The Application of AI-Powered Adaptive Learning as a Function of Improving Writing Skills and Academic Engagement in Iranian EFL Students

Objective: The present study aimed to investigate the effect of an AI-powered adaptive learning system on the improvement of writing skills and academic engagement among Iranian junior high school EFL students. Methodology: The research employed a quasi-experimental, pre-test/post-test design with a control group. From a population of ۸th-grade male students at a high school in Tehran, ۵۸ participants were selected via convenience sampling and randomly assigned to an experimental group (n = ۳۰) and a control group (n = ۲۸). The experimental group received instruction using the adaptive software Grammarly for Education over ۱۲ sessions, while the control group received traditional, teacher-fronted instruction. Instruments included a researcher-made writing task (inter-rater reliability of .۸۹) and the Reeve (۲۰۱۳) Academic Engagement Questionnaire. Data were analyzed using Multivariate Analysis of Covariance (MANCOVA). Findings: After controlling for pre-test scores, the experimental group significantly outperformed the control group in both writing skills (F(۱, ۵۵) = ۳۸.۴۷, p < .۰۰۱) and academic engagement (F(۱, ۵۵) = ۲۹.۱۲, p < .۰۰۱). The effect sizes (partial eta squared) were .۴۱ for writing skills and .۳۵ for academic engagement. Novelty (Value): This study is among the first in the Iranian educational context to demonstrate the mechanism of AI-powered adaptive feedback not only on the writing product but also on the process of learners' agentic engagement. Conclusion: AI-powered adaptive learning, through the personalization of feedback, simultaneously enhances writing performance and academic engagement, suggesting its potential as an effective supplementary tool in EFL classrooms.

کلیدواژه ها:

: Adaptive Learning ، Artificial Intelligence ، Writing Skills ، Academic Engagement ، English as a Foreign Language (EFL)