Exploring The Relationship Between Student Perceptions, Academic Performance, And Learning Behaviors in AI Assisted Art Education

Abstract

Generative Artificial Intelligence (GenAI) is increasingly influencing higher education, particularly in creative disciplines where concerns regarding creativity and skill development remain significant. This study employs a dual-dataset framework that combines a visual arts student survey [19] with a large-scale academic behavioral dataset [20] to investigate the relationships among AI perceptions, academic performance, and learning behaviors. The results reveal a cognitive paradox: while students generally recognize the educational benefits of GenAI, many remain concerned about creativity loss and career displacement. Regression analysis indicates that academic performance and AI literacy are significant predictors of positive AI perceptions, suggesting that higher-performing students are more likely to use AI as a learning support tool rather than a substitute for independent work. Furthermore, behavioral analyses reveal a non-linear relationship between AI usage and academic outcomes, where moderate use is associated with improved performance, while excessive reliance is linked to increased anxiety, reduced skill retention, and weaker learning outcomes. Based on these findings, the study recommends a scaffolded approach to GenAI integration that supports technical learning tasks while preserving creativity, critical thinking, and artistic skill development.