the Speech recognition under adverse listening conditions in young normally-hearing listeners

In the present study we aimed to investigate individual differences in cognitive inhibition, WMC, and how they relate to performance on a speech-recognition-in-noise task. Sixteen young normally-hearing individuals were presented with a cognitive test battery, as well as a sentence corpus masked by 5 different maskers, targeting 80% speech-recognition. One masker was a slightly modulated (10%) speech-shaped noise (SSN), 2 maskers were constructed by modulating the SSN with the envelopes from a single female talker, and the international speech test signal (ISTS). We also masked the target sentences with the ISTS, and a single female talker reading a passage in a Swedish newspaper. Our results showed that cognitive inhibition is significantly related to performance when maskers with meaningful, semantic information is used. The results further indicate that young normally-hearing individuals can take advantage of temporal and spectral dips to fill in missing information. Our findings suggest that choice of speech material is of importance for the outcome in speech-recognition-in-noise tasks. We further propose that tasks of cognitive inhibition can be used to predict performance in a speech-recognition task.