Binaural Speech Recognition in Noise and the Effect of Context A Senior Honors Thesis Presented in partial fulfillment of the requirements for graduation with research distinction in Speech and Hearing Science in the undergraduate colleges

Listening and understanding speech in noisy environments is a situation that many people encounter in their daily lives. This problem is exacerbated by old age or the presence of a hearing loss. The purpose of the present study was to investigate two components that contribute to successful word recognition in noise. The first component is the advantage of binaural over monaural listening and the second is the role of contextual cues. Ten normal hearing young adults aged 20-24 years old participated. Sentences from the Speech Perception in Noise (SPIN) test were presented to the subject through insert ear phones simultaneously with multitalker babble as background noise under three conditions: 1) noise and signal presented to the right ear, 2) noise and signal presented to the left ear, and 3) noise and signal presented to both the left and right ears. The subject was told to repeat the last word in each sentence and the responses were scored as correct or incorrect. Results show that subjects perform better on word recognition tasks in the binaural listening condition versus the monaural listening conditions. Performance was also better with high predictability sentences in which the context can be useful for word recognition. Literature has shown that older adults also rely on context for word recognition, so it can be expected that for the older adults who are hard of hearing word recognition would be most successful when listening with two ears and when presented with high contextual information. SPEECH RECOGNITION IN NOISE 3 Table of