김아름 Arum Kim , 서희 Hee Seo

DOI: Vol.27(No.4) 1-18, 2024

This study explores the characteristics of prompt literacy among learners by employing ChatGPT as a supportive tool in Korean language education, specifically analyzing the actual interaction logs between learners and ChatGPT. Over one semester, 15 Korean language learners completed three writing tasks, during which three notable characteristics of prompt literacy emerged: uniformity, discreteness, and passivity. Uniformity describes the tendency of learners to rely on limited and repetitive prompt formats, which constrain their ability to interact dynamically with AI. Discreteness reflects the lack of continuity in interactions, where exchanges with ChatGPT often appear segmented rather than connected, thus limiting deeper engagement. Passivity indicates a prevalent approach in which learners accept AI responses without critically evaluating or expanding on them, resulting in a more superficial use of ChatGPT. These findings highlight that learners tend to use ChatGPT primarily as a tool for information retrieval or translation rather than as a dynamic conversational partner, thereby underutilizing the AI’s full potential. To maximize the educational impact of generative AI, there is a clear need for instructional support aimed at strengthening learners’ prompt literacy skills and encouraging deeper, more interactive engagement with AI tools in language learning.

prompt literacy, ChatGPT, Korean language education, learner-AI interaction, AI-assisted writing

김정태 Jungtae Kim

DOI: Vol.27(No.4) 19-44, 2024

There is a lack of research on the use of the AI voice chatbot in speaking tests. This study compared ESWT (English Speaking & Writing Test) and CEST (Chatbot-based English Speaking Test) scores, analyzed the impact of interactional competence on the wholistic scores of CEST, and examined test-takers’ perceptions of item difficulties. Twenty-seven Korean university students, primarily at novice English language proficiency levels, participated in the study. They took both ESWT and CEST followed by surveys asking their self-perceived item difficulty for each test. The results of paired t-test showed that ESWT yielded higher mean scores across all analytical constructs and holistic scores with significant differences in pronunciation and coherence. A Ridge regression revealed that the interactivity and fluency constructs of CEST strongly influenced its wholistic score, highlighting the potential of AI chatbots to enhance EFL learners' speaking abilities. Participants, however, perceived CEST as more challenging than ESWT, pointing out the need for more preparation and response time due to their low proficiency levels. Regarding item difficulty, read-aloud and self-introduction tasks were considered the easiest, while opinion-giving and long consecutive questions were the most challenging in ESWT. Participants also found all items challenging overall, reflecting their limited language competence and underscoring the necessity of targeted speaking skill development. Based on these findings, the study presents discussions and practical implications for integrating AI chatbots into speaking assessment.

voice chatbots, speaking assessment, beginner, college students, English speaking

김혜숙 Hea-suk Kim , 김나영 Na-young Kim , 김윤진 Yunjin Kim , 손다정 Da-jung Son

DOI: Vol.27(No.4) 45-69, 2024

This meta-analysis investigated current trends in generative AI research in L2 and Korean language education internationally and domestically, aiming to guide future educational directions. Domestically, most studies focused on university students, with fewer studies on middle and high school students, elementary students, and teachers. Similarly, international studies also predominantly focused on university students. In Korean language education, research primarily involved adult learners, highlighting the effective use of AI tools with intermediate to advanced learners. Analysis of participant numbers revealed that most studies in English and Korean education involved 11-50 participants. Methodologically, many studies in English and Korean education employed quantitative methods, whereas international studies used qualitative approaches. Research on language skills predominantly focused on participant perceptions, with subsequent emphasis on speaking and writing. However, there were significant gaps in studies addressing listening and grammar skills. As a result of a comprehensive analysis of language skills, the research emphasized the area of writing, revealing that learners generally exhibited positive perceptions across the studies. To maximize the educational value of ChatGPT, it is crucial to teach students effective questioning techniques. Future research should focus on specific prompt usage methods and enhance self-directed learning using metacognitive strategies.

artificial intelligence, ChatGPT, meta analysis, English education, Korean education

방준성 Junseong Bang , 이상민 Sangmin-michelle Lee

DOI: Vol.27(No.4) 70-86, 2024

The integration of AI into education has catalyzed transformative changes, particularly in enabling customized and personalized learning experiences that were previously impractical in traditional classroom settings. While earlier AI automation systems focused on educational equality by providing uniform resources and opportunities, more recent AI-driven customized learning systems aim to achieve educational equity by providing differentiated support tailored to individual needs, ensuring comparable learning outcomes across diverse student populations. However, these AI-driven customized learning systems can inadvertently introduce bias in both data collection and algorithmic processing, potentially compromising educational equity. This risk is particularly pronounced in foreign language education, where learner populations exhibit significant demographic and cultural diversity, increasing the potential for data bias. With the planned introduction of AI English digital textbooks in 2025, addressing AI bias and educational inequality becomes critical, particularly in English education. This study systematically examines the sources of AI bias that occur at multiple stages: data collection, analysis, classification, and algorithmic processing. It identifies the manifestations and implications of these biases in educational contexts, and proposes comprehensive solutions that include both technical and non-technical approaches, drawing on existing literature. It also outlines recommendations for key stakeholders -developers, educators, and policymakers- to ensure equity in AI-driven educational systems.

artificial intelligence, customized learning system, bias, equity, data

안미리 Miri Ahn , 박인심 Insim Park

DOI: Vol.27(No.4) 87-111, 2024

This study analyzed the case of an international exchange-based online service-learning course conducted from the second semester of 2022 to the first semester of 2024, exploring the experiences of S Women's University students supporting Korean language and culture learning for multinational Korean learners at UNSW and Monash University in Australia. The research aimed to qualitatively analyze participants’ motivations, challenges, and the impact on their English learning to suggest improvements for service-learning courses linked to university English classes. Participants' motivations included personal growth, English proficiency improvement, multicultural communication skills development, introducing Korean culture and language, changing perceptions of teaching experiences, and enhancing problem-solving abilities. Students faced difficulties such as communication barriers due to language and cultural differences, challenges in planning unstructured volunteer activities, time zone differences, and varying levels of partner engagement. The study found that the service-learning activities significantly improved students’ confidence in using English, reduced fear of conversing with foreigners, enhanced speaking skills through continuous interaction, increased motivation for English learning, and improved understanding of effective communication elements. In conclusion, online global service-learning proved to be an effective educational method for enhancing learners’ English proficiency and learning motivation through authentic and regular communication experiences.

online service-learning, language exchange, cultural exchange, multicultural communication, English learning

장성임 Sung-im Jang , 김혜영 Heyoung Kim

DOI: Vol.27(No.4) 112-131, 2024

Digital literacy (DL) is a student-centered learning competency, often referred to as ‘new literacy’. This study aims to investigate digital environment for early childhood English education and the DL competency of English teachers. A survey was conducted with 249 Korean English teachers in early childhood education institutions. The questionnaire included class profile, digital tools and programs utilized in their classrooms, and instructional methods for integrating technology. The results indicate that early childhood learning environment is generally conducive to developing DL. Teachers reported using an average of four devices and more than three programs during English classes, dedicating about half of their class time to teaching storybooks, the alphabet and phonics. However, the approach to using technology remains teacher-centered, primarily involving material presentation to children. There is a lack of technology-assisted games, activities, and projects, resulting in limited opportunities for students to engage directly with devices and programs. Regarding DL teaching competency, teachers rated their skills below ‘occasionally (M=2.97)’ Except for ‘authenticity’, their competency in all other areas fell below 3.0 indicating minimal opportunities for students to develop DL. It is essential for educational authorities or early childhood institutions to monitor and enhance their English curricula and teacher competencies. DL workshops and teacher training programs should be provided for kindergarten English teachers.

digital literacy, teaching competency, early childhood English education, kindergarten English teachers

최희 Hee Choi

DOI: Vol.27(No.4) 132-159, 2024

This research investigates the impact of incorporating technology into English classes on students’ vocabulary ability, specifically through the use of digital flashcards, contextual video clips, and AI-supported collaborative writing tasks, along with associated vocabulary lists. The study involved 216 ninth-grade students over an eight-week period during the spring semester of 2024. Both pre-and post-tests were conducted to measure the effectiveness of the technology-integrated instruction versus traditional methods, which typically involved teacher-led lectures and physical textbooks. Additionally, surveys were administered before and after the experiment to determine if there were any significant changes in the perception of students towards learning English vocabulary and writing skills. Further qualitative data were collected through in-depth interviews with 16 participants to identify the advantages and drawbacks of using digital technologies in education. The statistical analysis using paired t-tests of vocabulary revealed significant progress in both the experimental and control groups, though no notable differences were found between the two. However, the survey results showed an improvement in the attitudes of the experimental group with regard to their vocabulary and writing classes. Interviews indicated that students particularly enjoyed engaging in game-based learning activities such as Quizlet Live and collaborative writing exercises. These results imply that while technological tools can boost student engagement and collaborative learning, educators need to carefully manage potential issues such as video pacing and minimizing distractions or over-reliance on AI tools.

Quizlet, digital technology, vocabulary, writing, AI, gamification

Joohoon Kang , Han Gil Kim

DOI: Vol.27(No.4) 160-178, 2024

This paper examines the role of multimodality in English Language Teaching (ELT), emphasizing its capacity to enhance meaning-making and communication through diverse semiotic modes, including text, images, gestures, and digital tools. The review aims to explore how multimodal approaches can enrich English Language Learners’ (ELLs) academic and social experiences. Using a systematic research synthesis methodology, the study analyzes 30 empirical studies in second language (L2) contexts, focusing on five key areas: (a) digital technologies, (b) EFL contexts, (c) adolescent ELLs, (d) in-and out-of-school settings, and (e) writing genres (narrative, expository, and argumentative). The paper concludes with pedagogical recommendations and future research directions to promote dynamic and inclusive ELT practices, while also addressing notable limitations. These findings provide valuable insights for ELT researchers and educators seeking to implement multimodal frameworks that support holistic and engaging language learning environments.

multimodality, multimodal composition, EFL, L2 literacy

Rakhun Kim

DOI: Vol.27(No.4) 179-199, 2024

This study examines the instructional efficacy and learner perceptions of AI-generated videos for teaching English caused-motion constructions and compares them with human made ones within the Korean EFL context. Employing generative AI to automate the creation of multimodal educational materials is expected to maximize instructional efficiency within Korean EFL settings. However, empirical research comparing AI-generated materials to traditional human-instructed content in language learning remains limited. This study addresses this gap using a mixed-methods design involving ninety high school students divided into the AI-generated and human-made video groups. Participants completed a perception survey and Elicited Writing Tasks (EWT) to assess engagement and learning outcomes. The results from the perception study demonstrate that Human-made videos were preferred for pacing and instructor support, highlighting the importance of human presence in instructional content. Nevertheless, AI-generated videos were also valued for their engaging visuals and efficiency, though participants suggested enhancements in avatar realism and speech clarity. A repeated-measures ANOVA revealed that both AI-generated and human-made videos significantly improved learners’ production of caused-motion constructions in the immediate posttest; however, the instructional effect was not sustained in the delayed posttest. The findings support using AI-generated videos as effective instructional tools and emphasize the need to balance technological efficiency with human-like features for optimal learning experiences. The pedagogical implications of the study are also discussed.

AI-generated video, human-made video, video-based learning, generative AI, caused-motion construction

Juhee Lee , Hansuk Son

DOI: Vol.27(No.4) 200-223, 2024

This study employed a mixed-methods approach to examine the impact of automated written feedback using AI on the English writing skills of Korean high school students. The participants consisted of 32 students learning English as a foreign language (EFL) at a high school in Korea. Over eight weeks, students participated in a 60-minute weekly writing intervention. This intervention included English writing exercises, during which students received automated written feedback from AI tools, revised their drafts using AI tools, and reflected on their learning experiences. Quantitative and qualitative data were collected through pre- and post-writing tests, questionnaires, weekly reflective comments, interviews, students’ AI prompts, and interaction records with AI. The results indicated significant improvement from the pre-test to the post-test, with notable gains in organization, grammar, vocabulary, and mechanics. However, no significant growth was observed in the content of the writing. Qualitative findings revealed that students did not utilize any prompts requesting feedback on the content of their writing, nor did their weekly comments indicate a focus on content development. The data also highlighted factors that facilitated or hindered students’ learning from the AI feedback process. This study provides pedagogical implications for the effective use of AI in EFL high school classrooms, emphasizing the potential benefits and limitations of automated feedback in supporting language development.

Artificial intelligence, automated written feedback, EFL writing, adolescents, mixed-methods