Abstract
Personal exposures to toxic airborne agents in the work environment often are approximately log-normally distributed. While the arithmetic mean of these distributions is independent of the averaging time of the measurement in a stationary environment, this is not the case for the variance. In general, the variance decreases with increasing averaging time, but the rate of decrease depends upon the autocorrelation structure of the exposure time series. If the autocorrelation structure can be determined, a knowledge of the parameters of the distribution of exposures at one averaging time allows the estimation of those for any other averaging time. Regardless of the autocorrelation situation, these results suggest that current ACGIH guidelines for assessing short-term exposures need examination and may require modification.