Masking additivity in the hearing-impaired

The purpose of the present study was to investigate how the effects of maskers combine in hearing-impaired listeners. The levels of two narrow bands of noise separated in frequency were selected so that each band individually produced the same amount of masking of a sinusoidal signal. Thresholds for the signal were then measured in the presence of both bands combined. For normal-hearing listeners, the effect of combining maskers was nonlinear, but, additive. A cube-root transformation of the masking produced by the individual and combined maskers preserved an additive relation among the effects of these maskers. This type of nonlinear additivity also characterized the results from the hearing-impaired listeners when the maskers were restricted to a region of normal sensitivity. In contrast, the effect of combining maskers was linear, additive without transformation, when the maskers were restricted to a region of significant loss. In many instances, the auditory filter model of masking may describe the detection performance of hearing-impaired listeners better than that of normal-hearing listeners.