Writing instruction in Indonesian vocational education faces significant challenges in engaging students and providing personalized feedback, yet limited empirical research has examined systematic ChatGPT integration in vocational English contexts. This classroom action research investigated and successfully improved tenth-grade students' writing skills at SMK Amaliyah Sekadau through systematic ChatGPT implementation following Kemmis and McTaggart's model across three instructional cycles. Twenty-six Computer and Network Engineering students participated, with data collected through observation checklists, field notes, writing assessment rubrics, and student response questionnaires. Results demonstrated a 37-percentage-point cumulative improvement from baseline 42.4% to post-Cycle 3 79.0% across five writing components: content, vocabulary, grammar, spelling, and mechanics. Student engagement increased significantly, with 85% expressing enhanced motivation and 88% perceiving ChatGPT as highly useful for writing development. This study revealed that systematic ChatGPT integration provided immediate personalized feedback, enhanced learner autonomy, and improved writing competence in vocational contexts, contributing empirical evidence for AI-infused English language teaching aligned with Indonesia's Merdeka Curriculum framework.
Copyrights © 2026