Evaluating Creativity in Generative AI Outputs: Dimensions and Cross-National Differences Among Art and Design Students

Résumé fourni par la source

Generative Artificial Intelligence (GenAI) has revolutionized the creative industries, prompting its widespread adoption in higher education. However, significant gaps remain regarding students’ perceptions of AI-generated visual outputs across different cultures. Employing a multi-stage methodological approach, this study first utilized focus groups to identify the dimensions underlying the perceived creativity of AI-generated visual outputs. Subsequently, a survey of 1,088 students from China and the U.S. was analyzed using structural equation modeling (SEM), multi-group analysis (MGA), and artificial neural networks (ANN). Results indicate that novelty, resolution, and affect are significant evaluation dimensions in both national samples. However, distinct cross-national differences emerged: aesthetics and inspiration were significant for the U.S. students, whereas diversity and transferability were pivotal for Chinese students. These findings provide an empirically supported framework for educators and researchers to inform GenAI integration in art and design education and foster deeper human–AI creative collaboration.