EVALUATING THE EFFICACY OF ARTIFICIAL INTELLIGENCE-BASED SPEECH RECOGNITION APPLICATIONS ON THE DEVELOPMENT OF PRONUNCIATION PRECISION AND PHONETIC AWARENESS IN ENGLISH AS A SECOND LANGUAGE CONTEXTS

This mixed-methods study explored the effects of AI-driven speech recognition technology on the accuracy of pronunciation and phonetic awareness in 100 participants who are learning English as a foreign language at the University of Lagos. The research method included a quantitative experiment with a quasi-experimental design and a qualitative focus-group interview to evaluate the students' experience. The experimental group received treatment in the form of individual practice with speech recognition software, while the control group got traditional phonetic instruction. Both groups were tested twice, and their segmental and suprasegmental accuracy were measured on a scale from 0 to 100. The initial mean score of the experimental group was 64, and it improved to 85 after the intervention (t(49) = –8.96; p < .001). The performance of the control group was within the same range, from 65 to 68. The qualitative results supported the findings by demonstrating that the students who used AI-driven practice had increased confidence, self-regulation, and phonetic awareness. Their creative expression and critical thinking were stimulated by the possibility of making mistakes in the computer-generated environment. Focus-group participants identified phonetic challenges in Nigerian English which were not addressed in the traditional curriculum, such as the pronunciation of long vowels, stressed syllables, and consonant clusters. Overall, the study demonstrated that the strategic use of speech recognition technology can facilitate the development of pronunciation skills and phonetic awareness in English as a foreign language learners