Rapid Release From Listening Effort Resulting From Semantic Context, and Effects of Spectral Degradation and Cochlear Implants.

The study analysed 20 normal-hearing listeners and 12 cochlear-implant listeners, so the groups are small. Two additional normal-hearing listeners were tested but excluded because of pupillometry problems, and 20% of trials were rejected for cochlear-implant and vocoded normal-hearing listeners, while 27% were rejected for normal-hearing listeners in the unprocessed condition. The cochlear-implant group was substantially older than the normal-hearing group, and older people can show less pupil dilation, so age may have affected the effort comparison. Participants were not screened for drugs that could affect physiological responses. The normal-hearing vocoded condition used degraded speech, not a cochlear implant, and the cochlear-implant participants had different devices, ages, and experience. Testing was done in quiet with one talker and a simple sentence repetition task, so it may not match everyday conversation with competing speech, noise, or faster turn-taking. Pupillometry requires specialized equipment and is described as impractical for individual clinical use at this time.

Do not conclude that people with cochlear implants cannot use semantic context to reduce listening effort. The study found a detectable context-related reduction for them, although it was smaller and delayed compared with normal-hearing listeners.