Toward Measuring How Well Hearing-Impaired Children Speak Average intelligibility scores for a group of 37 hearing-impaired and two normally hearing adolescents were determined by 50 normal listeners and were compared with nine acoustically measured speech variables. These nine variables included measurements of consonant production, vowel production, and prosody. Regression analysis of the variables showed that three of ... Research Article
Research Article  |   June 01, 1978
Toward Measuring How Well Hearing-Impaired Children Speak
 
Author Affiliations & Notes
  • Randall B. Monsen
    Central Institute for the Deaf, St. Louis, Missouri
Article Information
Research Articles
Research Article   |   June 01, 1978
Toward Measuring How Well Hearing-Impaired Children Speak
Journal of Speech, Language, and Hearing Research, June 1978, Vol. 21, 197-219. doi:10.1044/jshr.2102.197
History: Received March 15, 1977 , Accepted September 8, 1977
 
Journal of Speech, Language, and Hearing Research, June 1978, Vol. 21, 197-219. doi:10.1044/jshr.2102.197
History: Received March 15, 1977; Accepted September 8, 1977

Average intelligibility scores for a group of 37 hearing-impaired and two normally hearing adolescents were determined by 50 normal listeners and were compared with nine acoustically measured speech variables. These nine variables included measurements of consonant production, vowel production, and prosody. Regression analysis of the variables showed that three of the speech variables bore a multiple correlation of 0.85 with measured intelligibility scores. Two variables alone, the mean voice-onset-time difference between /t/ and /d/ and the mean second-formant difference between /i/ and /ɔ/, accounted for about 70% of the variance in the intelligibility scores. To cross-validate the reliability of these correlations, intelligibility scores were subsequently predicted for another group of 30 hearing-impaired adolescents and then compared with intelligibility scores as determined by another group of normal listeners. For this second group, the correlation between measured intelligibility scores and predicted scores was 0.86, which indicates that the reliability of the predicting variables is high. Five of the nine variables correlated more highly with measured speech intelligibility than did pure-tone audiometric thresholds. The average speech intelligibility of all 67 hearing-impaired subjects was 76%.

Order a Subscription
Pay Per View
Entire Journal of Speech, Language, and Hearing Research content & archive
24-hour access
This Article
24-hour access