The intelligibility of sentences in quiet and in noise in aged listeners

The monaural speech-reception threshold (SRT) for sentences was investigated in quiet and under four noise conditions for 80 male subjects (age 60-90) and 30 female subjects (age 71-89). The noise levels used were 28, 43, 58, and 73 dBA. The noise had the long-term average spectrum of speech. It is shown that a model developed by Plomp [J. Acoust. Soc. Am. 63, 533-549 (1978)], which interprets any hearing loss for speech (SHL) as a combination of a loss of class A (attenuation of both speech and noise) and a loss of class D (distortion of the sound signals), enables one to make an accurate description of the SRT values measured. Fitting this model, for each individual, to the SRT data yielded values for SHL in quiet (= A + D) and in noise (= D). These individual SHL values were studied in relation to one another and in relation to such parameters as the pure-tone average (PTA), the Fletcher Index (FI), and the increment in intelligibility score per dB near SRT. It was found that (1) subjects with the same SHL in quiet may differ considerably in their SHL in noise; (2) for an individual, PTA and FI are inaccurate measures for predicting SHL in quiet and in noise, and (3) the higher an individual's SHL in noise, the lower is the increase per dB of the intelligibility score for sentences in noise.