Comfortable loudness level: stimulus effects, long-term reliability, and predictability

Abstract

This paper reports the results of a series of investigations of comfortable loudness levels with particular reference to their application to hearing aid gain prescriptions. Experiment 1 studied the effects of several stimulus waveforms, bandwidths, and durations on comfortable loudness levels for normal and hearing impaired listeners. Speech band comfort levels were found to be significantly higher than equal-duration noise band or warble tone comfort levels. Comfortable loudness levels were found to be independent of warble tone modulation parameters and of stimulus bandwidth (stimuli did not exceed critical bandwidths). In Experiment 2, reliability of comfortable loudness levels was evaluated in hearing-impaired subjects over two consecutive 1-year periods. Results indicated that comfortable loudness levels were slightly less reliable than thresholds. In addition, the results were consistent with a hypothesis that exposure to amplified sound produces a small increase in comfortable loudness levels. In Experiment 3, data from 67 hearing-impaired subjects were used to develop regression equations for prediction of comfortable loudness levels. Thresholds at the test frequencies were combined with comfortable loudness data at 500 Hz and 4,000 Hz. The prediction method was then evaluated using a new group of 25 subjects. Accuracy of predictions of comfort levels was substantially better with the new method than with an older method that relied exclusively on threshold data. Relevance of the outcomes to hearing aid fitting procedures is discussed.