Effects of AI-assisted feedback on the essay writing performance of undergraduate English major students

Abstract

This study investigates the effects of AI-assisted feedback on the essay writing performance of undergraduate English major students. Quasi-experimental pretest–posttest control group design was employed; the study included 72 undergraduate students from Cebu Technological University-Moalboal Campus, Philippines. Students were assigned to the experimental group (AI-assisted feedback via Dola AI) and the control group (conventional teacher-led). Writing performance was measured using a validated analytic rubric covering content, organization, vocabulary, language use, and mechanics. To identify within-group and between-group differences, data were analyzed using paired-samples and independent-samples t-tests. Results showed significant increases in essay writing performance in both groups. Nevertheless, the experimental group demonstrated significantly greater gains across all components than the control group. The independent-samples t-test also confirmed a significant difference in mean gain scores (p < .001), with a moderate effect size (Cohen’s d = 0.53), indicating a meaningful impact of AI-assisted feedback. Results suggest that timely, consistently incentivized AI-generated feedback supports reformulation, leading to better writing. The study underscores the importance of teacher feedback, especially for higher-level writing practices. These findings indicate that AI-assisted feedback is most effective when teacher-guided and embedded within a systematic instructional approach. Conclusions on pedagogical integration and future work are presented.