An 'Off-the-shelf' Assistive Listening Device: Normal-hearing Children.

Testing Speech Intelligibility Outcomes With a Commercial Bone-conduction Headset in Children With Normal Hearing.

The cochlea, the sensory organ of hearing, is a structure of the temporal bone on the skull. In everyday life sounds are heard via air conduction. This means that vibrations in the air are conducted through our ear canals, via the eardrum and the middle-ear bones, to the cochlea. However, vibrations can be conducted to the cochlea via the bones of the head. Bone-conduction headsets have become popular for recreational use (for example cyclists and runners wear them to listen to music while exercising). When in a noisy environment, if a speech signal is delivered to a microphone connected via Bluetooth to the bone conduction headset, the person wearing the headset receives the speech signal as if the talker were closer to them. The ratio between the speech level and the noise level (SNR, signal-to-noise ratio) is increased, so that it is easier to understand the spoken message. A previous study carried out by the investigators has shown that this may help children with hearing loss due to otitis media with effusion ('glue ear'). The aim of the current study is to explore the potential of the headset to help children with auditory processing disorder (APD). Typically, children with APD have normal audiograms, but, in spite of this, they struggle to understand speech in a background noise. The headset can deliver the speech message to them. Currently, FM systems are used for children with APD in the classroom. These systems are effective, but their cost is high and provision may be limited. The feasibility of the use of the headset in a group of children with normal audiometric thresholds will be assessed. The study hypothesis is that using a bone-conduction headband improves speech recognition in noise and decreases listening effort even when air-conduction hearing thresholds are normal. Measures of speech recognition and listening effort will be done in quiet and in noise with and without the bone-conduction headset in order to measure the effect of using the headset on speech recognition when hearing thresholds are normal.

A repeated-measures model will be used. This means that participants in a single arm will be tested in all conditions. Speech recognition and listening effort outcomes will be measured in two conditions: with and without a bone-conduction headset.

The main outcome measure will be an age-appropriate speech test. This test is called "McCormick Toy Test" and it consists in measuring the level needed for the child to identify 71% the words in a closed-set list. A minimum of three reversals is needed to determine the outcome. The outcome is the level of speech in dB (A). Lower levels indicate better performance. The test will be performed in quiet and in noise, with and without the headset. The statistical analysis will depend on the characteristics of the data. Because the data are discrete, performing Analysis of Variance (ANOVA) will not be possible. Comparisons of thresholds obtained in different conditions will be done using non-parametric tests.

Outcomes will be measured in a single session. Four runs are required. The total testing time is around 12 minutes.

A dual task listening effort measure where the children are asked to pay attention to a screen and follow a simple instruction (press a button if they see a given shape and colour) while they also have to repeat words presented to them via loudspeakers. This paradigm has been used with children before. The listening effort outcomes are measured in terms of reaction time. A shorter reaction time indicates less effort. Differences in reaction time across conditions (without headset vs with headset) will be assessed performing ANOVA tests.

Outcomes will be measured in a single session. Four runs are required. The total testing time is around 12 minutes.

A short non-standardized questionnaire about acceptability of the device. The outcomes of the questionnaire will be qualitative data in order to detect patterns in the responses (i.e. dislike of the device, willingness to wear it publicly, etc.) and any patterns detected will be reported.

Inclusion Criteria: Children aged 6 to 11 years old. No developmental concerns No concerns about hearing loss. No concerns about language development. English dominant language. Exclusion Criteria: History, risk, or parent/carer/teacher concern of hearing loss. Developmental concerns Suspected language disability. Non-English dominant language.