Speech-on-speech recognition in children and adults: Effect of whispered speech
Margaret K. Miller, Lori J. Leibold, Lauren Calandruccio, Emily Buss · The Journal of the Acoustical Society of America · 2016
The harmonic structure of speech appears to provide adults with a robust auditory grouping cue, improving speech-on-speech recognition performance. This study evaluated the effect of harmonicity on speech-on-speech recognition in normal-hearing children (5-7 yrs) and adults (19-35 yrs). Masked word recognition was evaluated using a 3AFC procedure with a picture-pointing response. Targets were one-syllable words that lent themselves to illustration, and maskers were samples of two-talker speech. All stimuli were recorded by female talkers in two styles: whispered and natural. Thresholds corresponding to 70.7% correct word recognition were estimated adaptively by adjusting SNR while holding the overall level of the signal-plus-masker at 60 dB SPL throughout testing. Performance was evaluated for whispered target words presented in a whispered two-talker masker and for naturally produced target words presented in a naturally produced two-talker masker. Consistent with previous reports, adults required a more advantageous SNR to achieve criterion performance using whispered compared with natural speech. Children performed more poorly than adults in all conditions, and tended to show a smaller effect of speaking style than adults. These results suggest that children may be less adept than adults at using the harmonic structure of speech to segregate simultaneously presented voices.