ChatGPT integration in Chinese language pedagogy: a Vietnamese multi-institutional study

Abstract

This study investigates ChatGPT integration effectiveness in Chinese language education across three Vietnamese universities involving 350 students and 45 instructors. Using mixed-methods analysis, quantitative surveys revealed 78.3% improved learning motivation with artificial intelligence (AI)-enhanced personalized learning, while 82.1% valued immediate feedback capabilities. Significant disciplinary variations emerged: traditional medicine students demonstrated highest acceptance rates (M=4.3, SD=0.7) compared to translation studies students (M=3.7, SD=1.0), F(2, 347)=15.8, p<0.001. AI literacy was strongly associated with learning effectiveness (r=0.67, p<0.001), accounting for 45% of outcome variance. Qualitative analysis identified human-AI interactive negotiation competence (HAINC) development as critical, with 91% of instructors emphasizing systematic training needs. Despite benefits, 64.7% expressed human interaction concerns, revealing fundamental pedagogical tensions requiring collaborative human-AI frameworks. Findings provide evidence-based guidance for implementing discipline-responsive AI systems in Vietnamese higher education while maintaining pedagogical authenticity and addressing technological literacy gaps.