Effect of noise and reverberation on binaural and monaural word identification by subjects with various audiograms.

The Modified Rhyme Test (MRT) (Bell, Kreul, + Nixon, 1972; Kreul, Nixon, Kryter, Bell, Land, + Schubert, 1968) was used to test the word-identification ability of subjects with bilateral sensorineural hearing losses. The subjects were tested in a room with reverberation times, T, of .1 and .5 sec while listening monaurally, binaurally, aided, and unaided at various speech-to-noise ratios, S/N. A babble of 20 voices was used as a masker. Significant individual differences were found in a group of 15 subjects with various audiograms. The differences were not significant when a group was reduced to seven subjects with similar audiograms. The advantage of the binaural over monaural listening was significant and did not depend on reverberation or the use of hearing aids. Both the effects of S/N and T were significant. The effects on word-identification scores caused by a decrease in S/N from "quite" to +5 dB and by an increase in T from .1 sec to .5 sec were moderately correlated for individuals. There were individual differences in susceptibility to noise and reverberation, with subjects having asymmetrical audiograms being most susceptible. The analysis of consonant errors in noise or reverberation revealed no significant difference for the MRT. A significant difference emerged between consonant errors in the initial and final positions of the test words in both noise and reverberation.