Abstract
The effect of a speech " recoding " device on the articulation learning (by imitation) of sensori-neural, partially-hearing children was evaluated. It was found that a " frequency shift " of speech components above 4000 c.p.s. down to the residual hearing range of the children, when combined with selective amplification and amplitude compression, led to significantly better articulation learning of consonants with major spectral components in the " shifted " region, than when selective amplification and amplitude compression only were used, the improvement being manifest within only seven training sessions of five to ten minutes, under each condition of amplification. Further, " recoding " produced a small but significant improvement in articulation transfer-of-training, measured by ability to imitate the same consonants occurring in different phonemic contexts, uttered in a different voice to that used in training. However, the superior discrimination and articulation of " recoded " consonants did not suffice to establish the " recoded " words (in which the consonants were embedded) as substitute labels for corresponding pictures (or concepts) in preference to already " known " non-recoded word labels, emphasizing the need in such experiments to separately investigate the effects of a given training procedure or " recoding " technique on discrimination, articulation, and picture (or concept) labelling.
Get full access to this article
View all access options for this article.
