Problems associated with combining intersubject variabilities in hearing protector device (HPD) attenuation and susceptibility to noise induced hearing loss were considered. Variability in HPD attenuation was generally corrected for by using the "assumed" attenuation, an adjusted attenuation value. Arguments were set out for the need to establish the protection level at a set criterion, which should be met whether or not HPDs were used. In developing the arguments, aspects that were discussed included sources of variability in hearing protector attenuation (within subject variability and between subject variability), quality of instruction received, experience, and feedback. In addition, there were variabilities across different laboratories, which could be included with between subject variabilities. Using assumed values for the two types of variabilities, correction factors were worked out. A practical example using the attenuation of the Willson EP100 ear plug was conducted. According to the calculations, it was necessary to reduce the mean attenuation value by 3.2 decibels (dB) for frequencies below 1000 Hertz (Hz) and by 1.0dB for frequencies above it, for within subject variability. For between subject variability, these values needed to be reduced by 5.8dB for frequencies less than 1000Hz, and by 5.1dB for frequencies above it, at 90% protection level, and by 7.5dB and 6.6dB, respectively, at 95% protection level. This amounted to a reduction of about one standard deviation at 95%, and somewhat less than this at 90% protection level.
Proceedings, 1992 Hearing Conservation Conference, April 1-4, 1992, Lexington, Kentucky, Office of Engineering Services, University of Kentucky and NIOSH