Abstract
In this study of speech perception, subjects were children, 9 with sensorineural hearing loss and 20 with normal hearing. Two series of speech identification experiments were conducted. In Experiment 1, five isolated vowels, /a, i, u, e, o/, were used; Experiment 2, 12 consonant-vowel syllables, each comprised of one of the consonants /k, s, t, p/ in combination with one of the vowels /a, e, o/ were used. Segments consisted of the steady 12.8 to 102.4ms of each vowel in Experiment 1, and the initial 12.8 to 128ms of each consonant-vowel syllable in Experiment 2. In both experiments, the sounds spoken by an adult female speaker were temporally segmented in 12.8ms steps. Experiment 1 showed that, for normal-hearing children, only 51 ms were needed to identify vowels correctly. However, for children with sensorineural hearing loss, 108ms were insufficient for the correct identification of vowels. For all subjects in this experiment, the responses given tended to be vowels that resembled the F2-F1 values of the stimuli. When subjects with sensorineural hearing loss could not detect F2, their responses were based entirely on the F1 values. Experiment 2 showed that the aperiodic portion of a consonant-vowel syllable is sufficient for the correct identification of consonants by normal-hearing subjects. Subjects with sensorineural hearing loss seemed to use the contextual effects of consonant-vowel syllables in order to identify the consonant. Consonant-confusion patterns observed in subjects with sensorineural hearing loss were similar to those in subjects with normal hearing. Subjects with sensorineural hearing loss demonstrated poorer consonant categories than vowel categories. Although subjects with moderate hearing impairments did not differ in vowel-identification performance from those with severe impairments, they did differ in performance on consonant-vowel syllable identification. Analyses of the date from both experiments indicate that speech categories in hearing-impaired children are inferior to those in children with normal hearing.
