Implementing Learning Communities through the ELSA Application to Improve University Students’ English Pronunciation

Abstract

This study aimed to improve university students’ English pronunciation through the implementation of learning communities supported by the ELSA application. The research employed a Classroom Action Research (CAR) design involving first-semester students of Class 24B in the Nursing Study Program at dr. Soebandi University Jember. The study was conducted in two cycles, each consisting of planning, action, observation, and reflection stages. Learning communities were established to encourage collaborative practice, peer feedback, and shared responsibility for learning, while the ELSA application was used to provide automated pronunciation assessment and immediate corrective feedback. Data were collected through pre-test and post-test pronunciation assessments, classroom observations, and questionnaires to examine changes in students’ pronunciation ability, learning engagement, and perceptions toward technology-assisted learning. Quantitative data were analyzed by calculating mean scores and percentage improvements, while qualitative data were analyzed descriptively based on observation notes and questionnaire responses. The findings revealed a significant improvement in students’ English pronunciation, with the average score increasing from 62.3% in the pre-test to 71.0% in Cycle I and further improving to 81.5% in Cycle II. Observational and questionnaire data indicated increased student participation, collaboration, and positive attitudes toward the use of the ELSA application, although improvements in self-confidence were not evenly experienced by all students. Overall, the integration of learning communities and the ELSA application proved effective in enhancing English pronunciation and supporting technology-enhanced collaborative learning in higher education contexts