Grammar errors are a common problem encountered by senior high school EFL students when writing narrative texts. This study aimed to examine changes in students' simple past tense accuracy following the use of ChatGPT feedback in narrative writing. A pre-experimental one-group pre-test and post-test design was employed with 21 students from Class XI at SMAN 1 Sungai Raya, Kubu Raya Regency, West Kalimantan, Indonesia. The intervention was conducted in one session, during which students used ChatGPT 3.5 through their personal smartphones to obtain feedback on their narrative writing. The feedback focused on simple past tense accuracy, and students used the suggestions to revise their texts. Data were collected through pre-test and post-test writing tasks and analyzed using a paired-samples t-test and effect size analysis. The results showed a statistically significant improvement in students' scores (p = 0.010), with an effect size of 0.62, indicating a medium effect according to Cohen's criteria. These findings suggest that ChatGPT feedback was associated with improved simple past tense accuracy in narrative writing. However, the absence of a comparison group limits conclusions about its causal effects. Therefore, it cannot be concluded that the improvement that students got is solely due to the use of ChatGPT
Copyrights © 2026