Doctor of Philosophy
Communication Sciences & Disorders
A variety of noises, including multi talker babble (MTB), speech spectrum noise (SSN), and several other environmental auditory scenes that had MTB as the dominant noise, have been used in research in the field of audiology to evaluate speech perception in individuals with normal hearing and hearing loss. These different noises and environmental auditory scenes have been employed independently to evaluate speech perception performance. Given that there are several sounds in the real world that might impair communication, the evaluation of speech perception accuracy should encompass all noises that a listener may encounter in typical conversation. Combining several auditory scenes (including MTB) into a single noise for speech perception testing would be efficient and valuable. The studies presented here aimed to address this problem by integrating multiple realistic auditory settings that have previously been used to assess speech perception into a single noise (including MTB). In the first study, an ecological noise (EN) was produced by combining various natural noises with MTB. The impact of this EN on speech perception performance was then evaluated and compared to that of other noises (MTB & SSN). In the second study, the influence of the MTB in the EN was assessed by producing three different ENs: one without babble, one with babble, and one with reverse babble. In both studies, sentences from the QuickSIN, AzBio, and HINT were used to measure speech perception. Behaviorally, the findings of the current work showed that the speech perception performance for the EN used here was poorer compared to the other noises and the presence of MTB affected speech perception. Acoustically, the spectrum of the EN was similar to the other noises, however the EN was highly modulated in the low frequencies. Ecological noise creates a more difficult listening environment than other noises alone. Speech perception scores obtained using this EN may provide a more realistic expectation of communication in the natural world compared to MTB or SSN, which would be useful for counseling those with hearing loss. Findings from the current work also showed that the QuickSIN was the most sensitive test to measure speech perception in this EN.
Dissertation or thesis originally submitted to ProQuest.
Shukla, Bhanu, "SPEECH PERCEPTION PERFORMANCE IN ECOLOGICAL NOISE" (2022). Electronic Theses and Dissertations. 3261.