A Computer-based Perceptual Training of Korean L2 Speakers: Focusing on Speech-in-Noise Recognition Abilities

The current study aims to examine if a computer-based perceptual training on vowel/focus recognition can enhance speech-in-noise recognition abilities by L2 learners. As speech-in-noise recognition has been suggested to be able to represent L2 proficiency, this study examined if the improvement in other aspects of the target language (i.e., English) can be extended to the improvement of speech-in-noise recognition. Thirty-six Korean L2 learners of English completed two different types of training sessions, as well as speech-in-noise recognition tests in pre-, mid-, and post-training sessions. The results revealed that both training types contributed to improve participants’ speech-in-noise recognition abilities, however, the improvement did not persist after the first training session. The results indicate that speech-in-noise recognition might reflect L2 learners’ proficiency, and both segmental (vowel training) and suprasegmental (focus training) and recognition is related to speech-in-noise recognition.