Pupillometry Reveals That Context Benefit in Speech Perception Can Be Disrupted by Later-Occurring Sounds, Especially in Listeners With Cochlear Implants.

The cochlear implant group had only 10 participants, and 9 of 10 completed both condition pairs, so group patterns may be unstable. The normal-hearing group was younger than the cochlear implant group, with average ages of 26 years and 62 years, and age affects pupil responses. Three of 40 normal-hearing adults were excluded for poor eye tracking, and 13% of trials were dropped for contamination. The test used one sentence followed by a 2-second sound, not overlapping conversation, so it does not directly measure everyday speech exchange. Normal-hearing listeners heard noise-vocoded speech slowed by 20%, which was designed to be moderately challenging but not identical to cochlear implant listening. Pupil dilation does not scale linearly with difficulty and can shrink when listeners disengage, especially in older or hearing-impaired listeners. The study did not measure memory for the sentence, so it cannot say how much later sounds affect remembering what was heard.

Funding was provided by the National Institute on Deafness and Other Communication Disorders.

Do not conclude that cochlear implant users cannot use context. They did show a context benefit for final-word intelligibility, and pupil measures showed a context benefit in silence. The problem is that the effort benefit came late and was weakened by sounds after the sentence.