Quality judgments of hearing aid-processed speech and music by normal and otopathologic listeners

Ten normal listeners and ten listeners with high-frequency sensorineural hearing loss were given a paired-comparison task designed to elicit relative quality preferences among hearing aids when male speech, female speech, and music were processed by five different instruments. Results revealed notably high intrasubject test-retest reliability within the two groups of listeners for the male speech condition and in normal listeners for the female speech condition. Preferences assigned for any one of the three stimuli were found to be statistically correlated with those assigned for each of the other two stimuli. Furthermore, preferences of the otopathologic listeners were statistically equivalent those yielded by the normal listeners. Within the separate groups of listeners, rankings derived from the preferences of individual subjects were highly related to those of the other subjects within the respective group, suggesting the lack of a hearing aid-listener interaction. The relationship between preferences and "goodness" of selected electroacoustic characteristics, examined descriptively, was found to be equivocal. Overall results suggest that, while paired-comparison quality judgments may serve as an adequately reliable and adequately differentiating index of behavioral performance with hearing aids, incorporation of such a task within the traditional philosophy of hearing aid evaluation appears unjustifiable.