Hearing disability and hearing devices
Research type
Research Study
Full title
A large-scale study of how listening behaviour in normal and impaired hearing affects disability and provides potential for interventions
IRAS ID
236826
Contact name
Graham Naylor
Contact email
Duration of Study in the UK
4 years, 2 months, 31 days
Research summary
Research Summary
We are trying to find ways for hearing aids and hearing rehabilitation to be more attuned to the individual lifestyle and needs of each hearing-impaired patient. This will allow each patient to get the maximum possible benefit from her or his hearing aid(s).
To do this, we need to understand what aspects of an individual’s lifestyle are important for their feeling of hearing disability; is it a general average feeling across all activities, or is it driven by a few key situations? We will answer this by doing detailed surveys of people’s daily verbal communication activities, and analysing which aspects predict their overall feelings of disability. This will identify particular problem situations where hearing aids need to help most.
We also aim to develop hearing aids which can take advantage of instinctive changes in behaviour which occur in noisy situations (where hearing problems are greatest). Such changes are for example to speak louder or orient the head to one side.
To do this, we must first study whether hearing problems or the use of hearing aids make people behave differently from normal. We will do this using infrared cameras to track people’s body and eye movements during noisy conversations. Then we can identify helpful behaviours whose effect may be ‘amplified’ by a smart hearing aid, and unhelpful ones which a hearing aid could counteract.
Summary of Results
Our programme “Understanding and alleviating hearing disability: The contribution of natural behaviours” has made key contributions to understanding and alleviating hearing loss along three major themes: (1) understanding communication behaviours, by identifying speaking and listening behaviours that contribute to the complexity of people’s communication ecology, (2) real-life measures, by thoroughly investigating measures of listening effort and fatigue and developing/refining self-report methods in the process, and (3) behaviour-compatible technology, by revealing (through fine-grained behavioural measurement) the potential compatibility of hearing device innovation with people’s behaviour, as well as redefining the personalisation of hearing aids. In response to pandemic restrictions, we also developed online methodologies to innovatively address the virtual communication difficulties people with hearing loss face. From detailing the micro-behaviours in one-to-one conversations to reshaping our understanding of hearing healthcare, our programme has generated new insights and methods, and laid the groundwork for new models of hearing disability to incorporate natural behaviours.
1. Understanding specific auditory communication behaviours
To better understand communication behaviours, we collected multi-modal measurements of natural interactions between actively conversing interlocutors. This work was a landmark in that it coupled fine-grained motion capture with natural interaction – others have either looked at vocal production and/or micro-behaviours without the presence of a communication partner, or relied on crude judgments of behaviour in the presence of communication partners – but not the vital combination of both.
These studies have provided us with several insights on how hearing loss can affect communication: (1) in-person interaction produces many vital cues to our behaviour that are absent from previous controlled loudspeaker-based studies; (2) with appropriate calibration the analysis of natural conversation can be commensurate with those previous [artificial] studies; (3) studying conversations of 3+ interlocutors has the potential to yield greater insights than duos; (4) multi-modal measurement in these natural conversations provides us with rich datasets that can be explored and shared to create discoveries at multiple levels of inquiry (e.g., acoustic, behavioural, social, clinical). This work culminated in a review of theories and methods of face-to-face social interaction research, demonstrating the importance of employing interpersonal communication situations to understand the subjective, behavioural and cognitive experience of people with hearing loss in everyday life.
2. Behaviourally relevant measurement of auditory disability and intervention benefit.
To establish relevance in the study of auditory disability, one needs ecological validity. We therefore co-led an international workshop across disciplines to develop a working definition of ecological validity within hearing research, review diverse factors impacting it, and outline strategies to achieve greater ecological validity in such research. Our focus on understanding auditory disability aligns well with this drive towards ecological validity, by leading us both to develop behaviourally relevant measures of listening effort and fatigue, and to move measurement out of the laboratory and into the real world using ecological momentary assessment (EMA). We measured momentary listening fatigue in adults with and without hearing loss by having them respond to short questionnaires at different times of the day, every day for two weeks. This study highlighted how we tend to control our listening fatigue by changing our environments and lifestyles. A follow-up study explored if it was possible to set up and run an EMA study without face-to-face contact, such as during the COVID-19 pandemic. We found that it was indeed feasible to run an EMA study without contact, and identified some vital design features for ensuring compliance. This provides a critical tool for use in times of restricted contact or with participants in remote locales.
3. Hearing-aid fitting and device technology compatible with natural behaviour
Taking insights from our work on communication behaviour (see Section 1 above), we have looked at the potential for hearing devices to take advantage of our natural behaviours to enhance their function, and at how measurement of behaviour can enhance intervention outcome measurement. We have investigated some of the benefits which can arise if a hearing device has its directionality – the way in which it attenuates the sounds not in front of the wearer – steered by the wearer’s gaze. Being able to gather reliable gaze information from the ear canal (exploiting faint electrical signals arising from eye movement) would allow ear-worn hearing devices to be better steered by the listener’s intentions. We looked at methods of improving measuring eye gaze, then looking at how, in our conversation behaviours, we may attend with our head, our eyes or both. We found that people differ in their head-orientation behaviours in situations with multiple talkers. These individual differences are extremely important for the burgeoning field of ‘attention-steered’ speech enhancement, as they imply that the mapping from physiological electrical signals to steering in a practical device would need to be individually calibrated.
Along with designing and evaluating new hearing-aid technology with natural behaviours, we have taken a behavioural perspective on how hearing aids are fitted, examining the clinical process of personalisation through a psychophysical lens. Based on our previous work establishing just-noticeable and just-meaningful differences in speech intelligibility benefits– which has since made substantial impact in research, industry, education and even law – we examined in detail how (in)sensitive people with hearing loss are to changes in the settings of their hearing aids by measuring just-noticeable differences in the devices’ fundamental feature: gain. We then incorporated elements of clinical practice into our methods by studying preferences and descriptions of adjustments and showing that sensitivity to these changes is dependent on how long a clinician speaks to the patient during the adjustment. These results clearly have shown that current practice often adjusts settings on a scale too small to be discriminable, and that alternative practices should be considered.
REC name
West of Scotland REC 3
REC reference
18/WS/0007
Date of REC Opinion
23 Feb 2018
REC opinion
Further Information Favourable Opinion