Abstract
Cochlear Implants (CIs) have seen increasing use among older adults with severe to profound hearing loss who can no longer benefit from conventional hearing aids. Although the sounds produced by CIs are severely degraded as compared with natural speech, under ideal circumstances many post-lingually deafened young and older adult CI users can perform quite well on clinical tests of sentence comprehension and recall. A notable deficit appears, however, when CI users hear especially rapid speech, as in the case of artificially time-compressed speech. This may be in part due to a loss of acoustic richness in the time-compressed speech. Combined with the spectrally degraded nature of the CI signal,this would represent a data-limited situation, in the sense that no amount of effort, or allocation of resources, will improve performance. In part there is also a loss of ordinarily available processing time at the linguistic-cognitive level. In this case, performance may be improved by listeners allocating greater effort, hence resources, to the task. This may be referred to as a situation that is resource-limited. In this thesis young adults with age-normal hearing heard and recalled spoken narratives presented under three conditions: (1) Narratives that have been time-compressed, (2) time-compressed narratives with the lost time restored to the original playing time by inserting silent periods at linguistically salient points to allow listeners’ processing time to “catch up”, and (3) compressed and restored narratives processed with noise-band vocoding to simulate the sound produced by a CI. The data-limited versus resource-limited distinction is explored in both cases via a continuous measure of pupil dilation as an index of processing effort. Data from a small sample of CI users are also offered.