Speaker
Description
While the use of AI tools in English language classrooms has gained significant attention, how young learners emotionally respond to this technology remains under-explored, particularly in non-metropolitan settings. To address this gap, fieldwork was conducted at a language center in Cao Lanh City, Dong Thap Province, examining 50 primary school students (aged 7–11) who interacted with AI driven speech recognition applications for pronunciation tasks. Methodologically, we utilized a visual emoji-based questionnaire alongside brief semi-structured interviews to capture their immediate psychological states during these digital sessions. The field data reveal a clear, sharp split in student psychology. Although the children expressed high enthusiasm when interacting with virtual characters, automated error corrections and instant algorithmic scoring frequently triggered immediate language anxiety. This friction directly reflects the realistic mix of uneven digital habits and socioeconomic backgrounds typical of a provincial student cohort. With only small sample size and the initial novelty of the technology, these early findings should not be widely applied to all contexts. Even so, this research points to a clear, practical direction for future AI learning tools. First, software developers need to move away from strict "right or wrong" scoring. Instead, they should build gamified feedback systems that let children try again multiple times without losing points right away. Second, classroom teachers in provincial towns should prepare students mentally before they use the apps. By helping children know what to expect from the technology, teachers can change automated corrections from an anxiety trigger into a helpful learning tool that keeps them motivated over time.
Keywords: primary learners, affective responses, AI-powered tools, language anxiety, case study.
