Effect of reverberation and noise on the intelligibility of sentences in cases of presbyacusis

For 80 male subjects (age 60--90) and 30 female subjects (age 71--89) the monaural Speech-Reception Threshold (SRT) for sentences was investigated under five reverberation conditions at a constant noise level. The reverberation times used were between 0 and 2.3 s. The noise, with the long-term average speech spectrum, had a level of 52.5 dBA. Each subject was assigned to one of several subgroups formed on the basis of the maximum reverberation time at which the subject was still able to understand the sentences correctly The mean SRTs and the standard deviations are given, per subgroup, as a function of reverberation time. It is shown that, for each subgroup, the SRT in different reverberant sound fields can be expressed as a single number, namely the required Speech Transmission Index (STI) as introduced in room acoustics by Houtgast and Steeneken [Acustica 28, 66--73 (1973)]. Furthermore, it is shown that a model of SRT as a function of noise level, developed by Plomp [J. Acoust. Soc. Am. 63, 533--549 (1978)], can be combined with the STI model and can thus include the effect of reverberation.