubric (5-point scale) covering four dimensions; content validity was quantified using Aiken's V (Aiken, 1985). Data were analysed descriptively (mean, SD), for internal consistency (Cronbach's alpha), and for between-region differences (Mann–Whitney U); open-ended responses were examined through keyword-based content analysis. Results: Internal consistency was acceptable to excellent across sub-scales (α = 0.74–0.96). Teachers reported strong motivational/innovation needs (M = 4.39) and clear prioritisation of early-reading indicators (M = 4.44), moderate current-practice limitations (M = 3.78) and digital (AI/VR) readiness (M = 3.82). No statistically significant differences were found between Central and South Sulawesi respondents on any sub-scale (all p > 0.05, Mann–Whitney U). Expert validation of the storyboard yielded a mean Aiken's V of 0.865 (range 0.750–1.000), at the boundary of the acceptable (V ≥ 0.75) and high (V ≥ 0.87) validity thresholds. Teacher response to the storyboard was high (overall M = 4.22, SD = 0.63, α = 0.95). Conclusion: The AI-VR storytelling animation design is grounded in an empirically documented practitioner needs profile, meets acceptable-to-high expert content validity, and received a positive, internally consistent practitioner response, providing a design-stage foundation for prototype development at TRL 4 and beyond. The self-selected, geographically and educationally skewed nature of the online sample, and the small (n = 3) validator panel, are addressed as limitations.