An Analysis of Students’ Writing Ability in AI-Assisted Writing
Ulfa HaeraAndi Anto PatakVivit Rosmayanti
Aug 2026· International Journal of Language, Education, and Literature· 0 citations
TL;DR
It is suggested that AI integration alone does not guarantee improved writing outcomes without corresponding foundational skill development, as students lack the foundational knowledge necessary to utilize AI tools effectively.
Abstract
The integration of Artificial Intelligence (AI) tools in writing instruction has garnered widespread attention; however, empirical evidence on the actual quality of student writing produced with AI assistance remains limited. This study aims to examine students' writing ability in AI-assisted writing products. This research employed a qualitative case study design. Data were collected through document analysis of narrative texts from five Grade XII students at SMAN 12 Sinjai, Indonesia. The analysis was conducted using Brown’s (2018) five-component framework, assessing content, organization, grammar, mechanics, and style. The findings revealed that students’ writing ability in AI-assisted contexts remains predominantly low. Only two students demonstrated Moderate ability, while the rest showed Poor ability across multiple components. Specific deficiencies included extreme repetition in content, lack of logical flow in organization, pervasive past tense errors in grammar, missing punctuation in mechanics, and severely limited vocabulary in style. The study concludes that a fundamental disconnect exists between AI use and writing quality, as students lack the foundational knowledge necessary to utilize AI tools effectively. These findings suggest that AI integration alone does not guarantee improved writing outcomes without corresponding foundational skill development.
AI tools are frequently presented as a means to improve EFL student writing; however, the extent to which AI genuinely enhances writing skills—rather than merely composing polished final products—remains an open question. This longitudinal experiment involved 93 first-year Japanese university students completing English writing tasks over one semester. Students were divided into two groups: the self-editing group (control), and the ChatGPT-assisted group (variable). Writing samples were assessed on length, relevance, linguistic diversity, and accuracy. Analysis revealed that ChatGPT-assisted editing does little to improve students’ actual writing skills. While the final drafts of AI-assisted essays were relatively error-free, students’ pre-edited texts showed no statistically significant differences in overall writing quality when compared to the control group. This suggests that AI-assisted editing polishes the final composition rather than developing skills, calling into question the prevailing narrative that AI tools enhance writing ability.
AIツールは英語を外国語として学ぶ学生のライティング能力向上手段として頻繁に紹介されるが、AIが単なる完成度の高い最終成果物の作成ではなく、真にライティングスキルを向上させる程度については明確ではない。本研究では、日本の大学1年生93名を対象に、1学期にわたり英作文の課題を実施した。学生は自己編集グループ（対照群）とChatGPTによる支援グループ（実験群）の2群に分けられた。作文は、長さ、内容の関連性、語彙の多様性、正確性に基づいて評価された。分析の結果、ChatGPTによる編集支援は学生の書く力の向上にはほとんど寄与しないことが明らかになった。AI支援による最終稿は比較的誤りが少なかったものの、編集前の文章は対照群とは統計的に有意な差が認められなかった。このことから、AIは作文スキルを育成するよりも、最終的な文章を磨き上げるだけのものであり、AIツールがライティング能力を高めるという一般的な見解に疑問を投げかける結果となった。
James Thomas, Melissa Huntley· JALT Postconference Publicat...· 0 citations
The growing use of artificial intelligence in language learning has encouraged the development of digital tools that support academic writing, and Grammarly is one of the most widely used applications. This study explored EFL students’ perceptions of Grammarly as an AI-assisted writing tool in academic writing. A qualitative case study was conducted with seventeen fourth-semester students from the English Language Education Study Program at Universitas PGRI Ronggolawe Tuban during the 2025/2026 academic year. Data were collected through classroom observation, questionnaires, and semi-structured interviews, then analyzed using the interactive model of Miles, Huberman, and Saldaña. The findings show that students generally perceived Grammarly positively because it helped them identify grammatical errors, improve sentence structure, and revise writing more efficiently. Grammarly also increased students’ confidence and supported more independent learning through immediate feedback. However, participants also reported several challenges, including inaccurate suggestions, dependence on internet connectivity, and the need to evaluate AI-generated feedback critically. The study concludes that Grammarly is a useful complementary tool that should support, not replace, instructors’ guidance. Integrating Grammarly into EFL writing instruction with appropriate pedagogical support may help improve students’ academic writing skills and promote learner autonomy.
The integration of artificial intelligence (AI) technology into education has become increasingly important in supporting the writing proficiency of non-English major students in English for Specific Purposes (ESP) courses. This study investigates the effectiveness of ChatGPT in improving the clarity, coherence, and grammatical accuracy of students’ writing drafts. Employing a quasi-experimental design, the study examined differences in learning outcomes between a ChatGPT-assisted class (experimental group) and a traditionally taught class (control group). Data were collected through writing draft assessments, participant surveys, and interviews. The findings indicate that the experimental group demonstrated a significant improvement in writing performance, particularly in organizing and presenting ideas clearly and coherently. Furthermore, the personalized guidance provided by ChatGPT contributed to students’ learning autonomy and self-confidence. These findings highlight the potential of ChatGPT as a supportive tool for language learning while emphasizing the continued role of lecturers in validating recommendations generated by AI systems. Future research is encouraged to explore AI-assisted writing in more specialized disciplinary contexts and to investigate its long-term effects on students’ learning retention and writing development.
Rachmat Ari Wibowo· English Language and Educati...· 0 citations
Although generative artificial intelligence (GAI) increasingly impacts academic writing and integrity in higher education, empirical evidence on assessment strategies to reduce students' reliance on AI-generated texts remains limited. This study aimed to examine the effectiveness of the Contextual Writing Assessment in reducing AI-generated text dependence and improving writing performance, writing originality, academic integrity, and learner agency. A quasi-experimental mixed-methods design with an embedded explanatory approach was employed in an English course at the Faculty of Engineering at the University of
Muhammadiyah Tangerang. The participants comprised 150 second-semester students from six classes and four lecturers. Three classes were assigned to the experimental group and received Contextual Writing Assessment, while three classes served as the control group and received the conventional writing assessment. Data were collected through writing tests, AI-dependence questionnaires, academic integrity and learner agency questionnaires, originality scores, AI-use logs, reflective journals, draft analyses, observations, and interviews. The results demonstrated the effectiveness of the intervention, as the experimental group outperformed the control group in reducing AI-generated text dependence and exhibited substantial improvements in writing performance, originality, academic integrity, and learner agency. Qualitative evidence indicated more transparent AI use, meaningful revisions, stronger reflections, and reduced copy-paste behavior. The study concludes that Contextual Writing Assessment offers an ethical, process-based, and AI-resilient approach to writing assessment.
Y. Daniarti, Yanuardi· International Journal of Res...· 0 citations
The growing presence of artificial intelligence (AI) in higher education has changed the way students approach academic writing. While AI-powered tools offer practical support in generating ideas, organizing texts, and refining language, their increasing use has also raised concerns about the originality and authenticity of students’ written work. This study aims to examine how AI influences students’ authentic academic writing skills and to identify the patterns of dependence that emerge during the writing process. A convergent mixed-methods design was employed by integrating quantitative data from questionnaires completed by 71 third-semester English education students in Makassar, Indonesia, with qualitative evidence drawn from fifteen empirical and conceptual studies published between 2024 and 2025. Descriptive statistics were used to analyze the survey data, whereas thematic synthesis was applied to the literature findings. The results indicate that students rely on AI to varying degrees across different stages of academic writing. Five interrelated forms of dependence were identified, namely dependence on idea generation, language and text organization, revision and editing, cognitive and metacognitive processes, and writing autonomy. However, the findings suggest that AI is not inherently responsible for weakening students’ writing abilities. Instead, the erosion of authentic writing tends to occur when technological assistance replaces the reflective, critical, and self-regulatory processes that are central to academic writing. These findings underscore the importance of developing educational practices that encourage students to use AI responsibly while maintaining intellectual ownership and academic integrity.
Muhammad Yahrif, Suharti Sirajuddin, Muhamad Khaedar· Al Qodiri : Jurnal Pendidika...· 0 citations
It is argued that the thoughtful integration of AI feedback with teacher feedback, grounded in pedagogical principles and human judgment, can significantly enhance L2 writing instruction.
O. Jamoom, Saaid Ali Omar· Comprehensive Journal of Sci...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
