Human-AI collaborative writing strategy significantly improves Indonesian EFL students' argumentative writing skills and critical thinking abilities while fostering positive learning experiences.
Objective: The main goal of this study was to examine the effects of human-AI collaborative writing strategy on Indonesian EFL students' argumentative writing skills and explore their perceptions and experiences regarding the strategy implementation. The study specifically aimed to investigate whether this collaborative approach could effectively enhance students' ability to construct well-structured argument traits and develop critical thinking skills in argumentative essay writing.
Methods: The research employed a pre-experimental one-group pretest-posttest design with 30 university freshmen from various study programs who were members of a debate community at a public university in Surabaya, Indonesia. The study was conducted over three meetings within one week. The first meeting established baseline writing capabilities without AI assistance using pencil and paper. The second and third meetings implemented the human-AI collaborative writing strategy using multiple AI tools including ChatGPT, Gemini, Google Translate, DeepL, QuillBot, and Plagiarism Checker. The strategy incorporated four learning phases (pre-, while-, post-, and outside-learning activities) with ten key instructional steps, emphasizing an integrated writing approach that combined process, cognitive, and scaffolding methodologies. Data collection utilized essay tests administered as both pretest and posttest, with scoring based on adapted rubrics from Hamp-Lyons' Multiple-Trait (MT) and Facione's Holistic Critical Thinking Scoring Rubric (HCTSR). A six-point Likert scale questionnaire assessed student perceptions and experiences. Statistical analysis included descriptive statistics and paired sample t-tests, supplemented by qualitative analysis of argument structures using the AREL (Assertion, Reason, Evidence, Link-back) framework.
Key Findings: The study revealed statistically significant improvements in all aspects of argumentative writing skills. Overall argumentative writing abilities increased dramatically from weak (M = 18.80, SD = 4.18) to good levels (M = 56.87, SD = 4.18). The paired sample t-test showed significant enhancement in students' argumentative writing skills (M = 38.07, SD = 3.99, t(29) = 52.13, p < 0.05) with very large effect sizes (η² > 0.50) across all measured components. Specifically, students demonstrated substantial improvements in constructing well-structured argument traits (t = 54.87, df = 29, p = 0.000, η² = 0.99) and utilizing critical thinking skills (t = 29.65, df = 29, p = 0.000, η² = 0.96). The most commonly used improved argument pattern was C-R₁-E₁-R₂-E₂-I (claim-reason-evidence-reason-evidence-inference), resulting in more elaborated content with an average of 1,176 words across five paragraphs. Students showed overwhelmingly positive perceptions of the strategy, with approval ratings exceeding 85% across all measured aspects including writing ease (98.83%), effectiveness (98.33%), enjoyment (95.50%), and satisfaction (95.00%). All participants (100%) recommended incorporating AI tools in argumentative writing processes.
Implications: The findings demonstrate significant potential for integrating human-AI collaborative approaches in EFL writing instruction, particularly for developing both technical writing skills and higher-order thinking abilities. The study provides empirical evidence that AI tools, when properly integrated with human intelligence and structured pedagogical frameworks, can enhance rather than diminish critical thinking development. The research offers practical guidance for EFL educators seeking to implement AI-enhanced writing instruction while maintaining focus on cognitive skill development. The integrated writing approach framework presented in the study serves as a valuable model for combining process, cognitive, and scaffolding methodologies in technology-enhanced language learning environments. The positive student responses suggest strong acceptance and motivation for AI-assisted learning, indicating potential for widespread adoption in EFL contexts.
Limitations: The study acknowledges several important limitations that affect generalizability. The sample size was relatively small (30 participants) and consisted exclusively of debate community members who were already passionate about argumentative writing, limiting applicability to diverse EFL student populations. The pre-experimental design lacked a control group, preventing direct attribution of improvements solely to the intervention. Using identical essay topics for pretest and posttest introduced potential practice effects and content familiarity bias. The short intervention duration (one week) and absence of longitudinal follow-up provide limited information about sustained impact on writing skills and critical thinking development. The study did not examine student writing motivation as a potential confounding variable, and all participants had some prior familiarity with AI tools, which may have influenced results.
Future Directions: The research suggests several important avenues for future investigation. Studies should include larger, more diverse participant groups to enhance generalizability across different EFL contexts and student backgrounds. Longitudinal research is needed to examine the sustained impact of human-AI collaborative writing strategies on argumentative writing and critical thinking skills over extended periods. Future studies should incorporate control groups to better isolate the effects of the intervention. Research should explore the effectiveness of human-AI collaboration in other writing genres such as narrative or story writing to test broader applicability. Investigation of student writing motivation as a mediating factor could provide deeper insights into the mechanisms underlying strategy effectiveness. Additionally, research should examine the optimal balance between AI assistance and human cognitive engagement to maximize learning outcomes while preventing over-reliance on technology.
Title and Authors: "Effects of human-AI collaborative writing strategy on EFL students' argumentative writing skills" by Raga Driyan Pratama, Utami Widiati, and Lystiana Nurhayat Hakim.
Published On: May 19, 2025
Published By: Computer Assisted Language Learning