Abstract
51 pupils in a school for the deaf, age 9 yrs. 2 mos. to 24 yrs. 2 mos.[long dash]25 [male][male] and 26 [female][female], (21 congenitally deaf) were tested individually with the Maico D-5 Audiometer, and also for discrimination of speech sounds[long dash]vowels and consonants[long dash]by way of paired words. (These tests were presented over a multiple hearing aid since none of the cases could understand speech without amplification.) The I. Q.''s were available. Errors on vowel test and on consonant test are positively correlated. Neither age nor I. Q. was a factor. The higher the frequency loss, the more increase in errors in sound discrimination. Vowel losses are particularly concerned with losses of frequency at 1024 and 2048. Losses at 512, 1024, and 2048 are about equally correlated with errors in consonant loss. Residual hearing at 1024 and 2048 was equally important on the basis of discrimination of vowels and consonants. 5 different methods of determining % loss for speech were employed, the Sabine showing the largest range of % score and the Western Electric the smallest. Every method gave a high correlation between speech and % loss of hearing; when all 5 methods of loss were put together, correlation with speech loss ranged from 0.95 to 0.99. A. concludes {hat percentage method for computing hearing loss is more accurate than has been supposed; but which method was best could not be determined. The study will continue.

This publication has 3 references indexed in Scilit: