Advantage Accented? Listener Differences in Understanding Speech in Noise

Cross dialectal communication results in poorer performance than within-dialect communication in a variety of listening tasks. However, some listeners appear to be less affected than others, and this paper explores the factors behind interlistener variation in a listening in noise task. 63 native speakers of American English transcribed 120 HINT sentences, which were presented mixed with noise at -3dB SNR. The sentences had been recorded by six young males: two speakers of Standard American English (SAE), two speakers of Southern American English (STH), and two non-native, L1-Chinese speakers (NNS). Participants were asked to transcribe what they heard as best they could, and were scored on keywords correct. While everyone did much worse with NNS than SAE and STH, participants who self-reported being accented did significantly significantly better with STH (and trend better with NNS) than those who reported being unaccented. Additionally, participants who reported being in a good mood did significantly better with SAE sentences than speakers who reported being in a bad mood. Finally, there was a main effect of extraversion, such that extraverts did worse overall than introverts. The results suggest that individual differences account for some of the interlistener variation in cross-dialectal listening task, and exploring these metrics further may help us understand the cognitive mechanisms involved in processing unfamiliar dialects. This working paper is available in University of Pennsylvania Working Papers in Linguistics: http://repository.upenn.edu/pwpl/