Response times to speech stimuli as measures of benefit from amplification

The benefits of management of hearing disability, in particular by provision of a hearing aid, are traditionally assessed by the percentage improvement in performance on a speech identification task. To provide precise and stable results, such procedures require more time than is available in most clinical settings. In any stressed performance, e.g. an impaired individual trying to listen in noise, there is a trading relationship between accuracy and effort (the cost at which accuracy is achieved). If the control of performance naturally spends effort to stabilize high performance, then benefit from amplification may essentially comprise and be measurable as reduction in effort rather than improvement in accuracy. Certainly complaints of hearing disability emphasize fatigue from careful listening. Hence a hearing aid may not only enable hearing impaired persons to hear more of speech but may enable them to hear it more easily, thus reflecting a second dimension to disability and benefit.Ease of listening was investigated using auditory response times to speech stimuli of two levels of structure: single words and sentences. The speech material was presented to 44 experienced hearing aid users (mild to moderate sensorineural hearing impairment). The speech was presented both unaided and aided at presentation levels of 60, 70 and 80 dB SPL and signal-to-noise ratios of quiet and + 5 dB. Response times were taken to the tokens within each list that were correctly identified. Benefit is defined as the decrease in response time from the unaided to the aided condition. This index of benefit shows changes that are substantially greater in relative terms than the conventional change in per cent (%) correct identification. In addition, the otherwise problematic data subsets where the unaided performance is greater than 85% or where the aided/unaided difference lies within ±6% show significant changes on the response time measure. These changes in response time to speech stimuli are proportionately greater than the observed changes in response time to speech-shaped noise and can therefore be attributed to more easy perceptual processing rather than to a simple increase in loudness.