USABILITY, EFFECTIVENESS AND ENGAGEMENT OF TARI AI TOOLS FOR LANGUAGE LEARNING

This study investigates user perceptions of AI-supported tools developed by the Training and Applied Research Institute (TARI) at Ho Chi Minh City University of Foreign Languages - Information Technology (HUFLIT), designed to enhance language education. As artificial intelligence continues to transform educational practices, the research evaluates the extent to which localized AI tools facilitate usability, effectiveness, and engagement in linguistics learning. The data were collected from 271 participants using a structured questionnaire comprising 18 Likert-scale items and 6 open-ended questions to capture both quantitative and qualitative insights. Data cleaning was conducted using Python, followed by Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to validate the measurement scale. The EFA confirmed a three-factor structure aligned with the theoretical framework, while the CFA demonstrated good model fit (CFI = 0.969; RMSEA = 0.052). Reliability testing indicated strong internal consistency across all constructs of the measurement scale (overall α = 0.944). Composite Reliability (CR) and Average Variance Extracted (AVE) further supported convergent validity for most dimensions. The findings revealed predominantly positive user perceptions. The participants regarded the tools as accessible, interactive, and effective in supporting both conceptual understanding and practical application. The qualitative responses emphasized advantages such as personalization, efficient data processing, and enhanced motivation, while also acknowledging limitations in capturing humanistic dimensions of language learning. The study concludes that TARI AI tools hold significant potential to foster inclusive, engaging, and impactful language education. Continued refinement and user-centered development are recommended to ensure these tools remain responsive to diverse learner needs in dynamic educational contexts.