It has been well documented that children do not have the same auditory perceptual abilities that adults do. It has been hypothesized that children have underdeveloped auditory sensitivity at the peripheral level which may affect higher level phonetic processing. This may be the reason why children's speech recognition in reverberation and noise is significantly poorer than normal-hearing young adults. The purpose of this investigation is to quantitatively and qualitatively assess the consonant and vowel recognition abilities of normal-hearing children and adults as a function of age, presentation level, and presentation mode (monaural vs. binaural). In experiment 1, 80 normal-hearing listeners with 20 (10 male, 10 female) from each of the following age groups: (1) 6:0-7:11 years, (2) 10:0-11:11 years, (3) 14:0-15:11 years and (4) 18:0- 26:11 years, will respond to consonant and vowel stimuli monaurally presented at 30,40,50, and 60 dB sensation level (SL) (re: speech reception threshold) in four listening conditions of quiet (no reverberation, no noise), reverberation-only (1.5s), noise-only (+17dB S/N - consonants), and reverberation plus noise. Listeners' consonant and vowel sources will be arcsine transformed and submitted to a Three-way (4 age groups x 4 SLs x 4 listening conditions) Analysis of Variance with repeated measures on two factors. From this analysis, any differential effects of age, SL, and listening condition on children's consonant and vowel recognition abilities will be determined. Further, the SL at which all listeners will achieve maximum speech recognition in all listening conditions will be identified and used as the presentation level in experiment 2 which will quantitatively and qualitatively assess the monaural versus binaural consonant and vowel recognition in reverberation and noise. Listeners will respond to consonant and vowel stimuli in the four listening conditions mentioned above. The quantitative analysis will involve submitting listener' arcsine transformed consonant and vowel recognition scores to a Three-way (4 age group x 2 presentation modes x 4 listening conditions) Analysis of Variance with repeated measures on two factors. The qualitative analysis will involve submitting listeners' responses to a symmetric Individual Differences Scaling (SINDSCAL) analysis based on significant main effects and/or interactions between and among the independent variables. Results of this investigation may provide insight into how developing auditory systems interact with reverberation and noise in children's consonant and vowel recognition abilities.
Johnson, C E (2000) Children's phoneme identification in reverberation and noise. J Speech Lang Hear Res 43:144-57 |