Abstract
PURPOSE: This project evaluates whether diverse linguistic experience influences second language (L2)-accented English sentence recognition in Spanish-English bilingual and English monolingual school-age children and adults. METHOD: Four groups of participants completed the study: 22 Spanish-English bilingual school-age children, 21 English monolingual school-age children, 19 Spanish-English bilingual adults, and 19 English monolingual adults. Participants completed English sentence recognition and English language testing. Sentence recognition was assessed in speech-shaped noise at ascending target-to-masker ratios in three accent conditions: Korean (unfamiliar to all), Midland (familiar to all), and Spanish (familiar to the bilingual group). Speech recognition thresholds associated with 50% correct performance (SRT50) and asymptotic performance were extracted from psychometric function fits to behavioral data. RESULTS: Adults had lower (better) SRT50s in the Spanish and Korean accent conditions compared to children. Estimates of recognition in quiet were better for bilingual children tested in the Spanish and Korean accent conditions compared to monolingual children. The effect size was larger in the Spanish accent condition. There were no language group differences in performance among adults. CONCLUSIONS: For children, prior experience with a specific L2 accent offers significant benefit in accurately recognizing sentences spoken by talkers of that accent. This effect is not observed in adults, potentially reflecting differences in language experience between Spanish-English bilingual child and adult participants. Although greater diversity in linguistic input is also positively associated with recognition of unfamiliar accented speech in children, its effect is smaller than that of accent-specific experience. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.30827552.