Prediction of 8-hour noise dose from brief duration samples

The purpose of this study was to determine whether brief duration samples of noise dose measurements could predict the noise exposure levels of employees as measured by 8-hour dosimetry. Sampled noise dose measurements were obtained for 1, 5, 10, 15, and 20 minutes and 8 hours for four workers at each of four industrial sites. Statistical analyses indicated no significant differences or interactions among trials and sample durations. A simple linear prediction model provided the best fit for the four shortest of the five abbreviated sample durations, whereas an exponential regression model provided the best predictions using the 20-minute sample duration. The high variance and low correlations across all measures permitted only poor predictions from these abbreviated sample data. Based on these results, sampled noise measurements of short duration (1 to 20 minutes) do not appear to be a valid means for predicting the amount of noise exposure during an 8-hour shift.