Abstract:
This study investigates the impact of using Artificial Intelligence (AI) as an
assistant on students' writing achievement in teaching descriptive text. This study
aims to determine whether AI improves students' writing skills and identify the
challenges faced by students when learning descriptive text through traditional
techniques. Conducted at SMA Negeri 9 Medan, this study highlights issues such
as limited vocabulary and difficulty in expressing ideas. This study used an
experimental design involving two classes from grade X that were randomly
selected: X-3 as a control group using conventional teaching methods and X-5 as
an experimental group using AI (ChatGPT). The sample consisted of 30 students
in each class. Data were collected through pre-test and post-test assessments to
measure students' writing skills before and after the treatment, The assessment
instrument included a rubric to evaluate descriptive writing, focusing on clarity,
detail, imagery, organization, and overall effectiveness, then all data were
processed using SPSS to conduct a t-test independent sample test which then The
findings showed a significant increase in the average score of the experimental
group from a pre-test average of 45.77 to a post-test average of 86.67, compared
to the control group scores of 42.80 and 76.07, respectively. These results indicate
that the use of AI significantly improved students' descriptive writing skills while
reducing common challenges associated with traditional learning methods.
Consequently, the null hypothesis was rejected, which confirmed that AI
positively affects students' writing skills in a descriptive context. The implications
of this study suggest that integrating AI into educational practices can improve
student engagement and writing proficiency, especially in subjects such as
descriptive texts. This study contributes to the understanding of the role of AI in
language learning and provides practical insights for educators who want to
improve their teaching methodologies.
