Abstract
Spoken discourse in cochlear implant users is complex due to the perceptual and cognitive challenges associated with processing degraded signals. In this thesis I hypothesized that success in comprehension and memory for discourse should extend beyond solely accuracy but also should address the fatigue from effortful processing of extended discourse. I argue that one should define successful discourse processing as accuracy achieved with minimal effort. In this study, younger and older adults with normal (acoustic) hearing were tested for recall of speech passages using channel vocoding to simulate the sounds delivered by cochlear implants. Participants heard recorded speech passages interrupted at clause and sentence boundaries to allow them to self-pace through a passage. Passages were vocoded to 4 channel, 8 channel, 16 channel, and unprocessed speech. Recall accuracy and effort ratings showed that even as performance accuracy reached an asymptote, for both younger and older adults self-rated effort continued to decline as speech became clearer with more vocoder channels. The analysis of pause durations at sentence and clause boundaries reflected listeners taking time to absorb new information at the beginning of a passage, followed by faster pacing near the end of the passage as the integration of new details about the topic became easier. The vocoder simulations represented in this thesis argue for the clinical evaluation of cochlear implant success to move beyond tests of intelligibility for simple speech materials, and to introduce a definition of success that includes the effort needed to attain this success.