Student Research: Jason Woodruff
Accurate predictions of full shift noise exposures in the construction trades would allow employers to determine which employees should be enrolled in a hearing conservation program and use hearing protection before they are exposed. Previously collected task-based noise exposure data were used to generate task-based full shift noise exposure predictions for construction workers in three trades by combining task based exposure levels with predicted task durations or reported task durations for a particular work shift. The workers wore noise dosimeters which yielded time weighted averages for the shift for which task-based predictions were made. Workers were unable to predict their tasks or task durations accurately. Using the reported tasks the task-based noise exposure predicted distributions were biased low and could not account for high (8-hour time weighted average greater than 95 dBA) exposures. The predicted noise exposure distributions were less variable than the actual exposures. The task-based exposure assessment was also unable to stratify workers accurately into groups of overexposed and not overexposed based on occupational exposure limits. The construction sites that were sampled for the previously collected task based exposure data were quieter on average than the sites where the task-base predictions were made, thus accounting for the apparent bias in the predicted exposure distributions. Using a worker’s trade and task as the only predictor of noise exposure in construction is not an accurate way to estimate daily risk or assign a worker into a hearing conservation program or to use hearing protection for a particular shift. The observed exposures were too variable to be predicted on this basis and additional variables are needed in the model to generate accurate predictions.