Objective: The accurate perception of speech requires the processing of multidimensional information. The aim of this research was to examine linguistic influences on the auditory processing of speech in the presence of childhood hearing impairment.
Design: The processing interactions characterizing the linguistic and auditory dimensions were assessed with a pediatric auditory analog of the Pomerantz task (Pomerantz, Pristach, & Carson, 1989). The task yields measures of Stroop interference, the effect of irrelevant semantic content, and of Garner interference, the effect of irrelevant linguistic variability (Stroop, 1935; Garner, 1974a). Subjects were 100 normal-hearing children and 60 hearing-impaired children. Subjects were required to attend selectively to the auditory (voice-gender) dimension and to ignore the linguistic dimension. The logic of the task is that performance for the voice-gender dimension will be unaffected by what is happening on the irrelevant dimension if the dimensions are processed independently. On the other hand, if the dimensions are not processed independently, subjects will not be able to attend selectively and performance for the relevant dimension will be affected by what is happening on the to-be-ignored dimension.
Results: Both the normal-hearing and hearing-impaired children showed auditory Stroop and Garner interference effects, indicating that the auditory and linguistic dimensions were not processed independently by either group. However, the linguistic dimension exerted significantly less influence on auditory processing in the presence of childhood hearing impairment. Whereas normal-hearing children had remarkable difficulty ignoring irrelevant word input and focusing exclusively on voice-gender, hearing-impaired children were relatively successful at ignoring the linguistic dimension and attending selectively to the auditory dimension of speech. This result implies that the linguistic dimension of auditory speech input may have a different weight or processing value in the presence of childhood hearing impairment. It may be the case that hearing-impaired children encode spoken speech disproportionately in terms of the auditory dimensions, which offer important supplementary aids to speechreading. Further research is being carried out to address these possibilities.
Conclusions: Both Stroop and Garner interference were significantly reduced in the presence of childhood hearing impairment. This pattern of results suggests that multidimensional speech processing is carried out in a less stimulus-bound manner in the presence of childhood hearing impairment.