Research on the Impact Mechanism of Writing Instruction Supported by GenerativeArtificial Intelligence on Students' Learning Willingness

Abstract

With the rapid development of generative AI, its role in writing education has garnered widespread attention. This study constructs and validates a theoretical model to examine how AI-supported writing instruction affects students' learning willingness. Using survey data from Chinese university students analyzed via structural equation modeling, this study investigates the impact of performance expectancy, effort expectancy, perceived randomness, and perceived anthropomorphism on learning willingness. The findings reveal that generative AI significantly enhances learning willingness, with performance and effort expectancies mediating the effect of perceived randomness, while perceived anthropomorphism negatively moderates the relationship between effort expectancy and learning willingness. Based on these findings, the study proposes relevant educational implications and offers insights for the development of human-machine collaborative learning models.