Actual and predicted word-recognition performance of elderly hearing-impaired listeners.

Word-recognition scores in quiet and in noise were obtained from both ears of 101 elderly listeners demonstrating sensorineural hearing loss. These performance scores were compared to word-recognition scores predicted using Articulation Index analysis procedures. Negative difference scores (actual performance less predicted performance) would reflect aspects of the hearing impairment and/or the aging process that extend beyond the simple speech audibility constraints imposed by the hearing loss and masking noise. The distributions for both the left and right ears of difference scores in quiet revealed the majority of scores to be grouped near 0. In contrast, both distributions of difference scores in noise were normally distributed around means of approximately -25. These results suggest that the typical elderly hearing-impaired listener should be expected to demonstrate word-recognition performance in quiet similar to that of a normally hearing listener, given the same level of audibility of the speech material. On the other hand, in noise, this typical listener may be expected to demonstrate some word-recognition performance decrement, even after accounting for the audibility constraints of the hearing loss and noise.