Evaluating the Role of ChatGPT in Enhancing English Language Learning Outcomes: A Case Study of Gifted Secondary-Level Students

Abstract

This study examines which ChatGPT-related factors most strongly influence students’ language learning in a Vietnamese gifted high school (High School for Gifted Students in Social Sciences and Humanities, VNU). Building on prior studies and expert consultation, we propose four factor groups, including Speed (F1), Data diversity (F2), Convenience (F3), and Performance (F4), with sub-criteria that reflect common language learning uses. Data were collected through expert interviews (including high-achieving students and language-education stakeholders). To synthesize expert judgments and rank the relative importance of factors, the study applies a multi-criteria decision-making approach using the Weighted Sum Model (WSM), including weighting, normalization, and composite-score calculation. Results indicate that Data diversity is the dominant driver of perceived learning impact, achieving the highest composite value (0.90), followed by Speed, Convenience, and Performance. At the sub-criterion level, “Language and vocabulary sources” (F22) exerts the strongest influence (normalized mean 0.95). A second tier of influential criteria includes “Generating answers and responses” (F13), “Data resources” (F23), “Grammar correction” (F33), and “Multitasking, multimodal, and multilingual capabilities” (F42), all rated highly. Lower-rated items include personalization and several advanced capability dimensions. Overall, students value ChatGPT primarily as a rich linguistic input repository and an immediate feedback assistant. The findings suggest schools should integrate AI-supported language activities with clear guidance on verification, privacy, and academic integrity to maximize benefits while limiting risks. Schools should prioritize integrating ChatGPT into language learning as a curated, verifiable source of diverse linguistic input and fast feedback supported by explicit guidance on fact-checking, privacy, and academic integrity.