This study investigates the effectiveness of integrating the ELSA Speak as an artificial intelligence (AI)-based language learning application to enhance students’ speaking competence following an eight-week course intervention. A quantitative, one-group pre-test–post-test design was used with 32 undergraduate students enrolled in a Receptive Oral Language Skills course. Speaking performance was assessed in five reported dimensions: pronunciation, fluency, vocabulary, grammar, and confidence/comprehensibility. The reported mean scores increased from pre-test to post-test in all five dimensions. Pronunciation showed the largest gain (from 65.2 to 84.5; +19.3 points), followed by confidence (64.9 to 83.1; +18.2), grammar (66.2 to 82.5; +16.3), fluency (66.7 to 80.4; +13.7), and vocabulary (68.1 to 81.8; +13.7). A post-intervention questionnaire also indicated favorable perceptions: across the ten reported items, 78.1%–84.4% of respondents selected “Strongly Agree,” while the remaining 15.6%–21.9% selected “Agree.” Because the available results contain descriptive summaries only, no claim of statistical significance is made. The findings show improvement over the intervention period and positive learner perceptions, but the one-group design cannot establish that ELSA Speak alone caused the changes. The results support considering ELSA Speak as a supplementary practice tool alongside teacher-guided communicative activities, while further controlled research is needed.
Copyrights © 2026