APPLYING SPEECH RECOGNITION TECHNOLOGY IN TEACHING SPEAKING SKILL AT THE UNIVERSITY OF LABOUR AND SOCIAL AFFAIRS

Abstract

This study investigates the effectiveness of using Automatic Speech Recognition (ASR) technology, specifically the ELSA Speak application, to enhance English pronunciation among non-English major students at the University of Labour and Social Affairs. A quasi-experimental design was employed with 60 first- and second-year students divided into two groups: an experimental group using ELSA Speak for six weeks and a control group receiving traditional instruction. The results revealed that the experimental group showed significantly greater improvements in segmental accuracy, suprasegmental features, and overall intelligibility. Learner feedback also indicated high satisfaction with the app’s usefulness, ease of use, and timely feedback. These findings highlight the potential of ASR integration as an effective supplementary tool for pronunciation instruction in Vietnamese public university settings and suggest promising directions for technology-enhanced, personalized, and autonomous language learning.