Childhood hearing impairment: processing dependencies in multidimensional speech perception for an auditory level of analysis.

摘要

OBJECTIVE: Our purpose was to determine whether the influence of childhood hearing impairment (HI) on multidimensional speech processing is a purely linguistic effect or whether childhood HI also affects the processing of speech dimensions representing an auditory level of analysis. DESIGN: The processing dependencies characterizing the two dimensions of talker-gender and spatial location were studied in 40 children with HI and in two normal-hearing (NH) comparison groups representing similar chronological ages (N = 30) or similar vocabulary skills (N = 70). The processing interactions were assessed with a speeded selective-attention task requiring listeners to attend selectively to the gender of the talker and to ignore the spatial location and vice versa. The logic is that performance for the target dimension will not be affected by what is happening on the nontarget dimension if the dimensions are processed independently. On the other hand, if the dimensions are not processed independently, listeners will not be able to attend selectively and performance for the relevant dimension will be affected by what is happening on the irrelevant dimension. In the latter case, results may be analyzed in terms of Garner interference (the effect on performance of irrelevant variability in the to-be-ignored dimension) (Garner, 1974a) and Simon interference (the effect on performance of an irrelevant spatial source) (Simon, 1990). RESULTS: Overall results in all listeners, those with NH or HI, showed significant interference when the participants were attending to the gender of the talker and ignoring spatial location and vice versa. The talker-gender and spatial-location dimensions of speech were not processed independently by these children. When the processing interactions were compared between the NH and HI groups, the presence of childhood HI as a general rule significantly diminished the degree of interference from spatial location. The degree of interference from the gender of the talker, on the other hand, remained normal in the presence of childhood HI. All listeners seemed stimulus bound by the gender of the talker. The degree of Garner interference did not show age-related or degree of loss-related change. The degree of Simon interference showed significant change as a function of age in the children with mild-moderate HI, but not in the children with severe HI. The developmental function for Simon interference in the children with mild-moderate HI was delayed to a degree that corresponded to the duration of the auditory deprivation. CONCLUSIONS: The overall pattern of results indicates that the influence of childhood HI on multidimensional speech processing is not a purely linguistic effect.