A task-based statistical model of a worker's exposure distribution: Part I--Description of the model.

The authors present a task-based model to describe a single worker's exposures to a single airborne chemical toxicant. The model accounts for variability in short-term time-weighted average (TWA) exposure values within a task, and for variability in arithmetic mean exposure levels between tasks. For a given workday, the 8-hour TWA value is equated with the sample mean of an appropriate number of short-term TWAs arising from stratified random sampling of short-term TWAs with proportional allocation by task. The model accounts for autocorrelation in the stochastic process that generates successive short-term TWA values. Due to the underlying random process, a given type of workday with regard to the set of task times has an associated distribution of 8-hour TWA values; the variance of this distribution increases with increasing autocorrelation in the time series of short-term TWAs. A worker's total distribution of 8-hour TWAs is a mixture of these day-specific distributions weighted by the relative frequency of each type of workday; the variance of the total distribution increases with greater day-to-day variability in the array of task times.