Statistical Issues in sampling for compliance to exposure standards.
Symons-MJ; Flynn-MR; Rapport-SM; Truong-KN
Department of Biostatistics, School of Public Health, University of North Carolina, Chapel Hill, North Carolina 1996 Jan; :1-36
Statistical approaches for analysis of data from the limited number of samples collected by an industrial hygienist for checking compliance to an occupational standard were considered. Sampling for compliance usually has been guided by judgment selection, rather than true randomness, resulting in the creation of compliance samples which approximate a censored sample from the upper tail of the exposure distribution. A graphical based idea was developed, which appears useful with small numbers of compliance samples which are typically collected, less than ten. Assuming lognormality of the workplace exposures, considered as common with time weighted averages over hours of collection, and assuming the industrial hygienist has the ability to select the greatest and the least exposures, one is able to sketch a lognormal probability paper based inference scheme. Simple linear regression analysis provides an estimation of the median, the variance among jobs, and a subtotal of variance components. These variance components could include worker differences, variability in occasions from day to day sampling, and measurement error. Incorporating historical data will be necessary due to the few samples which are available from a compliance survey.
NIOSH-Grant; Grants-other; Workplace-studies; Industrial-hygiene; Sampling-methods; Statistical-analysis
Biostatistics University of North Carolina Chapel Hill, NC 27599-7400
Final Grant Report
NTIS Accession No.
Other Occupational Concerns; Grants-other
Department of Biostatistics, School of Public Health, University of North Carolina, Chapel Hill, North Carolina
University of North Carolina Chapel Hill, Chapel Hill, North Carolina