Exploring The Use of Perplexity AI in Improving Junior High School Students’ Speaking Skills

Abstract

This study investigated the use of Perplexity AI as a brainstorming and speech-outlining tool to improve the speaking skills of junior high school students. A qualitative descriptive research design with a case-study orientation was adopted to examine the language development process of fifteen eighth-grade students at a junior high school in Garut. The participants consisted of fifteen students (male and female) aged around 13–14 years with mixed speaking proficiency levels. Data were collected through classroom observations, reflective interviews, and the administration of pre-tests and post-tests. The students’ oral presentations were evaluated using an adapted analytical rubric measuring five core components: fluency, vocabulary, pronunciation, content, and confidence. The findings showed that all fifteen students experienced an increase in their speaking scores from the pre-test to the post-test. The mean pre-test score was 9.80, while the mean post-test score was 12.27, with an average improvement of 2.47 points. The highest improvements were found in vocabulary and content, followed by confidence, fluency, and pronunciation. Based on these results, it could be concluded that Perplexity AI served as an effective digital academic assistant that helped students overcome limited vocabulary and topic familiarity. However, the study was limited to a small sample, a short duration, and a single AI tool. This study recommends future researchers to explore other specific features of Perplexity AI in English language learning and suggests educators to use this tool as a practical reference for integrating artificial intelligence in speaking classes.