Aug 2026· Revista de educación· Vol 14· 0 citations· 37 references
Abstract
ChatGPT is increasingly being used to support English as a Foreign Language (EFL) academic writing in higher education, yet evidence on its role in Asian university instruction remains fragmented. This scoping review maps empirical studies on ChatGPT in EFL academic writing instruction in Asian higher education, focusing on pedagogical uses, reported outcomes and ethical concerns. Methods Following Arksey and O'Malley's framework and PRISMA‐ScR guidance, the review searched five databases for peer‐reviewed English‐language studies published between 2020 and 2025. Two reviewers independently screened and charted the studies, and findings were synthesised thematically through the PAGER framework. Twenty‐seven studies were included. ChatGPT was primarily used for pre‐writing, revision and feedback. Reported benefits included stronger writing quality, greater learner autonomy, improved motivation and reduced anxiety. Common concerns included overreliance, inaccurate or culturally biased feedback, uncertainty surrounding academic integrity and unequal access to AI technologies. ChatGPT can be a useful scaffold for EFL academic writing, but its value depends on guided integration, critical AI literacy and context‐sensitive policy. The findings support staged AI‐assisted writing activities for EFL instructors, curriculum and assessment designs that promote critical AI literacy and responsible AI use, and institutional policies that provide clear guidance, educator professional development and equitable access to AI technologies. Further longitudinal, comparative and theory‐informed research is needed to evaluate the long‐term effectiveness of these approaches.
Rationale for this study:
Empirical evidence on ChatGPT for EFL academic writing in Asian higher education remains fragmented, limiting understanding of effective pedagogical practices and associated challenges.
Why the new findings matter:
This review identifies how ChatGPT is currently used, its reported benefits and the conditions needed for responsible and effective implementation.
Implications for practitioners, institutional leaders and researchers:
The findings support integrating ChatGPT as a scaffold rather than a replacement for writing instruction. EFL instructors should adopt structured AI‐supported writing activities that promote critical AI literacy and independent writing. Higher education institutions should develop clear AI policies, redesign assessment to emphasise writing processes and responsible AI use and provide professional development for educators. Researchers should prioritise longitudinal, comparative and theory‐informed studies to strengthen the evidence base for AI‐supported academic writing.
ChatGPT is one of the large language models among artificial intelligence tools being integrated to support language and academic writing in higher education. This research was conducted with clear purposes: (1) to examine the roles of ChatGPT in enhancing the writing accuracy of undergraduate EFL learners; (2) to investigate the challenges and limitations students face when using ChatGPT for writing improvement; (3) to identify whether there are differences between male and female undergraduate students in their educational comprehension and perceptions toward using ChatGPT in English writing skills. The study was conducted quantitatively, with 226 undergraduate students participating via a Google Forms survey administered through purposive sampling. Roles, Challenges, and perceptions were measured with a 5-point Likert-scale instrument and presented in both inferential and descriptive statistics. Results revealed that EFL students heavily use ChatGPT for writing assistance, reporting for grammar error identification and correction (M = 3.78, SD = 0.96) and sentence clarity enhancement (M = 3.72, SD = 0.83). Regarding general perspectives, students valued ChatGPT for fast global information retrieval (M = 3.69, SD = 0.89), while confidence in relying on its output was lower (M = 3.35, SD = 0.89). Major challenges were found in proper citation and academic attribution (M = 3.74, SD = 0.81), and overly complex or technical outputs (M = 3.72, SD = 0.77). A statistically significant gender difference in confidence from inferential analysis with an independent samples t-test was indicated (t (224) = 2.88, p = .004), with male students rated on self-assessed confidence (M = 3.75) while female students rated (M = 3.51). However, there is no statistically significant difference across academic year cohorts (p > .05) analyzed with One-Way ANOVA. ChatGPT is valued as a writing tool for assistance with grammar correction and draft enhancement among undergraduate students, while challenges related to verifying information, practicing citations, and varying self-confidence between genders were found as constraints to ChatGPT adoption. Practitioners should consider digital literacy training to master the effectiveness of AI writing accuracy.
ABSTRAK
ChatGPT adalah salah satu model bahasa besar di antara alat kecerdasan buatan (artificial intelligence) yang diintegrasikan untuk mendukung pemelajaran bahasa dan penulisan akademik di perguruan tinggi. Penelitian ini dilakukan dengan tujuan yang jelas: (1) untuk menguji peran ChatGPT dalam meningkatkan akurasi penulisan pembelajar EFL (English as a Foreign Language) tingkat sarjana; (2) untuk menyelidiki tantangan dan keterbatasan yang dihadapi mahasiswa saat menggunakan ChatGPT untuk peningkatan kualitas penulisan; (3) untuk mengidentifikasi apakah terdapat perbedaan antara mahasiswa laki-laki dan perempuan dalam pemahaman edukatif serta persepsi mereka terhadap penggunaan ChatGPT dalam keterampilan menulis bahasa Inggris. Penelitian ini dilakukan secara kuantitatif, dengan 226 mahasiswa tingkat sarjana berpartisipasi melalui survei Google Forms yang disebarkan menggunakan teknik purposive sampling. Peran, tantangan, dan persepsi diukur menggunakan instrumen skala Likert 5 poin dan disajikan dalam statistik inferensial maupun deskriptif. Hasil penelitian mengungkapkan bahwa mahasiswa EFL sangat memanfaatkan ChatGPT untuk bantuan penulisan, terutama dalam mengidentifikasi dan mengoreksi kesalahan tata bahasa (M = 3.78, SD = 0.96) serta meningkatkan kejelasan kalimat (M = 3.72, SD = 0.83). Terkait pandangan umum, mahasiswa mengapresiasi ChatGPT untuk penarikan informasi global yang cepat (M = 3.69, SD = 0.89), sedangkan tingkat keyakinan untuk bergantung pada hasil (output) AI tercatat lebih rendah (M = 3.35, SD = 0.89). Tantangan utama ditemukan pada sitasi dan atribusi akademik yang tepat (M = 3.74, SD = 0.81), serta luaran yang terlalu kompleks atau teknis (M = 3.72, SD = 0.77). Analisis inferensial dengan independent samples t-test menunjukkan adanya perbedaan keyakinan yang signifikan secara statistik berdasarkan gender (t(224) = 2.88, p = .004), di mana mahasiswa laki-laki memberikan penilaian lebih tinggi pada keyakinan diri (M = 3.75) dibandingkan mahasiswa perempuan (M = 3.51). Namun, tidak terdapat perbedaan yang signifikan secara statistik di antara kohort angkatan akademik (p > .05) berdasarkan analisis One-Way ANOVA. ChatGPT dinilai sebagai alat bantu penulisan yang bermanfaat untuk koreksi tata bahasa dan penyempurnaan draf di kalangan mahasiswa, meskipun tantangan terkait verifikasi informasi, praktik sitasi, serta variasi rasa percaya diri antargender menjadi kendala dalam adopsi ChatGPT. Para praktisi sebaiknya mempertimbangkan pelatihan literasi digital untuk mengoptimalkan efektivitas akurasi penulisan berbasis AI.
A qualitative thematic analysis was conducted with 25 students from a Malaysian institution to gain a deeper understanding of their motivation for using ChatGPT in English language learning. The study employed Self-Determination Theory (SDT) and the results demonstrated that ChatGPT has the potential to contribute positively to aspects related to English language competence, social connections, and independent learning processes. Therefore, ChatGPT emerges as a helpful resource for enhancing grammar, writing, and conversational tasks, allowing educators to prioritize higher order thinking abilities and intricate areas of language training. The results suggest that ChatGPT supports learners' autonomy, relatedness, and competence, which are key components of Self-Determination Theory (SDT). Nevertheless, the study also highlighted certain obstacles, such as ChatGPT's tendency to produce erroneous or nonsensical information and its restricted capacity to ensure the accuracy of content in language learning situations. These findings emphasize the need for focused efforts to tackle these difficulties. Educators must acknowledge the limitations of the tool and complement ChatGPT interactions with human interactions to provide a comprehensive and efficient language learning experience.
Nagaletchimee Annamalai, Mohamed Nasor, Arathai Din Eak et al.· Frontiers in Psychology· 0 citations
GenAI is most educationally defensible when integrated as a guided formative-feedback resource rather than as a substitute writer or replacement for teacher expertise, as well as for assignment design, AI-use disclosure, feedback literacy, prompt literacy, and process-based assessment.
Bùi Thị Hương Thảo· International journal of soc...· 0 citations
This study addresses the limited evidence on the impact of ChatGPT integration on Arabic language learning among Generation Z students, particularly across linguistic, cognitive, motivational, and ethical dimensions. The study aims to synthesize recent evidence on ChatGPT-supported Arabic language learning and identify emerging patterns, challenges, and research gaps. A systematic literature review was employed using a PRISMA-informed approach to identify, screen, evaluate, and synthesize relevant studies published primarily between 2021 and 2026. The literature was analyzed thematically across linguistic development, motivation and engagement, learner autonomy, self-regulated learning, higher-order thinking, AI literacy, personalization, ethical use, and potential dependency. The findings indicate that ChatGPT can support Arabic reading, writing, speaking, grammar, vocabulary, translation, language practice, personalized feedback, motivation, and self-directed learning. Cognitive benefits were also identified when ChatGPT was integrated through structured guidance and active learner engagement. However, concerns regarding academic integrity, dependency, information accuracy, privacy, and inadequate human validation remain. The review concludes that ChatGPT is most effective as a guided human-AI learning support system rather than a replacement for teachers or learners. The study contributes theoretically by integrating technological, cognitive, and learner-centered perspectives and practically by highlighting the need for AI literacy, responsible prompting, teacher guidance, and human verification. Future research should examine Generation Z learners through longitudinal, experimental, and cross-skill studies in Arabic language education.
Nurfaizah, M. Hasyim, Yanti Puspita et al.· Indonesian Journal of Educat...· 0 citations
Although online cooperative learning has been widely examined in English as a Foreign Language (EFL) education, it is often conceptually blurred with less structured forms of collaborative learning. This review synthesised empirical articles on online cooperative learning in EFL contexts, focusing on study characteristics, reported outcomes, the operationalisation of Johnson and Johnson’s five elements, and implementation challenges. PRISMA 2020, Eric, Web of Science, Scopus, and ProQuest were searched for empirical English-language journal articles published up to June 2026. Twenty-five studies met the eligibility criteria and were appraised using the Mixed Methods Appraisal Tool (MMAT). Mixed-methods designs were the most common (56%, n = 14), university settings dominated the studies (80%, n = 20), and writing was the leading focus (40%, n = 10). Of the 17 studies using inferential tests of language-related performance, 16 reported statistically significant positive gains or associations, although the strength of evidence varied. Positive interdependence was explicitly reported in 20 studies and implied in 5. In contrast, small-group skills and group processing were explicitly reported in only[1] 3 and 7 studies, respectively. The main implementation challenges involved weak technological–pedagogical alignment, insufficient instructional scaffolding, and uneven participation and accountability. The findings suggest that online cooperative learning should be conceptualized not according to named cooperative structures alone, but according to how the online cooperative EFL learning environment, instructional configuration, and the five core cooperative learning elements are integrated. Future research should design clearly structured online cooperative EFL courses across more diverse educational settings, language areas, and learner groups, particularly for learners with lower language proficiency or limited digital literacy.
Ai-Jia Zhang, M. Yunus, Nur Syafiqah Yaccoba· International Journal of Lea...· 0 citations
This study explored the effectiveness of automated written corrective feedback (AWCF) in learning English as a foreign language among undergraduate students in Saudi tertiary education, focusing on how learners responded to and benefited from it. Using a mixed-methods approach, data were collected from students’ written drafts before and after using AWCF and from a structured questionnaire capturing their perceptions and engagement. The findings showed that AWCF helped students identify and revise grammatical and spelling errors, improving surface-level linguistic accuracy. Students also demonstrated high engagement, critically assessing feedback, particularly for local proper nouns and context-specific language use. Although students largely trusted AWCF, they did not accept all suggestions, demonstrating self-monitoring and writing autonomy. The findings underscore the importance of teacher involvement in addressing higher-order concerns. The study recommends integrating AWCF with teacher feedback to maximize learning outcomes. Future research should use larger samples to test whether similar findings could be discovered. This study also did not fully examine the complementary role of instructors in AWCF use. It is also recommended that future research examine the effect of combining AWCF with teacher feedback on writing development.
Unknown authors· Arab World English Journal· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
