Investigating the Effects of AI and Teacher-Based Explicit Correction on Learner Autonomy and Grammatical Accuracy

This study investigates the effectiveness of AI-based and teacher-based explicit correction on grammatical accuracy and learner autonomy among Arabian EFL learners at the university level. A total of 75 students were initially selected, but following a proficiency test, 54 homogenized EFL learners in terms of general English knowledge were divided into three groups: AI-based explicit correction (via ChatGPT), teacher-based explicit correction, and a control group with no feedback on grammar. The study employed a pretest-posttest design with grammar tests and an autonomy questionnaire administered before and after the intervention. Participants in both experimental groups were provided feedback on grammatical points like conditional type one, conditional type two, and active/passive voice through either AI tools or direct teacher feedback, while the control group received no feedback on the grammatical errors produced in writing texts. Results from the Kruskal-Wallis and ANOVA tests indicated that AI-based corrective feedback significantly improved both grammatical accuracy and learner autonomy compared to teacher-based feedback and the control group. These findings emphasize the potential of AI as an effective tool for enhancing language learning outcomes. The significance of the current study goes back to the role of AI and how it can play an influential role in both improving students'grammar and assisting them to become more independent and self-directed learners, which is an increasingly important skill in today's world. The implications for pedagogy and future research in technology-enhanced language learning are also discussed.