Promoting EFL Students’ Speaking Performance through ELSA Speak: An Artificial Intelligence in English Language Learning

Abstract

The presence of a technology-based speech learning application namely English Language Speech Application (ELSA), an artificial intelligence (AI) may promote EFL students' speaking performance because of its great features. This research attempts to examine EFL students' speaking performance by using the ELSA speak application, further explore their perceptions of using it, and find out the inhibiting factors influencing low performance in their speaking. A mixed-method research design was applied to obtain both quantitative and qualitative data. 21 EFL students (5 Males and 16 Females) from Universitas Teknologi Yogyakarta were involved in the data gathering process which was collected by using a test, questionnaire, and semi-structured interview. Furthermore, pre-test and post-test were conducted to examine their speaking ability, and a questionnaire using a 4-point Likert scale. In addition, a semi-structured interview was conducted to further explore the inhibiting factors of their low speaking performance. The findings revealed that students' speaking ability has significantly improved after using the ELSA application. It can be known the mean score of pre-test results was 75, meanwhile, the post-test score was 88. There was a 17% improvement after using the ELSA application. Furthermore, of 20 respondents, 18 (90%) of respondents feel confident in speaking English by using the ELSA Speak application. Only 2 (10%) of the respondents disagree and none strongly disagree about the statement. Of 20 respondents, 16 or 80% of respondents have positive perceptions that using ELSA may increase their speaking performance. Moreover, of 20 respondents, 19 or 95% of respondents perceived ELSA to be a great speech application that could boost their motivation to improve their speaking performance. 18 or 90% of respondents perceived that this application has a good quality learning design. Meanwhile, internal and external factors were identified as the inhibiting factors influencing their low speaking performance.