Englisia
Vol. 14 No. 1 (2026): Englisia: Journal of Language, Education, and Humanities

AI-supported personalized feedback in flipped EFL classrooms: A quasi-experimental study of language proficiency and learner engagement

Abdul Gafur Marzuki (State Islamic University Datokarama Palu)
Huriyah Huriyah (State Islamic University Datokarama Palu)
Erizar Erizar (Unknown)
Anita Anita (State Islamic University Datokarama Palu)
Risa Triassanti (State Islamic University Datokarama Palu)
Dewi Kencanawati (State Islamic University Datokarama Palu)
Muthmainnah Muthmainnah (State Islamic University Datokarama Palu)



Article Info

Publish Date
31 Oct 2026

Abstract

The increasing adoption of artificial intelligence in language education has created new opportunities for timely and personalized feedback, but questions remain about how such feedback should be pedagogically mediated. This study investigated the effectiveness of AI-supported personalized feedback embedded within a flipped classroom environment in improving university students’ English language proficiency and learner engagement. A quasi-experimental design was employed involving two groups of undergraduate EFL students. The experimental group received AI-mediated feedback during independent learning activities within the flipped learning structure, while the control group experienced conventional flipped instruction without automated feedback. Data were collected through pretest and posttest proficiency assessments, learning analytics capturing students’ revision behavior, and perception questionnaires. Quantitative data were analyzed using paired-sample t tests and ANCOVA to examine differences in learning outcomes between groups. The results indicated that students receiving AI-supported feedback achieved significantly higher gains in English language proficiency than those in the control group (ANCOVA, p < .05). The experimental group improved from a mean score of 3.02 in the pretest to 4.12 in the posttest, whereas the control group increased from 3.05 to 3.48. The analysis also indicated a large instructional effect. Learning analytics showed substantially higher revision frequency among students exposed to AI feedback, suggesting stronger engagement with iterative revision and self-regulated learning processes. At the same time, the findings should not be interpreted as evidence that AI feedback can replace teacher judgment, particularly for higher-order concerns such as argument quality, rhetorical purpose, and contextual appropriateness.

Copyrights © 2026






Journal Info

Abbrev

englisia

Publisher

Subject

Description

Englisia Journal (EJ) is open access and peer-reviewed journal that considers any original scientific article that expands the field of language studies in English Language Teaching and various other related applied linguistics themes. The journal publishes articles of interest to language teachers, ...