Speech Perception in Noise Predicts Oral Narrative Comprehension in Children With Developmental Language Disorder

Remote Microphones Support Speech Recognition in Noise and Reverberation for Children With a Language Disorder

PURPOSE

Children with typical hearing and various language and cognitive challenges can struggle with processing speech in background noise. Thus, children with a language disorder (LD) are at risk for difficulty with speech recognition in poorer acoustic environments.

METHOD

The current study compared the effects of background speech-shaped noise (SSN) with and without reverberation on sentence recognition for children with LD (n = 9) and typical language development (TLD; n = 9). We also investigated whether the use of a remote microphone (RM) improved speech recognition for children with LD.

RESULTS

Children with LD demonstrated poorer sentence recognition than peers with TLD in SSN. Both groups had poorer sentence recognition with SSN + reverberation than SSN alone. Notably, using an RM improved speech recognition for children with LD in SSN and SSN + reverberation.

CONCLUSION

We discuss educational implications and future research questions to identify how to optimally support speech recognition in noisy environments for children with LD.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.28037984.

Low-Level Auditory Processing Correlates With Language Abilities: An ERP Study Investigating Sequence Learning and Auditory Processing in School-Aged Children

Auditory processing and procedural learning deficits have been associated with language learning difficulties. We investigated the relationship of these skills and school-age language abilities in children with and without a history of late talking using auditory event related potentials (ERPs). Late talking (i.e., slow early language development) increases the risk of persistent language difficulties, but its causes remain unknown. Participants in this study were children with varying language abilities (n = 60). Half of the participants (n = 30) had a history of late talking. We measured procedural learning by manipulating the predictability of sine tone stimuli in a passive auditory ERP paradigm. Auditory processing was tested by examining how the presence of noise (increasing perceptual demands) affected the ERPs. Contrary to our hypotheses on auditory processing and language development, the effect of noise on ERPs did not correlate with school-age language abilities in children with or without a history of late talking. Our paradigm failed to reveal interpretable effects of predictability leaving us unable to assess the effects of procedural learning. However, better language abilities were related to weaker responses in a 75-175 ms time window, and stronger responses in a 150-250 ms time window. We suggest that the weak early responses in children with better language ability reflect efficient processing of low-level auditory information, allowing deeper processing of later, high-level auditory information. We assume that these differences reflect variation in brain maturation between individuals with varying language abilities.

The Effect of Background Noise, Bilingualism, Socioeconomic Status, and Cognitive Functioning on Primary School Children's Narrative Listening Comprehension

PURPOSE

This study focuses on 7- to 9-year-old children attending primary school in Swedish areas of low socioeconomic status, where most children's school language is their second language. The aim was to better understand what factors influence these children's narrative listening comprehension both in an ideal listening condition (in quiet) and for the primary school classroom, a typical listening condition (with multitalker babble noise).

METHOD

A total of 86 typically developing 7- to 9-year-olds performed a narrative listening comprehension test (Lyssna, Förstå och Minnas [LFM]; English translation: Listen, Comprehend, and Remember) in two listening conditions: quiet and multitalker babble noise. They also performed the crosslinguistic nonword repetition test and a digit span backwards (DSB) test. A predictive statistical model including these factors, the children's degree of school language exposure, parental education level, and age was derived.

RESULTS

Listening condition had the strongest predictive value for LFM performance, followed by school language exposure and nonword repetition accuracy. Parental education level was also a significant predictor. There was a significant three-way interaction effect between listening condition, age, and DSB performance.

CONCLUSIONS

Multitalker babble noise has a negative effect on children's narrative listening comprehension. The effect of multitalker babble noise could be explained by age differences in the ability to allocate working memory capacity during the narrative listening comprehension task, suggesting that younger children may be more vulnerable for missing information when listening in background noise than their older peers.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.25209248.