An investigation of the dependence of exposure variability on the interval between measurements.
Ann Occup Hyg 1994 Aug; 38(4):361-372
The variability which is found in measurements of airborne contaminants as the interval between data collections increases was investigated. Shift long measurements were taken for 53 workers exposed to on the job contaminants for five different data sets in 123 time series. A clearly increasing trend in the variance was noted with the time interval over which data were gathered. However, this was true for only two of the five data sets. Further analysis indicated that this finding was confined to 30 of the workers who contributed data and in 29% of the total number of time series analyzed. The combination of autocorrelation and nonstationary behavior explained the increase in 64% of the time series among the data where the trend was found. For a small group of workers in only one of the data sets and for a minority of cases among workers who contributed more than one time series to the analysis, there were findings of significant autocorrelation. The authors conclude that autocorrelation of shift long exposures did not seem to be pervasive and was therefore unlikely to present significant problems when implementing statistically based sampling strategies. They note that the issue of nonstationarity is equivocal. The time investigated was short, about 30 days, and the authors express concern as to whether the problem is more pronounced when longer time periods are used.
NIOSH-Publication; NIOSH-Cooperative-Agreement; NIOSH-Grant; Training; Air-quality-monitoring; Air-sampling; Air-sampling-techniques; Statistical-analysis; Industrial-hygiene
Environmental Sciences & Engr University of North Carolina 311 Pittsboro St Chapel Hill, NC 27599-7410
Annals of Occupational Hygiene
University of North Carolina Chapel Hill, Chapel Hill, North Carolina