The Development of English Communication Skills using Artificial Intelligence for Students in Ban Sai Dong School Under Surat Thani Primary Education Service Area Office 3

Abstract

This research aimed: (1) to design an instructional model that promotes the development of listening and speaking skills in English using artificial intelligence (AI) for students at Ban Saidong School, and (2) to evaluate the students’ listening and speaking skills after learning through AI compared to the 70 percent proficiency benchmark. The target group consisted of 22 Grade 4–6 students taking English course at Ban Sai Dong School in the first semester of the 2025 academic year. The research instruments included seven English lesson plans integrated with AI, totaling 18 instructional hours, and a listening and speaking skills assessment. Data were collected using a pre- experimental method and analyzed through percentage. The research findings revealed that: 1) The instructional design in the English subject, titled Real Communication, aimed to enhance students’ communication skills in real-life contexts. The content was related to students’ daily lives and followed five instructional stages: Warm-up, Presentation, Practice, Production, and Wrap-up. These stages helped students progressively acquire knowledge, practice it, and apply it meaningfully. AI played a central role by supporting listening and speaking activities. Before implementation, the teacher identified areas where AI could enhance students’ performance. AI was then used to ask questions and interact with students across the seven learning units, providing opportunities for authentic practice. Teachers evaluated student performance using a rubric. The seven units included self-introduction, daily routines, likes and dislikes, giving directions, expressing feelings, ordering food, and telling the time. These topics provided essential language for everyday communication, and 2) The test results showed that 9 out of 22 students (40.91%) achieved the benchmark of 7 out of 10 points, while 13 students (59.09%) scored below it. This indicates that AI-based instruction supported students’ listening and speaking development, though most learners did not yet reach the proficiency level required.