In the AI era: A project-based digital storytelling framework for art and design education

Synthesis: As Generative AI increasingly automates technical production in the creative industries, this study asks how art and design education should evolve to preserve and cultivate the deeply human capacities—emotional resonance, cultural interpretation, and narrative meaning-making—that AI lacks. The authors propose and evaluate a Project-Based Digital Storytelling (PBL-DS) Pedagogies and Teaching Strategies that positions AI as a supportive resource inside a structured creative process rather than an autonomous source of creativity.

The model is grounded in an integrative theoretical framework: Constructivism learning (Vygotsky), Dewey's Experiential Learning, and Amabile's Componential Model of Creativity, operationalized through the four-stage digital storytelling production process (pre-production, production, post-production, distribution). It was implemented and studied as an embedded case study of the 15-week "Creative Shanzhou" capstone project, run by the School of Art and Design at a Chinese university in partnership with the local Cultural and Tourism Bureau, in which 426 final-year undergraduates (guided by 48 mentors, ~1:9 ratio) translated local cultural heritage into Multimodal AI narratives. The research used a case-based, multi-method design combining expert evaluation of creative products, pre/post creative-thinking Assessment, and qualitative project analysis.

Key Findings

• Creative product outcomes were strong overall, strongest in Creativity's originality dimension: expert ratings of the 92 digital storytelling works gave novelty the highest mean (M = 4.21, SD = 0.72), followed by effectiveness (M = 3.96) and wholeness (M = 3.68), indicating students generated original ideas more successfully than they integrated and refined them.
• Wholeness showed the most variability (43% of projects scored below 3.5), pointing to extended Scaffolding needs for integrating narrative, technical execution, and multimodal coherence.
• Creative thinking gains: In the 31-student Animation subgroup, figural TTCT scores rose from 77.23 pre-test to 94.68 post-test—a mean difference of 17.45, t(30) = 4.55, p < .001, Cohen's d = 0.82 (large effect), with good inter-rater reliability (overall ICC = 0.86).
• AI served as a mediated tool, not a creator: Across representative projects (e.g., The Dream of Shanzhou, 3D Journeys of Ancient Shanzhou, Mermaid Legacy), AI supported information organization, visual reference, ideation, and technical editing (e.g., Runway), while students retained ownership of topic selection, narrative interpretation, cultural meaning-making, and final creative decisions.
• Human intentionality drives quality: the top-scoring wholeness example (Mermaid Legacy, 5/5) attributed its integration to sustained human engagement with source material and iterative refinement rather than technical polish from AI tools.
• Real-world validation: the project culminated in a three-day public exhibition of all 92 works, and 31 works were subsequently selected by the Cultural and Tourism Bureau for practical implementation, affirming the model's authentic, community-engaged impact.

What this means for practice

• Educators. Embed Generative AI at every production stage for research, ideation, prototyping, and editing, while requiring students to direct, evaluate, and integrate AI output against their own narrative and emotional goals.
• Educators. Add weekly task statements, progress reviews, and iterative feedback loops aimed at narrative coherence, technical execution, and multimodal synthesis: wholeness was the weakest and most variable rated dimension, with 43% of projects in the lowest score band.
• Curriculum designers. Anchor projects in authentic community partnerships and fieldwork so students work from firsthand cultural material that AI cannot replicate and pursue intrinsic task Motivation.
• Curriculum designers. Staff capstone projects at roughly one mentor per nine students (48 mentors for 426 students here) so guidance on strategic AI use stays consistent and personal.
• Designers. Stage a public exhibition and expert critique as the assessment event so authorship, originality, and creative responsibility become objects of reflection rather than AI literacy slogans.

Limitations

• The creative-thinking gain rests on a one-group pre-test/post-test design with 31 Animation students; the authors state it should not be read as causal evidence for the PBL-DS framework or for AI-supported tools.
• The study covers a single institutional context—a 15-week capstone at one Chinese university with 426 final-year students and 92 projects—so transferability across disciplines and institutions remains untested.
• AI-supported practice was inferred from representative project analysis rather than process data; no design logs, reflective journals, or AI interaction records were collected, so variation in how students actually interacted with AI was not captured.

Citation

Tian, Y., Tang, M., Li, G., & Dang, W. (2026). In the AI era: A project-based digital storytelling framework for art and design education. Computers and Education: Artificial Intelligence, 11, 100645.