Classroom study shows beginner learners engage with ChatGPT feedback, indicating the need for tailored teaching strategies.
Key Points
This study aims to understand how beginner Japanese learners engage with ChatGPT feedback during writing tasks. It focuses on emotional, cognitive, and behavioral responses to the AI's suggestions.
Examined four beginner learners of Japanese
Collected data through interviews, screen recordings, and reflective journals
Analyzed error types, prompt types, and accuracy of ChatGPT output
Participants valued English translations but experienced negative emotions from inconsistent feedback
Orthographic accessibility limited attention to ChatGPT suggestions, yet learners critically evaluated them
Learners frequently verified ChatGPT output with external resources and mainly used it for proofreading