The Impact of the AI-Based Application ELSA Speak on Students’ English Speaking Skills: A Pre-Experimental Study at the Madrasah Tsanawiyah Level

Abstract

Speaking skills remain one of the most challenging aspects of English language learning, especially for students in madrasah tsanawiyah (Islamic junior high schools) with limited access to interactive speaking practice. This study aims to examine the impact of using the ELSA Speak application in improving the English speaking skills of eighth-grade students at MTs Zainul Anwar Kraksaan, with a particular focus on pronunciation, intonation, and fluency. Employing a quantitative approach with a one-group pretest-posttest pre-experimental design, the study involved 18 students who used the ELSA Speak application regularly for one week. Data were collected through oral tests administered before and after the intervention and analyzed using paired sample t-tests. The results revealed a significant improvement in students’ speaking scores, with the average increasing from 50.61 to 59.50 and a significance value of 0.013. These findings indicate that ELSA Speak effectively enhances students’ speaking performance. In conclusion, the integration of AI-based learning tools such as ELSA Speak offers an adaptive and practical solution for developing speaking proficiency in English, particularly in educational settings with limited resources, while also supporting broader digital transformation efforts in language education at the secondary level.