Unpacking Co-Creation with AI: Understanding Students’ Creativity in a Game Design Course

Abstract

The proliferation of Artificial Intelligence (AI) is fundamentally reshaping creative work and education. While research has demonstrated AI's potential to enhance creative outcomes, the processes underlying human-AI co-creation remain insufficiently explored. This study investigates the dynamics of human-AI co-creation in an undergraduate game design course by leveraging Learning Analytics and Epistemic Network Analysis. Specifically, it examines how students’ creative thinking types (divergent, convergent, and evaluative thinking) interact with their contribution types (creating new ideas, extending ideas, refining ideas, and transforming ideas) during co-creation with AI. Results reveal the collaborative role of different thinking types in the student-AI co-creative process. Divergent thinking primarily drives the generation of new ideas, while convergent and evaluative thinking are essential for refinement and quality control. These cognitive-behavioral relationships evolve across learning phases, shifting from open exploration to focused integration. Furthermore, high-achieving students strategically employ convergent thinking and engage in continuous refinement and transformation with AI, whereas lower-achieving students tend to remain at the exploratory stage and rely on initial AI outputs. Practically, the findings underscore the importance of supporting students in shifting between creative thinking modes, promoting critical evaluation, and offering phase-specific guidance to facilitate effective and meaningful co-creation with AI in educational settings.

Ähnliche Arbeiten