The Effects of Prosody Training with AI Chatbot on the English Pronunciation Improvement of Korean EFL Learners
Sohyun Yoo, Hyunkee Ahn · Korean Journal of English Language and Linguistics · 2024
This study investigated the impact of prosody training using pitch contour feedback on the English pronunciation of Korean elementary school students. The aim was to evaluate whether this training could lead to significant improvements in pronunciation. The study involved 18 fifth-grade students who participated in a twoweek intensive program using AI PengTalk, a chatbot developed to support Korean elementary school students’ English learning on a national scale. The participants practiced aligning their pitch contours with those of native English speakers using the tool. The effectiveness of the training was evaluated through AI PengTalk’s automated pronunciation scoring and assessments by native English speakers, who rated pronunciation based on nativeness and intelligibility, using a 5-point Likert scale and a ratio scale, respectively. Additionally, the study explored the transferability of the training to unpracticed sentences. Although the automated scoring did not reveal statistically significant differences, assessments by native English speakers indicated notable enhancements in pronunciation regarding both nativeness and intelligibility. Furthermore, the training effect was generalized to novel sentences that were not explicitly practiced during the sessions. These findings suggest that prosody training with pitch contour feedback is a promising approach for improving English pronunciation among young Korean EFL learners.