The Impact of Hearing Loss on Speech Communication by Spanish-English Bilinguals

Kanae Nishi, PhD  

The overall goal of this project was to compare bilingual (Spanish/English) and monolingual (English only) listeners on consonant perception in noise performance to quantify the performance gap associated with listeners' language background. The focus was on consonants due to their high relevance to speech perception in individuals with hearing loss and to phonological differences between Spanish and English.  Adults who are bilingual since early childhood perform more poorly than monolingual adults do on speech perception in adverse listening conditions even though they can perform on-par with monolinguals in quiet. Young children with normal hearing and persons with hearing loss are also known to show greater detrimental impact of adverse listening conditions than adults with normal hearing. Therefore, treatment of bilingual children with hearing loss presents a greater challenge for speech-language pathologists and audiologists. However, in contrast to the cumulating evidence as to the impact of bilingualism on language development and availability of evidence-based assessment tools, there is no evidence as to the synergistic impact of bilingualism and hearing loss on children's language development, let alone assessment tools. As the first step to address this issue, this research project hypothesized that the size and detailed characteristics of bilingual disadvantage in adverse listening conditions for normal-hearing listeners varied with contextual cues available in utterances. To test this hypothesis, this project compared recognition of consonants embedded in speech materials with varying contextual cues (non-words, non-words similar to real words, real words, and real words in sentences with high and low context) with and without background noise.  This project consisted of two aims: first, to examine the influence of English language skill levels on consonant perception in Spanish-speaking bilingual listeners as compared to English-speaking monolingual listeners; and second, to examine the influence of hearing loss on consonant perception in Spanish-speaking bilingual listeners as compared to English-speaking monolingual listeners.

Altogether, results showed no striking bilingual disadvantage on average or for individual consonant recognition regardless of the background noise levels tested (-5 dB, 0 dB, 5 dB SNR, and quiet), contextual cues (non-words vs. real words), listener age (6-13 years old), or the addition of simulated mild-to-moderate sensorineural hearing loss. Although limited to the listening conditions tested (i.e., static noise, speech presented at 0° azimuth, within critical distance, no reverberation), these results provide new evidence that the detrimental impact of background noise on phoneme recognition is not greater for bilingual children than for monolingual peers. These results can lead to further investigation as to the interaction between phoneme recognition accuracy and the effective use of contextual cues in speech perception in more realistic adverse listening conditions for bilingual listeners.

This project demonstrated that the effect of simulated mild-to-moderate sensorineural hearing loss on consonant recognition accuracy in non-words is similar between monolingual and bilingual listeners with normal hearing. Further, at the phoneme level, regardless of the background noise levels tested, lexical status of stimulus words, or listener age, consonant recognition accuracy was similar between monolingual and bilingual listeners with normal hearing. This study produced a sentence bank which offers an unprecedented large library of sentences that is validated with a wide age range of listeners, and appropriate to use with school-age children.