Accounting for bias and uncertainty resulting from dose measurement errors and other factors

Analyses of data from epidemiological studies of persons exposed to radiation have usually been based on the assumption that doses are measured without error. This paper begins with a general discussion of measurement error, followed by discussion of approaches for handling both random and systematic error. However, dose measurement error is only one of many sources of uncertainty that have not usually been fully accounted for in epidemiological analyses. At the end of the paper, brief consideration is given to the general problem of quantifying uncertainties from several different sources. (Author)