Acceptable noise level in bilinguals: Consideration of signal and masker effects

Abstract

This study explores how bilingual listeners’ acceptable noise level (ANL) may be affected by the language of the signal, language of the masker, and talkers in the masker. ANL measures how much a listener tolerates background noise while listening to running speech. It differs from conventional speech recognition tasks in that it does not concern one’s ability to comprehend speech. Hence, ANLs are expected to be relatively free of bilingual background. In this study, the signal (running passages from New York State Department of Motor Vehicles driving manual) was presented in two languages (English versus Spanish). The maskers (Auditec babbles) were manipulated in language (English versus Spanish) and number of talkers (four versus twelve). Additionally, three groups of 12 listeners participated in the study—monolingual English, Spanish-English bilingual, and Russian-English bilingual. A 3×2×2×2 mixed, repeated design was carried out with listener group as the between-subjects factor, and signal and masker language and number factors as the within-subjects factors. Preliminary findings reveal a non-significant effect for all four factors; however, a marginally significant four-way interaction (p = 0.048) invites group wise analysis. Results may help establish new clinical approaches in assessing listeners’ speech perception difficulty in noise, regardless of language background.