12th Speech in Noise Workshop, 9-10 January 2020, Toulouse, FR

The effects of background noise, noise reduction and task difficulty on recall

Andreea Micula(a)
Linköping University, Linköping, Sweden | Oticon A/S, Copenhagen, Denmark

Elaine H. N. Ng, Fares El-Azm
Oticon A/S, Copenhagen, Denmark

Jerker Rönnberg
Linköping University, Linköping, Sweden

(a) Presenting

The Sentence-final Word Identification and Recall (SWIR) test was designed to investigate the effect of hearing-aid signal processing on memory for highly intelligible speech in noise. Previous findings suggest that people with high working memory capacity benefited more from advanced hearing-aid signal processing than people with low working memory capacity. However, people with low working memory capacity showed benefit when a less demanding version of the test was used. Thus, excessive task difficulty could have prevented capturing potential benefit from signal processing.

The aim of this study was to continue developing the SWIR test by manipulating task difficulty, as well as to investigate the effect of task difficulty predictability on recall. Moreover, the effects of noise and noise reduction on recall were investigated. Lastly, the correlation between working memory capacity and recall performance was analyzed for each task difficulty level.

Thirty-two experienced hearing-aid users with symmetrical moderate sensorineural hearing loss participated in this study. The SWIR test task consists of listening to lists of sentences and recalling the last word of each sentence after the list is finished. The SWIR test was administered with noise reduction on and off in competing speech and speech-shaped noise. The task difficulty was manipulated by varying the list length (three, five, seven and nine sentences per list). Half of the participants were informed about list length in advance (predictable task difficulty), while the other half were not (unpredictable task difficulty). Working memory capacity was measured using the Reading Span test.

The results revealed that recall performance was improved when noise reduction was on. However, this improvement was only significant in competing speech when task difficulty was unpredictable. Analysis of the probability of first recall suggested that there was a higher tendency to begin recall with the first list item when noise reduction was off when task difficulty was unpredictable in competing speech. When noise reduction was on, recall tended to start with the last list item. Thus, this finding may be attributed to potential effects of noise reduction on recall strategies. The results also showed a significant positive correlation between working memory capacity and recall performance on lists of five, seven and nine sentences. This finding suggests that the procedure of the test could be modified to be adaptive to individual cognitive capacity.

Last modified 2020-01-06 19:23:55