Earticle

This study attempts to investigate how AI-based technology, AI-recognition/recording application (Clova Note), affects language learners' use of metacognitive strategies when acquiring speaking skills using a metacognitive scaffolding tool. The study employed a mixed research method to obtain an integrative understanding of the learning effect. The garnered data were analyzed using paired sample t-test to compare pre-test and post-test and one-way ANOVA to assess the effect of using metacognitive strategies for speaking. The results from the study can be summarized as follows. First, language learners perceived that using AI-based Apps activated their awareness of accuracy and fluency in language learning. Second, this AI-based metacognitive scaffolding tool promoted problem-solving of metacognitive strategy, indicating that learners could manage their learning difficulties in speaking activities. Third, among the five sub-scales of speaking competence, comprehension and pronunciation were found to be the most effective. Upon these findings, noticeable pedagogical implications were discussed.

This action research investigated the impact of Clova Note, Naver’s speech recognition app, on supporting speaking practices among Korean university students in general English courses. Over one semester, the study involved 60 beginner and intermediate students. Students in this study completed read-aloud tasks using Clova Note. Data sources?surveys, journals, interviews, and peer instructor evaluations?were analyzed through a mixed-methods approach. Findings showed that repeated practice with Clova Note promoted autonomy, improved pronunciation awareness, and boosted speaking confidence, especially among beginners. Qualitative data showed students developed strategies, such as reducing reliance on translation tools and self-correcting based on transcript mismatches. Peer instructor evaluations showed significant fluency gains in intermediate students, while surveys indicated increased motivation and confidence in beginners. These results suggest speech recognition can effectively support independent speaking practice in EFL contexts, especially for less proficient learners in large classes.

본 연구는 한국어 학습자를 위한 자가 피드백의 도구로써 음성인식 텍스트 변환 프로그램의 활용 가능성을 알아보는 데에 목적이 있다. 이를 위해 본 연구에서는 클로바 노트(CLOVA Note)를 사용하여 초급 단계의 외국인 학부생 대상 말하기 수업에서 자가 피드백 수업을 진행하고 학습자의 자가 피드백 수행 양상 및 학습자 인식에 관한 설문조사 결과를 분석하였다. 본 연구에서 학습자들의 자가 피드백 수행 양상을 살펴본 결과, 우선 학습자들은 문법 범주에서 조사 오류를 발견하고 수정하는 것을 어려워했다. 어휘 범주에서는 전반적으로 오류 횟수가 많지 않았고 자가 피드백도 잘 수행한 편이었으나 그중 높임 어휘와 관련한 부분에서는 오류 횟수가 많았고 피드백 수행도 성공적이지 못했다. 따라서 조사와 같은 문법 범주는 단기간에 교육적 효과를 기대하기 어렵다는 점, 자가 피드백을 수행하기에 적절한 숙달도 단계는 초급 후반부 또는 종료 시점부터라는 점을 발견할 수 있었다. 설문조사 결과를 통해서는 학습자들이 자가 피드백 수업에 대체로 높은 만족도를 보였고 말하기 효능감을 느꼈다는 점, 자가 피드백 방식이 최상위권이나 하위권 학습자보다는 중위권에서 상위권에 해당하는 학습자들에 더 높은 교육적 효과를 기대할 수 있는 방식이라는 점 등을 확인할 수 있었다.

This study aims to verify the applicability of using a speech-to-text conversion program, specifically CLOVA Note, as a tool for self-feedback for Korean language education for beginners. The research focuses on conducting self-feedback sessions during speaking classes using the CLOVA Note with beginner level foreign undergraduate students. Data on patterns of self-feedback and perceptions of its effectiveness are collected and analyzed. Findings reveal that learners face difficulties in identifying and correcting grammatical errors, notably in the usage of particles. While the error rate in the vocabulary category is generally low and learners appear adept at self-feedback, significant errors and unsuccessful attempts at self-feedback are found in the area of honorific vocabulary. These results suggest that rapid pedagogical improvements in areas such as particle usage may be challenging to achieve and the optimal period for the introduction of self-feedback mechanisms appears to be during the later stages of the beginner level or upon course completion. Survey results indicate a high level of satisfaction and an improvement in perceived speaking efficacy among learners following the introduction of self-feedback sessions. Further observations reveal that the self-feedback methodology yields higher educational gains for students in the mid-to-upper proficiency range, rather than those at the extreme ends of the spectrum.

본 연구는 네이버 클로바노트를 발음 교육에 활용한 사례를 소개하였다. 한국어과 학생 38명을 대상으로 자가 발음 평가와 교사의 피드백을 수행하고 그 결과를 분석하였다. 녹음과제를 통해 학생들은 자신이 자주 범하는 오류를 파악했다. 학생이 자기의 오류를 확인하고 고치면서 발음의 정확도가 높아졌다. 본 연구는 발음교육 보조도구로 음성인식이 프로그램이 효과가 있음을 검증하였다.

The purpose of this study is to verify the effectiveness of semester-long pronunciation education using CLOVA Note voice recognition program. 38 students in Korean class were asked to describe what they experienced from the recording assignments using CLOVA Note, teacher feedback, and self-pronunciation assessment. Through this process, students identified the frequent errors they made, analyzed the errors and found the way to improve the accuracy of their pronunciation. The significance of this study is that the self-assessment of learners’ pronunciation using speech recognition technology was effective in pronunciation education.