The goal of the proposed work is to better understand several key factors that influence how hearing-impaired listeners recognize speech with and without amplification. This goal will be realized by evaluating normal-hearing and hearing impaired subjects under conditions that control for variations in performance due to audibility. The proposed projects were selected because of their implications for theories about reduced speech perception by the hearing-impaired, hearing aid selection, hearing aid benefit assessment and speech recognition testing. Recent developments in those areas reveal that the matters to be investigated have not been studied adequately to meet current needs. Methods developed in this laboratory over the past several years provide an improved way to study these issues. Their previous work suggests that these methods will offer new insights into the factors that influence speech recognition over a range of acoustic and listener conditions. This work will be guided by four specific aims. They are to: 1) Measure the intensity importance function and the effective dynamic range of two types of speech (monosyllabic words and connected speech) using a new procedure devised specifically for this purpose. 2) Study the effects of signal and masker levels on the speech recognition performance of hearing-impaired listeners using techniques that minimize variations due to audibility. 3) Study upward spread of speech self-masking in the hearing-impaired over a range of listening conditions and hearing losses. 4) Evaluate the consistency of individual performance with several types of speech (monosyllabic words and connected speech) and acoustic condition (masking and reverberation) and determine if the differences that remain after controlling for audibility loss are related to individual differences in the noise-masked pure-tone threshold measured with three methods. The investigations that will be carried out in support of these specific aims will use noise-masked, filtered and reverberated speech stimuli presented through earphones at multiple signal-to-noise ratios and intensity levels. The results will be interpreted within the framework of audibility theory, an extension of the classical articulation index theories of Fletcher & Galt (1950) and French & Steinberg (1947).
