AI integration in writing skills of MUET students: a quasi-experimental study

This quasi-experimental study examines the effects of artificial intelligence (AI)-assisted feedback on Malaysian University English Test (MUET) writing performance among Form 6 students. A sample of 80 students from two Malaysian secondary schools who were matched in their Lower 6 examination results was randomly assigned to experimental (n=40, AI-assisted feedback) and control groups (n=40, traditional teacher feedback). Both groups received the same classroom writing instruction in a four-week intervention, and only the feedback provision channels are different. The experimental group employed AI writing supports (Grammarly, Hemingway Editor, QuillBot) which generated real-time automatic feedback on grammar, cohesio，organization and lexis diversity. Conventional written teacher feedback was provided to the control group 48-72 hours after attending standard MUET preparation programme.