Measuring the Impact of Generative AI on College Chinese Writing: A Structural Equation Modelling Approach

Abstract

This study examines the impact of generative Artificial Intelligence (AI) on college Chinese writing instruction using Partial Least Squares Structural Equation Modelling (PLS-SEM). Drawing on an integrated framework combining the Technology Acceptance Model and Social Cognitive Theory, we investigate relationships between generative AI implementation strategies and learning outcomes through survey data from 376 undergraduate students across eight higher education institutions representing diverse institutional types and academic disciplines in China. Our findings confirm that both generative AI tool integration and instructor facilitation significantly enhance writing proficiency and critical thinking, with these relationships substantially mediated by student engagement with AI tools. The structural model reveals stronger effects on technical writing skills (β = 0.844) compared to critical thinking abilities (β = 0.709), with instructor facilitation demonstrating slightly greater influence than tool availability alone. The measurement model exhibits excellent reliability and validity, with all constructs showing Cronbach’s alpha values exceeding 0.87 and Average Variance Extracted (AVE) values above 0.79. These results provide empirical evidence for the effectiveness of generative AI in Chinese writing education while highlighting the crucial role of human guidance and student engagement in maximizing educational benefits, offering theoretical insights and practical guidance for educators seeking to integrate these emerging technologies into language instruction.