| 10 | Are AI Tools Effective in Reducing Cognitive Load or Boosting Writing Self-Efficacy for L2 Postgraduates? | Ziyi Zhang, Chunyan Wang, Yuhong Shang | 2025 | 327 | Postgraduate | Academic Writing | Composing research proposals in their disciplines | Long-term | | | | | Quillbot, Kimi, Xiao Lvjing, ChatGPT | AI tools significantly increased writing self-efficacy for postgraduates with lower L2 literacy without affecting cognitive load, but did not show similar benefits for those with higher L2 literacy. | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | | |
| 9 | Empowering medical students with Al writing co-pilots: design and validation of Al self-assessment toolkit | Laleh Khojasteh, Reza Kafipour, Farhad Pakdel, Jayakaran Mukundan | 2025 | 22 | Undergraduate | Academic Writing | Writing short papers (argumentative, cause-and-effect, problem-solution, analytical essays) on medical scenarios and case studies. | Long-term | | | | | ChatGPT | A comprehensive self-assessment toolkit designed for use with AI writing co-pilots was established as a valid and reliable instrument for supporting medical students' writing improvement. | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | | |
| 8 | Students-Generative Al interaction patterns and its impact on academic writing | Jinhee Kim, Sang-Soog Lee, Rita Detrick, Jialin Wang, Na Li | 2025 | 36 | Postgraduate | Academic Writing | Writing an academic essay on topics related to coursework | Short-term | | | | | GPT-4 | Students with high AI literacy showed a collaborative interaction pattern with GenAI, leading to significantly better academic writing performance in content, structure, and expression compared to students with low AI literacy. | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | | |
| 7 | The impact of Al-enhanced natural language processing tools on writing proficiency: an analysis of language precision, content summarization, and creative writing facilitation | Dan Zhao | 2025 | 295 | | English as a Foreign Language (EFL) education | Tasks focused on language precision, content summarization, and creative writing. | Long-term | | | | | Not specified | The experimental group that used AI-enhanced tools for Language Precision Enhancement showed significantly superior improvements in writing proficiency compared to other AI-supported groups and the control group. | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | | |
| 6 | Enhancing academic writing skills and motivation: assessing the efficacy of ChatGPT in Al-assisted language learning for EFL students | Cuiping Song, Yanping Song | 2023 | 50 | Undergraduate | Academic Writing | Completing IELTS academic writing tasks 1 and 2 | Long-term | | | | | ChatGPT | AI-assisted instruction via ChatGPT led to significant improvements in both academic writing skills and writing motivation compared to traditional instruction. | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | | |
| 5 | Al vs. teacher feedback on EFL argumentative writing, a quantitative study | Areen Alnemrat, Hesham Aldamen, Mohamad Almashour, Mutasim Al-Deaibes, Rami AlSharefeen | 2025 | 60 | Undergraduate | Argumentative Writing | Writing an argumentative essay on a given topic | Short-term | | | | | ChatGPT (GPT-4) | Both AI-generated and teacher-generated feedback led to significant improvements in writing performance, with no statistically significant difference found between the two feedback groups. | | — | — | | — | | | | — | — | | | — | | | | | | | |
| 4 | Enhancing EFL Writing Revision Practices: The Impact of AI- and Teacher-Generated Feedback and Their Sequences | Thi Thanh Thao Tran | 2025 | 14 | Undergraduate | Academic Writing | Writing opinion and cause-and-effect essays | Long-term | Quantitative | Essays, feedback records, and revision history are collected directly from users. | ANOVA, t-test, correlation | AI provides feedback based on prepared prompts that instruct it to act in specific roles, such as a 'writing coach' or 'grammar expert,' to deliver targeted feedback on areas like organization, vocabulary, and grammar. | Gemini | The integration of AI- and teacher-generated feedback yielded the highest frequency of revisions, with AI addressing surface-level issues and teachers focusing on higher-order concerns. | | | | | — | — | — | — | — | — | | | — | | | | | — | | AI provide feedback based on prepared prompts. Feedbacks from AI are more focused on low level skills, while feedback from teacher are more focused on high level skills. |
| 3 | Balancing syntactic complexity and clarity: the role of AI in enhancing academic writing proficiency | Akbar Bahari | 2025 | 466 | Postgraduate | Academic Writing | Writing a descriptive report and an argumentative essay (IELTS Academic Writing Test) | na | Mix | Pretest and posttest, experiment, with survey | ANOVA, post-hoc Tukey | AI serves as a co-writer or assessment partner, providing automated, instant, and personalized feedback to guide the revision process. | Grammarly, ProWritingAid, Hemingway Editor, Quillbot, Writefull, Turnitin Revision Assistant | AI-enhanced educational tools, particularly the Turnitin Revision Assistant, significantly improved academic writing proficiency among EFL postgraduate students compared to traditional methods. | | — | — | | — | | | — | | — | | | — | | | | | | Learners' perception on benefits and chalenges | |
| 2 | Shaping Human-Al Collaboration: Varied Scaffolding Levels in Co-writing with Language Models | Paramveer S. Dhillon, Somayeh Molaei, Jiaqi Li, Maximilian Golub, Shaochun Zheng, Lionel P. Robert | 2024 | 131 | Postgraduate | Argumentative Writing | Writing argumentative essays | Short-term | Quantitative | Essay writing, log data observation, survey | Generalized Linear Models (GLM) | AI provides high scaffolding (paragraph suggestion) and low scaffolding (sentence suggestion) | GPT-3 DaVinci (text-davinci-002) | High-level AI scaffolding (paragraph suggestions) significantly improved writing quality and productivity, especially for non-regular writers, whereas low-level scaffolding (sentence suggestions) did not. | — | — | — | — | — | — | — | — | — | — | | | — | | | | — | — | User satisfaction, content ownership, cognitive load | AI is used as an assistance tool to improve writing quality, rather than as a training tool to develop a writer's skill |
| 1 | Using LLMs to bring evidence-based feedback into the classroom: AI-generated feedback increases secondary students’ text revision, motivation, and positive emotions | Jennifer Meyer, Thorben Jansen, Ronja Schiller, Lucas W. Liebenow, Marlene Steinbach, Andrea Horbach, Johanna Fleckenstein | 2024 | 459 | High School | English as a foreign language | Writing and revising argumentative essays | Short-term | Quantitative | Essay writing | Regression | Providing personalised feedback | GPT-3.5-turbo | LLM-generated feedback increased students' revision performance, task motivation, and positive emotions compared to a control group revising without feedback. | — | — | — | — | — | | — | | — | — | | | — | | — | — | | | | AI provides general feedback, not targeted skill improvement. |