Does Perceived Talker Similarity Play Any Role in Perceptual Training?
Mentor 1
Hanyong Park
Start Date
10-5-2022 10:00 AM
Description
Training with multiple-talker materials has been shown more effective in second language perceptual training than that with single-talker materials. However, it has been rare in prior research to consider whether any multiple-talker training materials would produce such benefits. In this study, we investigated whether the trainee’s perceived talker similarity makes a difference in perceptual training of non-native phonemes. We trained two groups of native English speakers to learn unfamiliar Korean sounds for three days with materials consisting of either 3 dissimilarly sounding talkers or those of 3 similarly sounding talkers. The similarly sounding talker materials were intended for the trainees to perceive only single talker despite multiple talkers being present in the source materials. We reasoned that if the perceived talker similarity influenced the training benefit, the group trained with the dissimilarly sounding talker materials would perform better than the other group. After the training, the trainees were tested twice: once the day after the training was completed and again one week later. The purpose of the latter test was to determine the retention of the training effect after the break in exposure to the talkers. The results of these tests were not borne out as we predicted. The two groups produced similar scores. However, when we re-examined the results while splitting the groups into those who perceived multiple talkers in contrast to those who heard only one talker regardless of their original groups, the prior did better than the latter. This trend was evident through both testing sessions. These results indicate that the perceived talker similarity does matter in multiple-talker materials; multiple-talker training materials will benefit only when the trainees recognize multiple talkers. This may lead to future training materials utilizing decidedly different sounding materials to teach languages.
Does Perceived Talker Similarity Play Any Role in Perceptual Training?
Training with multiple-talker materials has been shown more effective in second language perceptual training than that with single-talker materials. However, it has been rare in prior research to consider whether any multiple-talker training materials would produce such benefits. In this study, we investigated whether the trainee’s perceived talker similarity makes a difference in perceptual training of non-native phonemes. We trained two groups of native English speakers to learn unfamiliar Korean sounds for three days with materials consisting of either 3 dissimilarly sounding talkers or those of 3 similarly sounding talkers. The similarly sounding talker materials were intended for the trainees to perceive only single talker despite multiple talkers being present in the source materials. We reasoned that if the perceived talker similarity influenced the training benefit, the group trained with the dissimilarly sounding talker materials would perform better than the other group. After the training, the trainees were tested twice: once the day after the training was completed and again one week later. The purpose of the latter test was to determine the retention of the training effect after the break in exposure to the talkers. The results of these tests were not borne out as we predicted. The two groups produced similar scores. However, when we re-examined the results while splitting the groups into those who perceived multiple talkers in contrast to those who heard only one talker regardless of their original groups, the prior did better than the latter. This trend was evident through both testing sessions. These results indicate that the perceived talker similarity does matter in multiple-talker materials; multiple-talker training materials will benefit only when the trainees recognize multiple talkers. This may lead to future training materials utilizing decidedly different sounding materials to teach languages.
