Acoustical and perceptual correlates of vocal effort in normal hearing and hearing-impaired children

Abstract

Audition is undoubtedly the most efficient means for early acquisition of oral cormnunication skills (Fry, 1978;Ling, 1976).Consequently, the speech of children with limited audition deviates from normal speech in almost all aspects (Calvert and Silverman, 1975). One problem area is voice quality, specifically vocal effort or harshness. Even though abnormal vocal effort is prevalent among 2 hearing-impaired speakers, there is a paucity of information describing the acoustical and perceptual features of this voice quality for hearing-impaired children. Whitehead and Lieberth (1979) explored vocal harshness in hearing-impaired adults and found a significant and systematic increase in the level of inharmonic energy, indexed as spectral noise, as the degree of perceived vocal harshness/tension became more severe. The constraints of the instrument used by Whitehead and Lieberth, however, precluded using children as subjects. To date, no research has been reported that specifically describes vocal effort or harshness in hearing-impaired children. An apparent need, therefore, existed to implement a research design with which hearingimpaired children could comply and that revealed data concerning the • perceptual and acoustical features of vocal effort in hearing-impaired children.The present study was designed to investigate the perceptual and spectrographic features of vocal effort in the speech of severely to profoundly hearing-impaired children and their normal hearing agemates. Recorded vowel and speech samples were obtained from ten normal hearing children, ten severely to profoundly hearing-impaired children attending Oral/Aural educational programs, and eight severely to profoundly hearing-impaired children attending Total Connnunication programs. The degree of perceived vocal effort for vowels and speech was evaluated, using a nine point equal-appearing-interval scale. In order to obtain a physical measurement for vocal effort, a digital wave analyzer was used to produce vowel spectra, and the amount of inharmonic (noise) components in each spectrum was indexed as spectral noise level.The results indicated: 1) there is a positive and significant correlation of moderate strength between spectral noise levels and perceived degrees of vocal effort for samples of hearing-impaired and normal hearing children; 2) hearing-impaired children tend to demonstrate higher vowel spectral noise levels, although this difference was significant only for the vowel /u/ and not for the vowel /a./; 3 3) hearing-impaired children are perceived as demonstrating significantly more vocal effort than normal hearing children; 4) Total Communication hearing-impaired students cannot be significantly differentiated from Oral/Aural students in terms of spectral noise levels; and 5) Total Communication students are perceived as demonstrating significantly greater vocal effort than Oral/Aural students for the vowel /u/ and for speech.