ASSESSING THE EFFICACY OF AI-INTEGRATED TASKBASED LEARNING IN MEDIA NEWS TRANSLATION: A CASE STUDY AT NGUYEN TAT THANH UNIVERSITY

Abstract

ABSTRACT: This study examines the effects of AI-Integrated Task-Based Learning (AI-TBL), compared with conventional Task-Based Learning (TBL), on students’ independent English-Vietnamese media news translation performance, critical AI evaluation, post-editing skills, and perceptions of AI-supported learning at Nguyen Tat Thanh University. Using a quasi-experimental mixed-methods design, the study compares an experimental group receiving AI-TBL with a control group receiving conventional TBL. AI-TBL involves independent translation, AI consultation, human-AI comparison, critical evaluation, post-editing, and reflection. Data are collected through pre- and post-tests, a six-dimension translation rubric, critical AI-evaluation and post-editing tasks, questionnaires, reflection logs, and semi-structured interviews. The simulated findings indicate that the AI-TBL group achieves greater improvement across the assessed translation competencies, particularly accuracy, contextual appropriateness, journalistic style, and readability, as well as critical AI evaluation and post-editing. Students also report both learning benefits and concerns about inaccurate AI output and over-reliance. The study suggests that AI is most effective when used as a learning scaffold that supports human judgment and independent translation competence. Keywords: artificial intelligence; generative AI; task-based learning; media news translation; translation competence; AI literacy; post-editing.