Aug 2026· Language, Technology, and Social Media· Vol 4, pp. 448-459· 0 citations· 23 references
TL;DR
The results imply that the use of AI‑mediated feedback, when systematically integrated with teacher support, can be an effective and scalable way to advance EFL learners’ writing skills in similar contexts.
Abstract
The present paper focused on examining the application of GPT‑5.5 as an automated corrective feedback tool for improving the writing skills of 60 Omani pre‑intermediate EFL students enrolled in a foundation program. A quasi‑experimental pretest–posttest design was used, and the participants were divided into two groups, namely an experimental group and a control group, each receiving instruction on advantages‑and‑disadvantages essays. While both groups followed the same syllabus and classroom procedures, only the experimental group utilized GPT‑5.5 AI tool for iterative feedback and revision, supplemented by teacher explanation and clarification. The analytic scoring of the pretest and posttest compositions demonstrated that both groups made progress, yet the experimental group scored substantially better on the posttest, with a large effect size. The results imply that the use of AI‑mediated feedback, when systematically integrated with teacher support, can be an effective and scalable way to advance EFL learners’ writing skills in similar contexts.
The rapid advancement of artificial intelligence (AI) has created new opportunities for providing feedback in writing instruction. This study examined the effect of ChatGPT-generated feedback on students’ narrative writing performance and compared its effectiveness with teacher feedback. A quasi-experimental design was conducted with two groups of Grade XI students at SMAN 65 Jakarta. The experimental group received feedback from ChatGPT, while the control group received teacher feedback. Students completed pre-tests and post-tests consisting of narrative writing tasks assessed using Jacobs’ ESL Composition Profile Rubric. Data were analyzed using the Wilcoxon Signed-Rank Test and Mann–Whitney U Test due to non-normal distribution. The results showed that ChatGPT-generated feedback significantly improved students’ writing performance (p < 0.05). However, no significant difference was found between the improvements achieved by the ChatGPT and teacher-feedback groups (p = 0.285). These findings indicate that ChatGPT can be a useful supplementary tool in writing instruction, while teacher feedback remains essential for providing personalized guidance.
Priskila Go, H. B. Sihombing· JURNAL ILMIAH PENDIDIKAN KEB...· 0 citations
Introduction:
Generative artificial intelligence tools such as ChatGPT are increasingly used to provide writing feedback, yet their effectiveness relative to teacher feedback and across distinct writing dimensions remains underexplored, particularly in languages other than English. This study compares the effects of ChatGPT-assisted feedback and teacher feedback on the writing achievement of B2-level learners of Turkish as a foreign language.
Methods:
An explanatory sequential mixed-methods design was adopted. The study group comprised 18 students enrolled at the Turkish Teaching Center (TÖMER) of Recep Tayyip Erdoğan University. Over an eight-week intervention, the experimental group (
n
= 9) received feedback through ChatGPT and the control group (
n
= 9) received feedback from the teacher, using an identical seven-category feedback guideline. Quantitative data were gathered through a pre-testpost-test control-group quasi-experimental design and analyzed with t-tests; qualitative data were obtained through a written process-evaluation form and analyzed thematically.
Results:
Writing achievement increased significantly in both groups, yet no significant difference emerged between the groups in their post-test scores. At the sub-dimension level, ChatGPT proved more effective in structural areas such as grammar, vocabulary, and spelling and punctuation, whereas the teacher held a clear advantage in inter-sentential planning and organization. The qualitative data indicated that ChatGPT fostered learner autonomy by lowering the anxiety of being evaluated, while the teacher was more capable of providing motivation and building contextual depth.
Discussion:
The two feedback sources are statistically equivalent overall but display a division of labor across writing dimensions, complementing rather than replacing each other. The study argues that artificial intelligence should be integrated into a hybrid model that enriches the feedback process rather than replacing the teacher.
Generative AI tools such as ChatGPT are increasingly integrated into the teaching of English writing, and previous research confirms their ability to provide personalized feedback to correct grammatical and lexical errors and improve writing quality. However, few empirical studies have examined how ChatGPT shapes students’ argumentative writing self-efficacy. To fill this research gap, a circular framework of writing-feedback-revision was used for three weeks in this study. Forty-five Chinese English majors were divided into a high-level group (n = 22) and a low-level group (n = 23) according to their vocabulary test scores, and then randomly assigned to the ChatGPT feedback group and the teacher feedback group. The research instruments included an argumentative writing assignment, a writing self-efficacy questionnaire, and a vocabulary assessment. This paper examines the effects of feedback at different levels of writing on writing competence and self-efficacy. The results reveal three key findings. First, learners who received teacher feedback (mean rank = 27.39) made greater writing progress than those who used ChatGPT (mean rank = 19.19), with low-proficiency students showing the weakest improvement with AI feedback (mean rank = 13.63). Second, ChatGPT significantly improved the writing self-efficacy of low-level learners, but there was no corresponding gain in their writing performance. Third, teacher feedback had a consistent positive effect on both writing outcomes and self-efficacy. The evidence suggests that ChatGPT cannot completely replace writing instructors, especially for low-proficiency learners, as over-reliance on AI may inhibit writing growth. Nevertheless, ChatGPT can still be used as a viable aid to correct errors and refine expressions to develop writing confidence, and advanced learners need teacher-led training to develop critical and creative thinking. This study provides empirical evidence for AI-enhanced writing instruction and provides differentiated instructional designs for learners with different proficiency levels.
Chenlong Zhang· Journal of Contemporary Educ...· 0 citations
Grounded in the reinforcement principle of behaviorism and Keller's ARCS-V (Attention, Relevance, Confidence, Satisfaction, and Volition) motivational model, this study explored the effects of Grammarly as an Automated Written Corrective Feedback (AWCF) gadget, combined with teacher feedback and the ARCS-V motivational model, on enhancing the writing performance and self-efficacy of Saudi EFL university students. The study implemented a quasi-experimental design with equivalent groups. The participants included 84 first-year Saudi female EFL students at King Khalid University, divided into an experimental group (43 students) that was exposed to both Grammarly and teacher feedback via the ARCS-V model, and a control group (41 students) that received only teacher feedback. To collect data, pre- and post-writing tests evaluating six writing subskills and a self-efficacy scale covering four domains were utilized. Quantitative data were analyzed using independent and paired-samples t-tests. Results demonstrated that the experimental group significantly outperformed the control group in both writing skills and all four self-efficacy domains, reflecting the positive effects of integrating AWCF with teacher feedback via the strategic application of the ARCS-V model. It has been noted that Grammarly is effective at providing feedback on grammar, punctuation, spelling, and vocabulary. However, in some contexts, it could not provide relevant explanations on overall coherence and organization. Therefore, the teacher's feedback is crucial. This study recommends integrating AI tools, human feedback, and motivation models for writing instruction.
Eman Mahmoud Ibrahim Alian· Technology in Language Teach...· 1 citation
This study examined ChatGPT-4.0 as a feedback tool in Chinese as a Second Language writing instruction, focusing on writing performance and writing self-efficacy. An eight-week quasi-experimental intervention was conducted with 53 intermediate-to-advanced CSL undergraduates in East China. The experimental group received ChatGPT-assisted feedback during writing, while the control group relied on traditional teacher support. A mixed-methods design was adopted, combining pre- and post-test writing assessments, questionnaire measures of self-efficacy, and follow-up interviews. Findings indicated that students using ChatGPT showed significantly higher post-test writing scores and greater gains in writing self-efficacy, particularly in ideation, conventions, and self-regulation. Regression analyses further suggested that ChatGPT use and post-test self-efficacy were both associated with improved writing outcomes. Interview data revealed that learners valued the immediacy, availability, and low-pressure nature of AI feedback, which appeared to support correction, idea development, self-monitoring, and confidence building, while also raising concerns about overreliance and formulaic expression. Overall, the results suggest that ChatGPT may serve as a useful feedback resource in CSL writing when implemented with structured guidance, teacher mediation, and attention to learner uptake of AI-generated feedback.
Guan Huang, Huijing Chen· Language Learning & Tech...· 0 citations
This study investigated the impact of a metacognitive instruction (MI) intervention on EFL writers’ metacognition, feedback uptake, and revision performance in a Generative AI (GenAI)‐assisted feedback environment. The MI intervention integrated prompt engineering into EFL writing pedagogy following a four‐step MI sequence. Participants were 141 EFL learners recruited from a Chinese university, with 71 allocated to the experimental group (EG), which received the MI intervention, and 70 assigned to the control group (CG), which engaged with GenAI feedback without additional scaffolding. Following a pretest–posttest quasi‐experimental design, data were collected from multiple sources, including questionnaires, reflective journals, think‐aloud protocols, and draft and revised writing products. Results showed that the MI intervention significantly improved students’ strategy knowledge and all dimensions of metacognitive strategy use. EG participants demonstrated more elaborate uptake of GenAI feedback on global writing issues. Moreover, the EG achieved greater feedback‐informed revision gains in content and organization, with no significant differences observed in grammar, vocabulary, and mechanics. This study has important pedagogical implications for designing and implementing situated metacognitive scaffolding to support EFL writers’ critical and effective engagement with GenAI‐assisted feedback.
Daisy Yan· International Journal of App...· 0 citations