본 연구는 Deepseek와 ChatGPT가 중국 대학 한국어 학습자들의 쓰 기 텍스트에 대해 제공하는 피드백 양상을 살피고, AI 피드백의 활용 방 안을 모색하는 데 목적이 있다. 이를 위해 한국어 학습자 30명을 대상으 로 쓰기 텍스트를 수집하고, 두 도구가 텍스트에 대해 제공한 피드백을 분석하였다. 분석 결과, 두 도구 모두 평가 기준에 기반하여 피드백을 제 공하고, 학습자가 수용하기 쉽도록 문제점과 개선 방향을 명확하게 제시 하였다. 대부분의 피드백은 정확도가 높은 편이었으나, 평가 범주와 오류 유형이 부합하지 않는 경우도 일부 나타나고 있었고, 특히 Deepseek의 경우 쓰기 주제를 잘못 파악하여 내용적 측면에서 부적절한 피드백을 제 공한 사례가 많았다. 두 도구 모두 쓰기 텍스트의 내용에 따라 피드백을 제공하여 우선순위는 두드러지지 않았고, Deepseek가 ChatGPT에 비해 지지적인 어조를 더 많이 사용하는 것으로 나타났다. 본 연구는 생성형 AI를 활용한 한국어 교육 및 연구를 위해 기초적인 자료로 사용할 수 있 을 것이다.

This study investigated the effect of automated writing evaluation (AWE) feedback on Korean university students’ revision behavior. Specifically, it explored the extent to which the quality of essays differed between first drafts and second drafts. Furthermore, the way individual students dealt with incorrect feedback in revising essays was examined. In this study, ETS’ Criterion program, one of the most popular AWE programs, was employed. In cases where incorrect feedback was detected, the first and second drafts were closely compared to find out how students altered the text in response to the incorrect feedback, and this was classified into one of three categories: successful changes, unsuccessful changes, and text deletion. The results showed that students’ second drafts were rated significantly higher than first drafts by two NS raters, which indicates that many students followed the revision suggestions made by the Criterion program. As students’ Criterion scores increased, their ability to identify and make successful changes in response to incorrect feedback improved. The findings of the study can advance our understanding of AWE use in an EFL context and should contribute to broader examination of how Korean university students engage in revision of their essays.

This paper examined effects of three types of written corrective feedback (CF) on L2 learners’ grammatical accuracy in writing, focusing on the use of verb tense and articles. Within Ellis’ (2008) typology of CF, the present study focused on direct, indirect, and metalinguistic CF. The feedback was provided at each stage of writing, and the participants were to write three new pieces of writing. Changes in the accuracy of the grammatical targets were measured. Results indicated that despite the limitations of the small number of the participants, positive changes in the grammatical accuracy were found with metalinguistic CF for verb tense and with indirect CF for articles in new pieces of writing. In addition, unlike previous studies, a negative role of direct written feedback was observed in the present study.

This study investigated which type of corrective feedback, direct feedback or indirect feedback, would be more effective for reducing the targeted errors−verb, noun, and determiner errors−in L2 writing. The 20 Korean EFL college students were divided into two feedback conditions: direct feedback and indirect coded feedback. The results of the study showed that only the errors in the noun category, not the other two categories, showed a significant decrease over time. There were no significant differences in the effect of the differential feedback. The post surveys and interviews revealed that the students in both groups had difficulty in understanding their errors and feedback on them, and they wanted to receive a more detailed explanation about their errors in order to fully understand them. The overall findings indicate that different types of corrective feedback need to be combined with other methods such as a one-to-one conference or mini-lesson for further helping students understand their errors and reduce other errors in the future. This study provides L2 teachers with valuable information on how to respond more effectively to the needs of L2 learners.

This study investigates the effects of indirect and direct error feedback on Korean college students’ accuracy improvement in writing and their responses to the feedback they received. The 32 participants of the study were divided into four groups of different error feedback conditions: indirect feedback, indirect feedback along with metalinguistic feedback, direct feedback, and direct feedback along with metalinguistic feedback. Direct feedback was full, explicit error correction, while indirect feedback was coded error correction in which errors were underlined and marked with code. Metalinguistic feedback was marginal explanations about errors. Analysis of error rates in the students’ first and last essay writing did not show statistically significant difference in the accuracy improvement of the four groups. In other words, indirect error feedback was not more effective than direct feedback. However, the gain score in each group showed that indirect feedback and direct feedback were more effective when they were combined with metalinguistic feedback. In particular, the students’ learning diary clearly indicated that the students valued error feedback on their writing and that follow-up feedback was crucial in providing indirect error feedback.

대화일지 쓰기는 학습자와 교사 간에 자연스럽게 의사소통을 할 수 있는 장이다. 대화일지를 통해 학습자는 부담감 없고 자유롭게 자기 생각을 표현하고 교사는 학습자의 글을 피드백해 줄 수 있다. 본고에서는 대화일지 쓰기를 통해 교사의 피드백 유형이 학습자의 쓰기 능력에 어느 정도 효과가 있는지 살펴보고 대화일지가 학습자의 쓰기 학습에 대한 자신감, 불안감, 부담감에 어떤 영향을 주었는지를 실험을 통해 밝혀 보고자 하였다. 실험은 8주 동안 중급과정에서 한국어를 배운 15명을 대상으로 하였다. 그 결과 대화일지 쓰기를 통한 교사의 피드백이 학습자의 쓰기 능력과 불안감 감소에 어느 정도 효과가 있었음을 알 수 있었다. 이 중 쓰기 능력에서는 대화일지를 통한 형태 중심․의미 중심의 피드백이 기존의 직접적인 피드백보다 담화 구성 능력 향상에 긍정적인 영향을 주었음을 알 수 있었다.