<![CDATA[Although artificial intelligence (AI) is increasingly used in EFL speaking instruction, little is known about its role as a pre-speaking preparation tool for genre-based monologue tasks. This study investigates the effect of AI-assisted speaking preparation on the monologue speaking performance of 9th-grade students in the context of English as a Foreign Language (EFL). Employing a mixed methods design, this study integrated a pre-experimental one-group pre-test–post-test design with qualitative content analysis of student prompts used during AI-assisted preparation. A total of 20 students from a Madrasah Tsanawiyah in Bekasi participated as the single group. Quantitative data were collected through pre-test and post-test assessments of monologue speaking performance in an animal-themed report text task. Quantitative results indicated a statistically significant improvement in speaking scores (t = -9.790, df = 19, p < .001), with an overall mean N-Gain of 0.25, categorized as low. Qualitative analysis of the prompts of three students suggested that more specific and strategic prompts were associated with greater speaking improvement, although this pattern is illustrative and needs confirmation in larger samples. Because the study used a one-group design without a control group, the results indicate that speaking scores improved after AI-assisted preparation, not that AI alone caused the improvement. The findings suggest that how students interact with AI may matter as much as having access to it, and that explicit instruction in prompt formulation may be needed.]]>
Copyrights © 2026
