The effects of noise and hearing loss on conversational dynamics

Understanding how normal-hearing and hearing-impaired people communicate in their everyday life is an important area of research within hearing diagnostics and hearing aid (HA) development. Traditionally, aspects of speech comprehension and speech production are studied in isolation and results from these studies are generalized to real-world performance.

However, the act of conversing is not just the sum of speech comprehension and production. It involves a complex overlap between the two, manifested in a dynamic feedback process between interlocutors, i.e., conversational partners. Thus, there is a need to study people in interaction to understand the impacts of hearing loss and HA amplification on communication.

In this thesis, the effects of hearing loss, noise interference, and HA amplification in measures of temporal dynamics and speech production between interlocutors were investigated in remote and face-to-face conversations. It was shown that in the presence of noise compared to quiet, both normal-hearing (NH) and hearing-impaired (HI) interlocutors responded later and with more variability when taking a turn.

They also produced longer units of connected speech, i.e., interpausal units (IPUs), and they spoke louder. When conversing with an HI interlocutor in the presence of noise, NH participants altered their conversational dynamics more than when conversing with an NH partner. The NH interlocutors changed their strategy from conversing at negative to positive signal-to-noise ratios (SNRs), and they slowed down their speech rates, which were unaltered in NH/NH conversations.

HI interlocutors answered with more variability and produced longer IPUs than their NH partners. When conversing remotely, the HI interlocutors tended to dominate the conversation. When receiving simple HA amplification in face-to-face conversations, the HI participants spoke faster, timed their responses with greater precision, and produced shorter IPUs.

These results suggest that the considered outcome measures were sensitive to communication difficulty and the improvements provided by the HA amplification, even in quiet. Therefore, the proposed measures of communication performance have the potential to be used in the development of new experimental paradigms that evaluate how hearing loss affects real-life communication and to assess the benefit of hearing devices in compensating for these deficits.