Application of Generative Models for Personalized Design Education: Adaptive Approaches in Art-pedagogical Learning
TL;DR
The findings show that, when supported by a structured pedagogical model, generative AI can serve as an effective educational agent that fosters learner independence in creative decision-making and provides a structured pedagogical pathway for personalized design instruction.
Abstract
This article introduces Generative Design Adaptive Learning Environment (G-DALE), an adaptive generative learning framework designed to personalize creative development in art and design education. The framework combines learner profiling, adaptive sequencing of generative tasks, multimodal feedback, and real-time instructional support to align learning activities with individual creative needs. We conducted a four-week study with 72 participants to evaluate students’ engagement with generative Artificial Intelligence (AI) tools and the influence of adaptive mechanisms on their performance. The results demonstrate notable gains in creative quality, including deeper conceptual thinking, improved visual coherence, and greater stylistic stability (paired-samples t-tests, all p < 0.001, Cohen’s d = 0.65–0.77). Students also showed higher productivity through faster iteration cycles and increased design variation. Cluster analysis revealed distinct learner adaptation profiles, indicating that personalization emerges through differentiated patterns of cognitive scaffolding, refinement guidance, and style alignment. Overall, the findings show that, when supported by a structured pedagogical model, generative AI can serve as an effective educational agent that fosters learner independence in creative decision-making and provides a structured pedagogical pathway for personalized design instruction.