Application of Plomp’s model to recognition of the Hearing in Noise Test sentences

Abstract

Sensitive measures of speech recognition are needed for adequate characterization of the effects of various types and degrees of hearing loss. The Hearing in Noise Test (HINT), by Nilsson et al. [J. Acoust. Soc. Am. 95, 1085–1099 (1994)], was developed to quantify the benefits of binaural amplification by measuring a reception threshold for sentences (RTS) in the soundfield in quiet and various noise arrangements. The purpose of the present study was to expand the use of the HINT as a tool to characterize effects of hearing loss in monotic listening conditions. Modeled after Plomp’s [J. Acoust. Soc. Am. 63, 533–549 (1978)] adaptive, speech-recognition paradigm with Dutch sentences, the HINT sentences were presented in quiet and in increasing noise levels via insert receivers to listeners with normal and mild-to-severe hearing impairments. Following Plomp’s analysis, two variables were derived for each listener with hearing loss from the function relating RTS to noise level: A and D, which represent attenuation and distortion effects, respectively. As expected, the pure-tone average was more strongly correlated with A than with D. Derivation of these two variables was also explored using an abbreviated procedure of only two RTS values, one in quiet and one in noise.