Comparing three audio-visual training methods on improving English vowel identification in Cantonese-speaking secondary school students learning English as a second language

Abstract

In Hong Kong, advanced English learners still demonstrate difficulties discriminating English vowels, particularly tense-lax contrasts. This study compared the effectiveness of three intensive and high variability audio-visual training programs in improving English vowel identification in Cantonese-speaking secondary school students. Sixty-seven 12- to 18-year-old students were randomly assigned to one of the three training groups -- the audio-visual group (AV), the enhanced audio-visual group (AV-E) and the hybrid audio-visual group (AV-H), which differed by whether mouth posture was provided as part of the visual feedback and whether focused training was provided for the more confusing vowel pairs. All participants attended 13 English vowel identification training sessions with feedback and were tested at four time points. Results showed that after 8 to 10 hours of perceptual training, all three groups of students, who had received formal English education for 7 to 13 years, significantly improved English vowel identification accuracy in untrained real and non-words produced by novel speakers. No significant difference was found in the training effect of the three methods but the AV-H training method, which included a component of focused training on more confusing vowel pairs, required significantly shorter training duration and was, thus, the most efficient.