The impact of rise time discrimination, language ability, memory, and attention on speech recognition in noise

Background: The ability to perceive and recognise speech in background noise is crucial for effective communication in everyday life. However, what factors contribute to this ability is not well understood. Several studies have highlighted that cognitive factors such as attention and memory are relevant for speech recognition in noise. The role of auditory processing is less well-established, however. Auditory processing is not a unitary skill and encompasses several sub-skills that target binaural processing and analysis of spectral and temporal cues. This study consequently aimed to investigate the importance of auditory processing skills, specifically the perception of rise times and amplitude modulation, cognitive skills, such as working memory and selective attention, and language ability on speech recognition in noise and to determine which of these abilities predicts speech recognition in noise.

Participants and Methods: Forty-eight young normal-hearing participants (aged 19.3 to 35.6 years) participated in the current research. They all underwent the measures of auditory processing skills, including the perception of rise time and amplitude modulation, working memory, selective attention, and language ability. The dependent variable was the sentence recognition score in noise as measured by The Test of Listening Difficulties – Universal (ToLD-U). Predictor variables included the scores on rise time discrimination tasks, Parsing Syllable Envelopes (ParSE) test, backward digit Span, Test of Everyday Attention, and AudiCloze. Correlation and regression analysis were performed.

Results: The results of correlation analysis showed that only the AudiCloze, a measure of language ability, was significantly correlated with the ToLD-U score, r(46) = -.59, p < .001. There were no significant correlations between ToLD-U and other measures. However, the correlation with selective attention was close to significant, r(46) = -.27, p = .063. Regression analysis showed that AudiCloze and selective attention scores explained 58% of the variance (R² = 0.58, F(2, 45) = 30.9, p < .001) in speech recognition in noise performance on ToLD-U.

Conclusion: The findings are consistent with previous literature that language and selective attention are significant predictors of speech recognition in noise in young normal-hearing adults. The chosen auditory processing skills were not associated with speech recognition skills, at least for young adults. It is possible that the role of auditory processing is more relevant in children when language and attention skills are still maturing. To assess this hypothesis, future studies will evaluate the association between auditory processing skills and speech recognition in noise in school-aged children.