Modelling the Effect of Personal Hearing Protection and Communications Devices on Speech Perception in Noise

http://cradpdf.drdc.gc.ca/PDFS/unc80/p530801.pdf


Christian Giguère, Ph.D.


A predictive model of speech intelligibility was developed and validated for use in
workplace environments characterized by hazardous noise levels requiring the use of
hearing protection devices. The model takes into account the characteristics of the noise field, the signal-to-noise ratio, the attenuation of the device, and the hearing status of the listener. Accurate prediction of speech intelligibility required consideration of both the audibility (threshold) and distortion (suprathreshold) components of a hearing loss, as measured by the pure-tone audiogram and the clinical Hearing-In-Noise Test. The mean prediction error over all conditions tested was -0.1% with a standard deviation of 14.6%. The model could be applied in the context of two important applications of direct relevance to the Canadian Forces military environment: (1) the optimal selection of hearing protectors, and (2) the establishment of functionally-based hearing standards for
the personnel. The predictive procedures developed in this research could form the basis for a comprehensive hearing protector selection and speech communication assessment system. Additional work is needed to extend the speech perception model to headsets and other communication devices. Speech production with hearing protection devices should also be further studied.