Comparison of hearing protector ratings for statistically significant differences

During the development of the American National Standards Institute S12.68 standard for estimating effective A-weighted sound pressure levels when hearing protector device (HPD) are worn, the need to provide a mechanism to compare HPD ratings was identified. The current Noise Reduction Rating (NRR) is a single number that describes the potential of a product to reduce noise at the user’s ears. The Noise Reduction Statistic for A-weighting (NRSA) describes a range of protection performance and estimates the uncertainty associated with the rating. The uncertainty for NRSA is estimated using a bootstrap procedure. This method can be applied to current ratings and comparisons between rating evaluations from different laboratories. This paper will discuss NRR and NRSA error bars ratings and present a modification of the bootstrap method to estimate directly the statistical significance of differences. Results from two different studies of hearing protector attenuation measurements and training protocols will be examined. The studies’ results will also be compared against the manufacturers’ published attenuation values.