Earticle

Despite growing attention, translanguaging is often treated as a universal practice, with insufficient attention to how it is shaped by Global South English as a Foreign Language (EFL) conditions. This systematic review analyzes empirical studies on translanguaging in the Thai EFL context. Drawing on García et al.’s (2017) and Cenoz and Gorter’s (2021) translanguaging frameworks, the study synthesizes findings from 13 peer -reviewed studies published between 2009 and 2025. Using the Preferred Reporting Items for Systematic Reviews and Meta -Analyses (PRISMA) framework, this review examined how EFL teachers applied translanguaging across different educational levels. The findings showed that translanguaging had become a common instructional approach in Thai EFL. Teachers generally adopted linguistically resource-oriented stances, implemented planned pedagogical designs, and employed frequent translanguaging shifts to scaffold learning and manage classroom interaction. Moreover, the use of the whole linguistic repertoire emerged as the most prev alent practice, while metalinguistic awareness activities were limited, and integrated language curricula were largely absent at the institutional level. The discussion reveals a field in transition, where translanguaging functions simultaneously as both a practical teaching strategy and an emerging pedagogical orientation within an English-hegemony educational system. This study also offers implications for classroom practice, teacher education, and language education policymakers.

ถึงแม้การเช-ือมโอนภาษาจะได้รับความสนใจเพิ-มขึ@น แต่โดยทัว- ไปหลักการนี@มักถูกมองว่าเป็นวิธีการสอน ดังนั@นการศึกษาท-ีเก-ียวข้องกับการเช-ือมโอนภาษา ภายในเงื-อนไขของการสอนภาษาอังกฤษเป็นภาษาต่างประเทศในบริบทซีกโลกใต้จึงยังไม่ได้รับความสนใจเท่าที-ควร การศึกษาเชิงทบทวนวรรณกรรมอย่าง เป็นระบบนี@ทำการวิเคราะห์บทความท-ีเป็นการศึกษาเชิงประจักษ์เก-ียวกับการเช-ือมโอนภาษาในบริบทการสอนภาษาอังกฤษเป็นภาษาต่างประเทศของ ประเทศไทย โดยอาศยั กรอบแนวคิดของ García และคณะ (2017) และ Cenoz และ Gorter (2021) การศึกษานี@สังเคราะห์ผลการวิจัยจากงานวิจัยท-ีผ่านการ ประเมินโดยผู้ทรงคุณวุฒิจำนวน 13 เร-ือง ซึ-งตีพิมพ์ระหว่างปีค.ศ.2009 ถึง 2025โดยใช้กรอบแนวทาง Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) เพ-ือศึกษาว่าครูผู้สอนภาษาอังกฤษเป็นภาษาต่างประเทศใช้การเช-ือมโอนภาษาในชั@นเรียนระดับต่าง ๆ อย่างไร ผลการศึกษาพบว่า การเชื-อมโอนภาษาเป็นกลยุทธ์การสอนที-พบได้แพร่หลายในบริบทการสอนภาษาอังกฤษเป็นภาษาต่างประเทศในประเทศไทย ครูผู้สอนมีจุดยืนที-มองว่า ภาษาเป็นทรัพยากรการเรียนรู้ มีการออกแบบการเรียนการสอนล่วงหน้า และใช้การเช-ือมโอนภาษาแบบกะทันหันอย่างสมํ-าเสมอ โดยมีจุดมุ่งหมายเพ-ือ อำนวยการเรียนรู้และจัดการปฏิสัมพันธ์ในชั@นเรียน นอกจากนี@ยังพบว่ามีการใช้คลังภาษาทั@งหมดของผู้เรียนมากท-ีสุด ในขณะท-ีมีการใช้กิจกรรมท-ีส่งเสริม ความตระหนักรู้เชิงอภิภาษาเพียงเล็กน้อย และแทบไม่พบหลักสูตรภาษาแบบบูรณาการในชั@นเรียนภาษาอังกฤษเป็นภาษาต่างประเทศเลย บทอภิปราย ชี@ให้เห็นว่าการสอนภาษาอังกฤษเป็นภาษาต่างประเทศในประเทศไทยกำลังอยู่ในช่วงเปล-ียนผ่าน โดยการเช-ือมโอนภาษาทำหน้าท-ีเป็นทั@งกลยุทธ์สอน ภาษาอังกฤษเป็นภาษาต่างประเทศและแนวทางการจัดการเรียนการสอนแบบใหม่ที-อยู่ภายใต้ระบบการศึกษาที-ให้ความสำคัญกับการใช้ภาษาอังกฤษเป็น ภาษาหลักในการเรียนการสอนเพียงภาษาเดียว การศึกษานี@ยังเสนอข้อเสนอแนะเชิงนัยสำหรับการปฏิบัติในชั@นเรียน การพัฒนาวิชาชีพครู และผู้กำหนด นโยบายดา้ นการศึกษาภาษาอีกดว้ ย

This research investigates how the Global Citizenship Approach (GCA) influences the multicultural perspectives and English proficiency of EFL (English as a Foreign Language) learners. The study included 30 CEFR B1-level students from Swinburne Vietnam and employed a mixed-methods research design. The EFL learners were exposed to the GCA curriculum, with assessments conducted before and after a two-month intervention. Additionally, participants in the experimental group (n=15) were invited for in-depth interviews to provide insights into their experiences and perceptions of engaging with the GCA. The results showed a statistically significant improvement in language proficiency scores after GCA implementation, indicating a positive correlation between exposure to the GCA approach and English proficiency scores. Furthermore, thematic analysis of the interviews highlighted a deeper appreciation for cultural diversity and heightened awareness of global issues. These findings offer valuable insights into the effectiveness of incorporating a Global Citizenship Approach into EFL curricula in Vietnamese higher education and inform educational practices that foster intercultural competence among language learners.

데이터 기반 학습(Data-driven learning; DDL) 연구는 최근 중등교육 맥락으로 관심을 확장하고 있으나, 이 맥락에서 멀티모달 입력이 학습을 어떻게 지원할 수 있는지는 아직 실증적으로 검토되지 않았다. 본 연구는 멀티미디어 학습의 인지 이론(CTML)에 기반하여 중등 학습자를 위한 멀티모달 DDL을 적용한 최초의 시도 중 하나이다. 연구의 탐색적 성격을 고려하여, 학습 결과와 학습자 인식을 살펴보기 위해 단일집단 설계를 사용하였다. 참여자는 CEFR A2 수준의 한국 중학생 43명이었다. 멀티모달 DDL 수업은 40분씩 총 3차시로 진행되었으며, 각 차시에는 상대적 지위, 사회적 거리, 요청 부담이 낮은 상황과 높은 상황에 각각 적절한 요청 표현을 한 가지씩 다루었다. 학습자들은 차시마다 멀티모달 플랫폼에서 영상을 시청하고 안내 질문에 응답한 뒤, 협력 학습지를 활용하여 짝토의와 요청 표현 말하기 연습을 수행하였다. 사전ㆍ사후ㆍ지연 사후 검사 결과, 사회화용론 점수는 즉시 사후 검사에서 상승했으나 화용언어론에서는 뚜렷한 변화가 나타나지 않았고, 두 지식 영역 모두에서 장기 유지의 증거는 제한적이었다. 플랫폼의 안내 질문에 대한 응답, 인식 설문, 그리고 인터뷰 자료는 참여자들이 영상 이해와 구두 담화 완성 시험 수행에 어려움을 겪었음에도, 멀티모달 단서와 시각적 강화에 주목하여 대화 상황의 사회화용론적 특성을 추론했음을 보여주었다. 본 연구 결과는 중등학습자를 위한 멀티모달 DDL이 사회화용론 발달을 지원할 가능성을 시사하며, 화용언어론적 발달과 장기 학습을 지원하기 위한 향후 연구 및 교수 설계의 방향을 제시한다.

Although data-driven learning (DDL) research has begun to extend to pre-tertiary settings, the potential of multimodal input for supporting learners in these settings has not yet been empirically examined. This study presents one of the first implementations of multimodal DDL for pre-tertiary learners, drawing on the Cognitive Theory of Multimedia Learning (CTML). Given the exploratory nature of the study, a single-group design was used to examine learning outcomes and learner perceptions. Participants were 43 Korean middle school learners at the CEFR A2 level. The main intervention consisted of three 40-minute sessions, each targeting one request expression appropriate for low-PDR situations and another appropriate for high-PDR situations (six total). In each session, participants first used a multimodal DDL platform to view videos and answer guiding questions and then completed pair discussions and speaking practice with a collaborative worksheet. Pre-test, post-test, and delayed post-test scores indicated an immediate increase in sociopragmatic scores, but not pragmalinguistic scores, with little evidence of long-term retention in either domain. Participants’ answers to the guiding questions, questionnaire responses, and interviews revealed that, despite difficulties in video comprehension and spoken discourse completion tests (DCTs), they drew on multimodal cues and visual enhancements to infer sociopragmatic features of communicative situations.

This study examines how novice Japanese high school EFL learners deploy epistemic-claim formats (e.g., I know and the contextually related Japanese form, wakaru) to sustain progressivity in classroom peer interaction. It addresses two research questions: (a) in what sequential environments these formats emerge, and (b) how they shape participation to sustain progressivity under linguistic constraint. Drawing on multimodal conversation analysis of about 450 minutes of naturally occurring classroom peer interaction in task-based activities (91 pair recordings), the study shows that these formats occur in environments of delay, such as hesitation or word search, where progressivity becomes vulnerable. Functioning as warrants for speakership, epistemic-claim formats allow speakers to claim entitlement to the relevant domain, which recipients locally ratify as eligibility. Across the collection, these formats secure self-repair space, resist premature peer assistance, and enable collaborative intervention based on displayed shared epistemic access. These findings extend conversation-analytic work on epistemics by illustrating how explicit epistemic claims function as participation-management resources in novice L2 interaction. Pedagogically, the study suggests that learner silence and hesitation should be treated as interactionally organized moments of formulation rather than epistemic insufficiency, calling for instructional practices that support self-repair and interactional competence development.

本研究は、日本の高等学校における初級EFL学習者が、教室内の学習者同士の対話において、 進行性 (progressivity) を維持するために、認識主張の形式 (epistemic-claim formats: I know, わ かるなど) をどのように用いているかを検討するものである。本研究は以下の2つの問いに取 り組む。(a) これらの形式は、どのような連鎖環境で出現するのか、(b) これらの形式は、言 語的制約の下で学習者の参加のあり方をどのように形づくり、相互行為の進行性をどのよう に維持するのか。分析は、マルチモーダル会話分析 (multimodal conversation analysis) を用い、 課題遂行型活動における教室内会話 (約 450分: 91 件のペア会話録音) を対象に行った。分析 の結果、これらの形式は、ためらいや語彙検索など、進行性が脆弱になる遅延場面において 出現する。そして、発話継続の保証 (warrant) として機能し、話者はそれを根拠として関連す る知識領域への権利 (entitlement) を主張する。さらに、この権利は受け手の後続行為を通し て、局所的な適格性 (eligibility) として承認される。また、これらの形式は自己修復の機会を 確保したり、求められていない他者からの支援を退けたり、共有された認識的アクセスを示 すことで協働的な介入を正当化したりする働きをもつ。これらの知見は、明示的な認識主張 の形式がL2初級学習者同士の相互行為における参加管理の資源として機能することを示し、 認識性 (epistemics) に関する会話分析的研究を拡張するものである。また教育的示唆として は、学習者の沈黙や躊躇は、認識的な不足の表れではなく、発話形成が進行中であることを 示す相互行為的に組織化された局面として捉えるべきであり、自己修復の機会を確保し、相 互行為能力の発達を支える指導実践が求められる。

While the term English academic literacy has become a buzzword in the literature, the construct of academic literacy has rarely been investigated empirically, especially in an English-as-a-foreignlanguage context. The first aim of this study was therefore to validate an English academic literacy scale that reflects the prevalent multidimensional perspective of academic literacy in the literature. The second aim of the study was to identify latent profiles of English academic literacy endorsed by Chinese university students and the associations between these profiles and students’ English learning achievement. Data were collected from 739 English major students from two Chinese universities. Confirmatory factor analysis results corroborated the hypothesized five-dimensional structure of English academic literacy grounded in the various theoretical perspectives, including Linguistic Knowledge, Reading Skills, Writing Skills, Subject Content Knowledge, and Social Learning Practice. Subsequent latent profile analysis (LPA) revealed three distinct profiles: “Struggling Literacy Learners”, “Developing Literacy Learners” and “Proficient Literacy Learners” among the English major students in this study. Pairwise Wald tests results indicated significant associations between English academic literacy profiles and learning achievement. This study ends by providing theoretical and empirical implications for English academic literacy education in an EFL context.

尽管‘英语学术素养’（English academic literacy）一词已成为文献中的热门概念，但学术素养这 一建构却鲜有实证研究，尤其是在英语作为外语（EFL）的环境下。因此，本研究的首要目的在 于验证一个英语学术素养量表，以反映文献中主流的学术素养多维视角。本研究的第二个目的， 则是为了识别中国大学生所具有的英语学术素养潜在剖面（latent profiles），并探究这些剖面与 学生英语学习成绩之间的关联。研究数据收集自中国两所大学的739 名英语专业学生。验证性 因子分析结果证实了基于多种理论视角所假设的英语学术素养五维结构，包括：语言知识、阅 读技能、写作技能、学科内容知识以及社会学习实践。随后的潜在剖面分析（LPA）显示，本研 究中的英语专业学生群体存在三种截然不同的剖面：‘挣扎型素养学习者’（Struggling Literacy Learners）、‘发展型素养学习者’（Developing Literacy Learners）和‘熟练型素养学习者’（Proficient Literacy Learners）。两两瓦尔德检验（Pairwise Wald tests）结果表明，英语学术素养剖面与学习 成绩之间存在显著关联。本研究最后为外语环境下的英语学术素养教育提供了理论与实证层面 的启示。

본 연구는 영어 학업 저성취자를 분류하기 위한 8가지 머신러닝 모델을 비교하고, 축소 모델이 유의미한 예측 성능을 유지할 수 있는지 검토하였다. 이를 위해 한국 고등학생 728명을 대상으로 자기보고식 설문을 통해 2017년도에 수집된 이차 데이터가 활용되었으며, 생물학적, 사회문화적, 인지적, 비인지적 영역에 걸친 23개의 예측 변수가 모형 검증에 활용되었다. 모형 검증은 70:30 훈련 및 검증 데이터 분할, 훈련 데이터에만 적용된 ADASYN 기법, 그리고 10겹 층화 교차검증으로 구성된 엄격한 검증 체계 하에서 실시되었다. 그 결과, 로지스틱 회귀 모델이 가장 균형적인 분류 성능을 보이는 것으로 나타났다. 집단 순열 중요도 분석 결과, 거주 지역, 자기조절 효능감, 과제난이도 효능감, 영어 흥미, 영어 학습 총 기간, 영어 자신감, 수행접근 목표, 사교육 경험, 인지 전략, 보완 전략이 상위 10개의 중요 핵심 예측 변수로 확인되었으나, 변수별 안정성에는 차이가 확인되었다. 상위 10개 변수를 기반으로 구성한 축소 로지스틱 회귀 모델은 중첩 교차검증에서 전체 모델보다 모든 지표에서 다소 낮은 성능을 보였으나, 그 차이는 크지 않아, 예측 변수군의 절반 이하만으로 전체 모델 성능의 상당 부분을 유지하는 것으로 나타났다. 이러한 결과는 분류 기반 예측이 영어 학업 부진 위험 학생을 식별하는 데 있어 중간 수준이지만 의미 있는 유용성을 지님을 시사하며, 축소 모델이 자료 수집 부담을 크게 줄이면서도 선별 목적의 교육적 활용에 실용적인 선택지가 될 수 있음을 시사한다.

This study compared eight machine learning models to classify English academic underachievement and examined whether a reduced-feature model could retain useful predictive performance. Using secondary self-reported survey data from 728 Korean high school students, the study analyzed 23 predictors across biological, sociocultural, cognitive, and non-cognitive domains. Under a rigorous validation framework involving a 70:30 train-test split, training-only ADASYN, and 10-fold stratified cross-validation, Logistic Regression showed the most favorable overall balance of performance. Grouped permutation importance identified region, self-regulatory efficacy, task-difficulty efficacy, interest in English, total years of English learning, English self-confidence, performance-approach goal, private tutoring experience, cognitive strategy, and compensation strategy as the top 10 predictor groups, although their stability varied. Under nested cross-validation, a reduced Logistic Regression model based on these feature groups performed slightly below the full model across all indices, though the differences were modest, suggesting that most of the full model’s predictive performance was retained with fewer than half of the predictor groups. The findings indicate that classification-oriented prediction has moderate but meaningful utility for identifying students at risk of English academic underachievement and that reduced-feature models may remain a practical option for screening-oriented educational applications by substantially reducing data collection burden.

Paraphrasing is a pivotal academic writing skill that enables writers to restate source ideas accurately in sufficiently transformed language while mitigating patchwriting. Yet many EFL learners struggle to preserve meaning while reformulating source language. This study examined three questions: the extent of change in Vietnamese EFL undergraduates’ paraphrasing performance following sustained strategy instruction, which dimensions of paraphrase quality showed the greatest gains, and what challenges learners reported during the intervention. A mixed-methods one-group pretest–posttest intervention design was used with 42 undergraduates in Vietnam over six weeks. Participants completed pre- and post-intervention paraphrasing assessments scored with an analytic rubric, while reflective journals and semi-structured interviews provided qualitative evidence. Paired-samples t tests and effect sizes examined performance gains, and qualitative data were analysed thematically. Results showed statistically significant improvement in overall paraphrasing performance. Analytic dimension scores indicated the largest proportional gain in meaning preservation, followed by syntactic transformation and idea organisation. Qualitative findings showed that lexical and grammatical control remained recurring constraints during strategy application, alongside concern about semantic distortion, time pressure, and overreliance on technology during independent practice. The findings support sustained strategy instruction while highlighting the need for continued scaffolding in lexical development, meaning verification, and responsible technology use.

Paraphrasing là kỹ năng quan trọng trong viết học thuật, giúp trình bày chính xác ý tưởng nguồn và hạn chế patchwriting. Nghiên cứu xem xét thay đổi năng lực paraphrasing của sinh viên EFL Việt Nam, thành phần chất lượng tăng nhiều nhất, và khó khăn trong sáu tuần giảng dạy chiến lược. Thiết kế hỗn hợp một nhóm tiền kiểm–hậu kiểm áp dụng với 42 sinh viên. Người tham gia hoàn thành đánh giá trước và sau can thiệp; nhật ký và phỏng vấn cung cấp dữ liệu định tính. Kiểm định t, kích thước hiệu ứng và phân tích chủ đề được sử dụng. Kết quả cho thấy năng lực paraphrasing tổng thể cải thiện có ý nghĩa thống kê. Điểm thành phần cho thấy mức tăng lớn nhất ở bảo toàn ý nghĩa, tiếp theo là biến đổi cú pháp và tổ chức ý tưởng. Dữ liệu định tính cho thấy hạn chế từ vựng, ngữ pháp, lo ngại sai lệch ý nghĩa, áp lực thời gian và phụ thuộc công nghệ vẫn tồn tại. Kết quả ủng hộ giảng dạy chiến lược và nhu cầu hỗ trợ từ vựng, kiểm chứng ý nghĩa, sử dụng công nghệ có trách nhiệm.

Multimodal feedback, particularly video feedback, has been shown to offer significant benefits for language development. However, research exploring dialogic multimodal feedback (DMF) for diverse learner needs in L2 academic writing remains limited. This study adopted a mixed-method case study design to examine student and lecturer perceptions of DMF affordances and the perceived impacts of DMF on students' L2 academic writing. The study was conducted with a class of 31 Vietnamese university students in an English-medium instruction (EMI) course requiring a formative writing assignment. Data were gathered through a survey, follow-up interviews, and a teacher reflective journal. Qualitative data were analysed thematically, whilst quantitative data were analysed using SPSS and SurveyMonkey. The findings revealed three main types of DMF affordances: pedagogical (multimodal sources addressing diverse learning needs), technical (replay, transcript, and reaction functions), and social (discussion capabilities). Both the students and the lecturer expressed more favourable perceptions of DMF than of conventional text feedback. DMF was perceived to foster deeper understanding through multiple modalities, to engage students with varying preferences, to support individual learning goals, and to reduce teacher workload when appropriately designed. The interactivity function empowered students to question and reflect on feedback, transforming passive monologic practices into reciprocal knowledge co-construction. Notably, different learners perceived and acted upon the available affordances in different ways, contributing to different perceived impacts on writing. Some limitations were identified, along with recommendations for optimal implementation.

Phản hồi đa phương thức, đặc biệt là phản hồi qua video, có tiềm năng mang tới những lợi ích đáng kể cho quá trình dạy học ngôn ngữ. Tuy nhiên, vẫn còn rất ít nghiên cứu về phản hồi đa phương thức mang tính đối thoại (dialogic multimodal feedback – DMF) cho kỹ năng viết học thuật bằng ngoại ngữ/ngôn ngữ thứ hai của sinh viên với nhu cầu học tập đa dạng. Nghiên cứu này sử dụng thiết kế nghiên cứu trường hợp có kết hợp định tính và định lượng (mixed-method case study design) để tìm hiểu quan điểm của các bên liên quan về những cơ hội học tập (affordances) và ảnh hưởng của phản hồi đa phương thức mang tính đối thoại đối với kỹ năng viết học thuật bằng ngoại ngữ/ngôn ngữ thứ hai của sinh viên. Nghiên cứu được thực hiện với 31 sinh viên trong một khóa học chuyên ngành bằng tiếng Anh (EMI), trong đó sinh viên được yêu cầu làm bài viết để đánh giá quá trình. Các công cụ nghiên cứu chính bao gồm bảng hỏi cho sinh viên, phỏng vấn sâu với sinh viên và nhật ký phản tư của giảng viên. Phân tích chủ đề được áp dụng với dữ liệu định tính, trong khi đó, dữ liệu định lượng được phân tích qua SPSS và SurveyMonkey. Kết quả nghiên cứu cho thấy, phản hồi đa phương thức mang tính đối thoại mang lại ba nhóm cơ hội học tập chính cho sinh viên về khía cạnh sư phạm (các nguồn thông tin đa phương thức hỗ trợ nhu cầu học tập đa dạng), khía cạnh kỹ thuật (dễ dàng tua lại, dùng bản ghi chép lời thoại hay các biểu tượng phản hồi), khía cạnh xã hội (dễ dàng thảo luận với giảng viên). Cả giảng viên và sinh viên đều yêu thích phản hồi đa phương thức mang tính đối thoại hơn so với hình thức phản hồi bằng văn bản truyền thống. Phản hồi đa phương thức mang tính đối thoại được cho là có thể giúp sinh viên hiểu nhận xét sâu hơn thông qua kết hợp đa phương thức, tăng cường sự tham kết của sinh viên với sở thích học tập đa dạng, hỗ trợ mục tiêu học tập cá nhân, và giảm khối lượng công việc của giảng viên nếu được thiết kế hợp lý. Tính năng tương tác giúp trao quyền cho sinh viên để sinh viên đặt câu hỏi chủ động, phản tư về nhận xét của giảng viên, chuyển đổi hình thức phản hồi một chiều sang quá trình đồng kiến tạo tri thức hai chiều. Đáng chú ý là những người học khác nhau có quan điểm và hành động khác nhau dựa trên những cơ hội học tập sẵn có, dẫn tới quan điểm khác nhau về tác động của hình thức nhận xét này lên kỹ năng viết. Sinh viên và giảng viên cũng chỉ ra một số hạn chế và đề xuất để áp dụng phản hồi đa phương thức mang tính đối thoại hiệu quả hơn.

본 연구는 호텔영어 학습에서 인공지능(AI) 리터러시 함양을 목표로, Coh-Metrix 프로그램을 활용하여 호텔 영어 교재 3종을 대상으로 가독성, 어휘 정교성, 통사적 응집성 등 열 가지 언어적 차원을 분석하였다. 분석 결과, 세 가지 제약 범주가 확인되었다. 지시적·의미적 중첩, 후기 습득 추상 어휘를 동반한 높은 유형-토큰 비율, 그리고 확장된 동사 구간을 포함한 높은 명사구 밀도이다. 이러한 언어적 특성은 학습자의 정보처리 부담을 가중시킬 수 있으나, 호텔 영어 교재들은 이를 완화할 명시적 지원을 충분히 제공하지 못하고 있었다. 이러한 코퍼스 기반의 격차를 줄이기 위해, 본 연구는 언어적 제약 범주를 AI 기반의 개입을 통한 비계(Scaffolding) 교수 모형을 제안한다. 이 모형은 지시적·의미적 중첩을 확인하기 위한 AI 응집성 감사 (AI Cohesion Auditor), 어휘 정교성을 위한 통제된 용어집 및 연어 생성 (Controlled Glossary and Collocation Generator), 통사적 밀도를 위한 이중 수준 재작문 (Two-Level Rewriter), 그리고 이 세 기술을 학습자 주도적 AI 활용으로 통합하는 제약 기반 프롬프트 설계 (Constraint-Based Prompt Designer)로 구성된다. 각 개입은 해당 AI 리터러시 영역과 연계되며, 프런트 데스크, 하우스키핑, 식음료 서비스 맥락에 걸쳐 제시된다. 본 연구는 호텔 영어교재에서 다루어 지지 않은 언어적 제약을 AI 리터러시 개발을 위한 훈련의 장으로 재구성함으로써, 향후 실증적 검증이 필요한 이론적 기반의 접근을 제공한다.

This study is aimed at addressing the need for artificial intelligence (AI) literacy within English for Specific Purposes (ESP) education in the hospitality industry. The study used the Coh-Metrix program to examine readability, lexical sophistication, and syntactic cohesion across ten linguistic dimensions in three hotel-English coursebooks. In the findings, the study revealed three constraint categories: weak referential and semantic overlap, elevated type-token ratios with late-acquired abstract vocabulary, and high noun phrase density with extended pre-verbal spans. Such features may increase processing demands, yet the coursebooks offer limited explicit support to help learners manage this increased load. Building on this corpus-informed gap, the study proposes an AI-enhanced scaffolding instructional model that maps each constraint category onto a targeted intervention. They are the AI Cohesion Auditor for referential and semantic overlap, the Controlled Glossary and Collocation Generator for lexical sophistication, the Two-Level Rewriter for syntactic density, and the Constraint-Based Prompt Designer, which synthesizes these skills into learner-directed AI use. Each intervention aligns with a corresponding AI literacy domain and is illustrated across reception, housekeeping, and food and beverage contexts. The study repurposes constraints unaddressed by coursebooks as training grounds for AI literacy development, offering a theoretically grounded approach that remains to be empirically validated.

The increasing use of artificial intelligence (AI) has expanded opportunities for oral English practice beyond traditional classrooms. In vocational tourism education, however, oral English functions primarily as a workplace communication skill rather than an academic objective, and the pedagogical value of AI-assisted language learning remains underexplored, particularly in practice-oriented contexts. This qualitative study examines AI-assisted oral English learning in a vocational tourism programme in China, drawing on focus group discussions with 12 tourism students and semi-structured interviews with two English instructors. Participants perceived AI-assisted tools as providing accessible and low-pressure opportunities for oral English practice, with reported benefits including reduced speaking anxiety, increased participation, and support for tourism-specific language use. At the same time, participants reported notable limitations, difficulty processing lengthy or complex AI responses, and difficulties sustaining extended spoken interaction. Crucially, both students and teachers emphasised that the pedagogical value of AI-assisted learning depends on pedagogical scaffolding, task design, and institutional support rather than on technology alone. Overall, the study positions AI-assisted oral English learning as a context-dependent pedagogical resource rather than a self-sufficient instructional solution, offering qualitative insights into its conditional value in vocational tourism education.

人工智能（AI）的广泛应用为传统课堂之外的英语口语练习提供了更多可能。然而，在职业旅游教育中，英语口语主要服务于职场交际，其教学目标和应用情境具有较强的实践性，人工智能辅助语言学习在这一领域的教学价值仍有待深入探讨。本研究采用质性研究方法，以中国某职业院校旅游专业为研究情境，通过对12名旅游专业学生开展焦点小组讨论，并对2名英语教师进行半结构式访谈，考察人工智能辅助英语口语学习的实际体验与教学应用。研究发现，参与者普遍认为，人工智能辅助工具能够提供较为便捷、低压力的口语练习机会，有助于缓解口语表达时的紧张感、提高参与意愿，并为旅游专业相关语言的练习提供支持。同时，参与者也提到，人工智能在实际使用中仍存在一些局限，如较长或较复杂的回复不易理解，以及难以维持持续、深入的口语互动。学生和教师均强调，人工智能辅助学习能否在旅游英语教学中发挥作用，不仅取决于技术本身，还受到教师支架、任务设计和制度支持等因素的影响。总体而言，本研究表明，人工智能辅助英语口语学习更适合作为一种结合具体教学情境使用的辅助性资源，而非独立的教学解决方案，并为人工智能在职业旅游英语教育中的合理应用提供了质性研究依据。并为理解其在职业旅游教育中的条件性价值提供了质性研究视角。

생성형 인공지능(AI)의 급속한 발전은 제2언어(L2) 쓰기 환경을 변화시키고 있으며 이에 따라 AI 보조 쓰기에 대한 학습자의 수용은 중요한 연구 주제로 떠오르고 있다. 본 연구는 수정 및 확장된 기술수용모형(Technology Acceptance Model, TAM)을 바탕으로 한국의 신입 예비교사들이 AI 보조 영어 쓰기를 어떻게 수용하는지를 살펴보았다. 자료는 정규 대학 학기가 시작되기 전 오리엔테이션 기간에 232명의 1학년 예비교사를 대상으로 수집되었다. 설문지는 지각된 사용용이성(Perceived Ease of Use), 지각된 위험(Perceived Risk), 지각된 유용성(Perceived Usefulness), 주관적 규범(Subjective Norms), 자기효능감(Self-efficacy), 사용의도(Intention to Use), 그리고 자기보고식 AI 보조 쓰기 사용(Self-reported AI-assisted writing use)을 측정하였다. 기술통계 및 상관분석 결과, 참여자들은 전반적으로 AI 보조 쓰기에 대해 긍정적인 인식을 보였으나 높은 지각된 위험 점수는 수용도가 아니라 AI 사용에 대한 우려가 높음을 의미한다. 확인적 요인분석(CFA)은 제안된 7요인 측정 구조를 부분적으로만 결과를 나타냈으며, 이에 따라 후속 회귀분석은 탐색적 분석으로 해석하였다. 그 결과, 주관적 규범과 자기효능감은 지각된 유용성을 유의하게 예측하였으며, 지각된 유용성, 자기효능감, 주관적 규범은 사용의도를 유의하게 예측하였다. 또한 사용의도와 자기효능감은 자기보고식 AI 보조 쓰기 사용과 유의한 관련성을 보였다. 반면, 지각된 사용용이성과 지각된 위험은 관련 회귀모형에서 유의한 기여를 보이지 않았다. 이러한 결과는 대학 입학 시점의 예비교사들이 보이는 AI 수용에 대한 기초 자료를 제공하며 영어 쓰기 교육에서 AI 리터러시 교육의 중요성을 강조한다.

The rapid growth of generative artificial intelligence (AI) has transformed second language (L2) writing, which makes learners’ acceptance of AI-assisted writing an important area of research. This study investigated newly admitted Korean pre-service teachers’ acceptance of AI-assisted English writing using an adapted extended Technology Acceptance Model (TAM). Data were collected from 232 first-year pre-service teachers during university orientation, prior to the start of their formal university coursework. The questionnaire measured Perceived Ease of Use (PEOU), Perceived Risk (PR), Perceived Usefulness (PU), Subjective Norms (SN), Self-efficacy (SE), Intention to Use (IU), and self-reported AI-assisted writing use. Descriptive and correlational analyses indicated generally favorable perceptions of AI-assisted writing, although high PR scores represented concern instead of acceptance. The confirmatory factor analysis provided partial support for the proposed 7-factor measurement structure; therefore, the subsequent regression analyses were exploratory. The regression analyses found that SN and SE predicted PU, while PU, SE, and SN significantly predicted IU. IU and SE were also significantly associated with self-reported AI-assisted writing use. PEOU and PR did not make significant contributions in the relevant regression models. These findings provide baseline evidence of AI acceptance among newly admitted pre-service teachers and emphasize the importance of responsible AI literacy in English writing instruction.

This study investigated the separate and combined effects of processing instruction (PI), which focuses on input-based explicit rules, and syntactic priming (SP), which focuses on output-based implicit learning through repetition, on object relative clauses (ORCs) in the English as a Foreign Language (EFL) classroom in Japan. Participants comprised 71 Japanese EFL learners from a technical college, randomly assigned to four groups (PI, SP, PI+SP, and control). All groups undertook a pre-test, intervention session, immediate post-test, and delayed post-test. The three intervention groups received ORC instruction, whereas the control group did not. In each test, participants processed 24 sentences and answered comprehension questions. Accuracy and reading times were compared across groups and clause types. The findings demonstrated that instruction groups improved in both comprehension accuracy and reading speed after the intervention compared with control group, and these gains lasted for four weeks. However, no significant differences were observed among the three intervention groups. These findings highlighted that since the combination of PI and SP did not outperform the isolated approaches, teachers may choose either based on instructional feasibility.

Энэхүү судалгаагаар Японы Англи хэлний Гадаад хэл (EFL) ангид тодорхой дүрмүүдэд анхаарлаа хандуулдаг боловсруулалтын заавар (PI) болон давталтаар дамжин далд суралцахад анхаарлаа хандуулдаг синтаксик прайминг (SP)-ийн тусдаа болон хосолсон нөлөөллийг судалсан. Оролцогчид нь техникийн коллежийн 71 Японы EFL суралцагчдаас бүрдэж, санамсаргүй байдлаар дөрвөн бүлэгт (PI, SP, PI+SP, болон хяналтын бүлэг) хуваарилагдсан. Бүх бүлгүүд урьдчилсан шалгалт, интервенцийн хичээл, шууд дараах шалгалт, хойшлуулсан дараах шалгалтыг хийсэн. Гурван интервенцийн бүлэг ORC зааварчилгаа авсан бол хяналтын бүлэгт заагаагүй. Тест бүрт оролцогчид 24 өгүүлбэр боловсруулж, ойлголтын асуултанд хариулсан. Нарийвчлал болон унших хугацааг бүлгүүд болон өгүүлбэрийн төрлүүдээр харьцуулсан. Судалгааны үр дүнгээс харахад хяналтын бүлэгтэй харьцуулахад интервенцийн дараа ойлголтын нарийвчлал болон унших хурд хоёулаа сайжирсан бөгөөд эдгээр өсөлт дөрвөн долоо хоног үргэлжилсэн. Гэсэн хэдий ч гурван интервенцийн бүлгийн хооронд мэдэгдэхүйц ялгаа ажиглагдаагүй. Эдгээр үр дүнгээс харахад PI болон SP-ийн хослол нь тусгаарлагдсан аргуудаас давж гараагүй тул багш нар сургалтын хэрэгжих боломжид үндэслэн аль нэгийг нь сонгож болно гэдгийг онцолсон.

본 연구는 대학 EFL 학습자들이 L2 글쓰기 과정에서 ChatGPT를 협력적으로 활용하는 방식과 이러한 경험이 이후 개인의 AI 활용에 미치는 영향을 살펴보았다. 한국의 영어교육과 1학년 학생 17명이 15주간의 작문 수업에 참여하여 ChatGPT를 활용한 협력적 글쓰기, 개별 글쓰기 과제, 성찰일지 및 인터뷰를 수행하였다. 분석 결과, 학생들은 ChatGPT를 사용하는 하나의 공통된 방식을 따르기보다, ChatGPT를 언제 활용할지, 제안을 어떻게 평가할지, 글쓰기에서 어떤 역할을 부여할지에 대해 각 집단마다 서로 다른 협력적 활용 방식을 형성하였다. 또한 시간이 지나면서 학생들은 ChatGPT를 단순히 정확한 언어 표현을 제공하는 도구로 보기보다, 그 제안을 토론하고 비판적으로 평가해야 하는 협력적 파트너로 인식하게 되었다. 이러한 협력 경험은 이후 개인적으로 ChatGPT를 사용할 때에도 보다 전략적이고 성찰적이며 목적의식 있는 활용으로 이어졌다. 연구 결과는 교육에서 ChatGPT의 역할이 기술 자체의 기능에 의해 결정되기보다 학습자 간 협력적 상호작용을 통해 협상되고 형성된다는 점을 강조한다.

Focusing on the collaborative use of generative AI in L2 writing, this study examines how university EFL students engage with ChatGPT in group writing tasks and how these experiences shape their subsequent individual use of AI. Seventeen first-year English education majors in Korea participated in a 15-week composition course where they completed collaborative writing tasks using ChatGPT, individual writing assignments, reflective journals, and interviews. Analysis of these data revealed that students did not adopt a single approach to using ChatGPT. Instead, each group developed distinct collaborative practices regarding when ChatGPT should be consulted, how its suggestions should be evaluated, and what role it should play during writing. Over time, students shifted from viewing ChatGPT primarily as a source of correct language to treating it as a collaborative partner whose suggestions required discussion and critical evaluation. These collaborative experiences also influenced students’ subsequent individual engagement with ChatGPT by encouraging more strategic, reflective, and purposeful use of AI. The findings highlight the importance of understanding ChatGPT as a resource whose educational role is negotiated through collaborative interaction rather than determined solely by its technological capabilities.

Despite the increasing prevalence of online tests in the past decade, only a limited number of studies have explored their impact. In order to fill this gap, a study was conducted using a regression discontinuity design to investigate the transition from a paper-based test to an online-based Test of English for International Communication (TOEIC®). Students in high schools and universities take this test for the further advancement of their English proficiency and for placement purposes; thus, the ramifications of the test’s outcomes have been large. The sample size for the research was 2,019 participants, which was the case during the end of their second year of college. The participants took the test over a four-year period from 2019 to 2022. While paper-based format was used in the first two years, online-based format was used in the last two years, with there being a cutoff point between 2020 and 2021. Data were analyzed using descriptive statistics and regression analysis to examine the overall score trends. The results revealed that the two test formats differed considerably and produced contrasting outcomes. These findings highlight the need for caution when interpreting scores from different test formats, and suggest several avenues for future research.

본 연구는 한국 초등학생들의 AI 활용 영어 쓰기 과정에서 나타나는 ChatGPT 활용 양상을 자기조절학습의 관점에서 탐색하였다. 특히 연속적 쓰기 과업을 수행하는 과정에서 프롬프트 사용 방식, 피드백 수용, 사후 성찰에서 나타나는 변화에 초점을 두었다. 이를 위해 초등학교 6학년 4명을 대상으로 다중 사례 연구를 실시하였다. 분석 자료는 학습자-ChatGPT 상호작용 로그, 과업별 자기보고서 및 인터뷰, 초고 및 수정본, 사후 인터뷰를 포함하였다. 연구 결과, 세 가지 분석 차원에서 모두 공통적인 변화 양상이 관찰되었다. 먼저 프롬프트 사용 측면에서 참여자들은 직접적인 텍스트 생성 요청에서 벗어나 점차 ChatGPT 응답의 초점, 범위 및 수준을 자신의 목적에 맞춰 조절하고 구체적인 도움을 요청하는 방식으로 전환하였다. 피드백 수용 측면에서, 자신의 의사소통 의도와 언어적 직관을 바탕으로 ChatGPT의 제안을 평가하고, 선택적으로 수용하거나 거부하는 경향이 보다 구체적으로 드러났다. 사후 인터뷰 분석에서는 ChatGPT와의 상호작용을 조절하고 피드백을 선별적으로 통합하는 방식에 대한 비판적 성찰이 관찰되었다. 다만 이러한 비판적 성찰이 항상 ChatGPT의 성공적인 전략적 활용으로 이어지는 것은 아니었으며, 학습자의 제 2 언어 지식과 주체성이 조절변수로 작용할 가능성이 시사되었다. 이러한 결과는 초등 수준의 EFL 학습자도 ChatGPT 를 단순한 텍스트 제공 도구가 아니라 자신의 글쓰기 의사결정을 지원하는 상호작용적 자원으로 활용할 수 있고, 자기조절적이고 전략적인 AI 활용 능력을 발달시킬 수 있음을 시사한다. 제 2 언어 글쓰기에서의 자기조절적 AI 활용을 지원하기 위해, 학습자의 제 2 언어 자신감과 주체성을 촉진하는 지속적인 스캐폴딩의 필요성이 제언된다.

This study examined how Korean elementary EFL learners engaged in self-regulated use of ChatGPT during AI-assisted writing, focusing on prompting practices, feedback uptake, and reflective evaluation across successive writing tasks. Adopting a multiple case design, the study involved four sixth-grade Korean EFL learners. Data comprised learner-ChatGPT interaction logs, structured self-reports, post-session interviews, initial and revised drafts, and a post-study interview. The findings revealed con-sistent directional shifts across all three analytical dimensions. Regarding prompting, participants pro-gressively moved away from direct text generation requests toward more dialogic, evaluative, and guid-ance-oriented strategies. Regarding feedback uptake, participants increasingly evaluated ChatGPT sug-gestions against their own communicative intentions and linguistic intuitions, selectively incorporating or rejecting feedback. Regarding reflective evaluation, participants demonstrated growing awareness of how to regulate their ChatGPT interactions and selectively incorporate its feedback. Notably, evaluative awareness alone did not uniformly enable strategic use; linguistic knowledge and learner agency ap-peared as moderating factors. The findings suggest that elementary-level EFL learners are capable of developing self-regulated and strategic orientations toward ChatGPT, repositioning it from a text pro-vider to an interactive resource supporting their own writing decisions. Sustained instructional scaffold-ing targeting linguistic confidence and learner agency is recommended to support self-regulated AI use in L2 writing.

여러 영문법서에서는 on the contrary 가 선행문에 제시된 명제를 반박하는 표현이라고 설명한다(예: Cowan, 2008; Larsen-Freeman & Celce-Murcia, 2016; Swan, 2017). 그러나 Lake(2004)는 It doesn’t seem ugly to me; on the contrary, I think it’s rather beautiful 과 같은 예에서 on the contrary 가 선행 진술을 반박하는 것이 아니라 오히려 재확인한다고 지적한다. Lake 의 주장은 해당 예에 대해서는 타당하지만, 이를 그대로 받아들일 경우 on the contrary 는 서로 모순되는 두 의미를 지닌 자가반의어적(contranymic) 부사구가 된다. 본 연구는 Brown Corpus 와 Frown Corpus, 그리고 Corpus of Contemporary American English 의 두 하위 코퍼스에서 추출한 on the contrary 용례를 분석하고, 12종의 주요 참고문법서가 이 표현을 어떻게 기술하는지 검토한다. 분석 결과, Lake 가 제안한 교육적으로 혼란스러운 설명은 다음과 같은 일관된 설명을 채택함으로써 해결될 수 있음을 보인다. 즉, on the contrary 는 항상 반박을 나타내며, 선행문이 부정문인 경우에는 no, not, never, no longer 와 같은 부정어 뒤에 오는 부분만을 반박한다.

Many grammars explain that on the contrary contradicts the proposition expressed in the preceding statement (e.g., Cowan, 2008; Larsen-Freeman & Celce-Murcia, 2016; Swan, 2017). Lake (2004), however, points out that on the contrary reiterates, rather than contradicts, the previous statement in examples such as It doesn’t seem ugly to me; on the contrary, I think it’s rather beautiful. Although Lake’s argument is not incorrect, adopting his argument results in on the contrary being a contranymic adverbial, that is, an adverbial with two contradictory meanings. By analyzing the tokens of on the contrary retrieved from the Brown Corpus and the Frown Corpus, as well as from two sub-sections of the Corpus of Contemporary American English, and surveying 12 well-known reference grammars on their treatment of the phrase, this paper shows that the pedagogically confusing explanation proposed by Lake can be remedied if we instead adopt the following explanation: on the contrary is always used for contradicting; preceded by a negative statement, however, on the contrary contradicts only the part that comes after a negative element such as no, not, never, and no longer.

The increasing integration of generative artificial intelligence (AI) in language education has transformed how written feedback is delivered in EFL writing classrooms. While previous studies have examined linguistic outcomes and revision quality, less attention has been paid to learners’ affective engagement with AI-generated feedback over time. This study investigates how EFL students’ emotional responses to ChatGPT-generated feedback evolve across three genre-specific paragraph-writing tasks. Adopting a 16-week longitudinal qualitative design, the study involved 36 undergraduate EFL students enrolled in Paragraph Writing course. Students engaged with ChatGPT feedback while revising three genres: comparison–contrast, cause–effect, and argumentative paragraphs. Data were collected through 108 reflective journal entries and analysed using thematic analysis. In addition, students writing artifacts provided contextual evidence of students’ responses to feedback. The findings suggest that students’ affective engagement developed progressively from initial confusion, frustration, and uncertainty toward motivation, confidence, and greater autonomy. The results also indicate that emotional engagement fluctuated in response to cognitive and rhetorical demands. Teacher mediation played an important role in clarifying feedback and supporting students’ emotional adaptation. These findings highlight the dynamic nature of affective engagement in AI-mediated feedback and suggest the importance of pedagogical scaffolding when integrating AI tools in EFL writing instruction.

With the recent advent of GenAI, preparing L2 students to write an effective Curriculum Vitae (CV) or resume, in English, a typical genre in English for Professional Purposes (EPP) courses, presents unique challenges and opportunities. GenAI tools can create professional-looking CVs, yet L2 students may struggle to analyze them or recognize issues of using AI-generated CVs when they apply for jobs. To develop L2 writers’ CV-writing skills in the GenAI age, an intervention was conducted within an EPP course at a Hong Kong tertiary institution. The intervention lasted six hours and involved various activities designed to demonstrate the opportunities and limitations of using GenAI tools to assist with CV writing. Data were collected from 24 students through pre- and post-intervention questionnaires. Findings indicate a slight improvement in students' confidence and perception of the importance of CV writing, though challenges remain in differentiating AI-generated content and in addressing concerns about accuracy and policy restrictions. The study underscores the need for EPP courses to integrate AI literacy, ensuring students can effectively leverage AI while maintaining their unique voice in professional CVs.

Reflective Practice (RP) has been widely recognized as one of the crucial elements of teacher education and professional development. However, limited studies have explored how Indonesian Pre-Service English Language Teachers (PELTs) conceptualize and implement RP in language education. This study aims to examine Indonesian PELTs’ conceptualization of RP and their justification of their practices in a higher education context. Informed by a qualitative case study design, the study involved 50 fifth-semester PELTs from an English Education Department in West Java, Indonesia. Data were collected through reflective writings and semi-structured interviews, which were then analyzed thematically. The findings show that participants conceptualized RP in two interconnected ways: RP as evidence-based interrogation and RP as introspective process. They employed different ways of undertaking RP, including journals, social media, videos, peer discussions, AI applications, and creative writing, to evaluate their experiences and support personal and professional growth. This study highlights the significance of RP in cultivating reflective practitioners and supports its integration into language teacher education programs.

This small scale narrative inquiry examined how five Indonesian EFL pre-service teachers interpreted challenges experienced during school-based teaching practicum in relation to their emerging professional identities. The participants were purposively selected English Education students at a private university who had completed teaching practicum in the preceding semester. Data were generated through individual, offline, face to face semi-structured interviews lasting approximately 45-60 minutes. The data were analyzed using thematic analysis to identify patterns in participants’ accounts of classroom challenges, interpretations, responses, and projected teacher identities. The accounts centred on classroom management, students’ speaking confidence, instructional engagement, learner diversity, and limited teaching experience. Across these experiences, participants repositioned themselves as relational authorities, supportive facilitators, responsive and adaptive teachers, reflective planners, and committed continuous learners. The findings represent retrospective and provisional identity construction rather than verified long term transformation. The study contributes an incident to identity meaning-making sequence in which consisting of challenge, appraisal, pedagogical response, and projected professional identity that can support structured reflection in EFL teacher education.