Towards a physiological measure of listening effort
thesisposted on 2022-03-28, 09:42 authored by Kelly Miles
This dissertation investigated the clinical feasibility of two physiological measures of listening effort; changes in pupil dilation and changes in alpha power (a cortical oscillation between 8-12 Hz). It was motivated by two factors: (1) sustained listening effort can lead to adverse health consequences or could adversely affect healthy behaviours (such as withdrawing socially) yet it is not clinically assessed; and (2) listening effort is a multi-faceted construct, and it remains unknown whether current physiological measures claimed to assess listening effort evaluate the same processes. We conducted laboratory studies on young adults with normal hearing and cognition to better understand how manipulating task difficulty can lead to changes in subjective and physiological measures of listening effort. Specifically, listener-internal factors were investigated by evaluating how working memory capacity, measured with a reading span task, interacted with subjective and physiological measures of listening effort. Listener-external factors were evaluated by examining the effects of channel-vocoding and performance parameters on subjective and physiological measures of listening effort. We also assessed how different data processing strategies and statistical approaches used across studies affect the results and interpretation of physiological outcome measures. The results of these studies indicated that while working memory capacity predicted perceived listening effort ratings, this was not the case for the two physiological measures. Perceived listening effort ratings appeared to be driven by estimated performance, and not the effort required to perform a speech recognition task. On the other hand, the physiological measures were both sensitive to changes in channel-vocoding, and the pupil response was further sensitive to performance levels and task accuracy. The physiological measures were not correlated with each other, suggesting each may be assessing a different aspect of listening effort. Finally, differences in data processing and statistical approach greatly altered the results and subsequent interpretation of the findings. This dissertation provides an opportunity to advance the understanding of listening effort in an experimental setting, and was conducted within the overarching context of exploring the viability of a physiological tool to assess listening effort in a clinical environment.