tion ability would be related to the strength of the spectral correlation was supported, though a related hypothesis that identification ability would be similarly related was not supported. The dissimilarity of the internal spectra from different stimuli was quantified using the spectral distance measure. A hypothesis that discrimination ability would be related to this measure was also supported. This work was significant for its development of psychophysical techniques, and their application to measuring the internal spectra of musical sounds for cochlear implantees, hearing-aid users and normally hearing listeners. It was also significant for the novelty of applying forward masking to the study of multiple-electrode stimuli derived from complex sounds in the regular environment. The spectral shapes thus measured were found to be related to the ability of subjects to perceive and discriminate sounds. These experimental techniques could easily be applied in the future to non-musical sounds, such as speech. ·
ported. The best implantee showed comparable performance to the normally hearing listeners, though performance was poorer with some of the other hearing-impaired subjects. The spectral correlation was consistently stronger for lower frequencies than higher frequencies. Considered with the weaker correlations observed in hearing impairment, the strength of the spectral correlation was interpreted as being
"This work was significant for its development of psychophysical techniques… …measuring the internal spectra of musical sounds for cochlear implantees…"
influenced by auditory or hardware analysis-filter bandwidth, as well as signal processing for the cochlear implants and hearing aids. The discrimination ability of the normally hearing subjects was close to perfect, though the performance of the hearing-impaired subjects in all groups was markedly poorer. Identification performance ranged from 10% to 75%, showing no systematic relationship to the degree of hearing loss. Identification of sounds in the absence of temporal cues was considered somewhat difficult for even the normally hearing listeners. A hypothesis that discrimina