Abstract
This study examines the usefulness of the audiogram as a predictive instrument in speech intelligibility. Much of the early work on the audiogram centred on speech reception (e.g. Harris, Haines and Myers, 1956 but Montgomery 1967 made a case'for the application to speech production especially when profoundly deaf children are involved. Speech intelligibility was measured by playing back to groups of listeners, tapes consisting of sentences read by 15–16-year-old schoolchildren. Listeners had to fill in two missing words in the sentences. Analyses of the pattern of hearing loss, of the loss at individual frequencies and finally, of combinations of the loss at different frequencies were made. The results indicate that in most circumstances, subjects may be placed in four distinguishable groups, separated at 65dB, 85dB and 105dB. When the simple mean of five frequency losses, centred on 1,000Hz is taken, five groups have significantly different levels of speech intelligibility.

This publication has 8 references indexed in Scilit: