Estimating uncertainty in physical measurements, observational and environmental studies: lessons from trends in nuclear data

The authors analyze the time trends in five large nuclear data sets and describe a practical approach which utilizes these data in estimating the range of uncertainty in the individual physical measurements, observational and environmental studies and their meta-analytic syntheses. The time trends in the sequential series of measurements of the same physical quantity show a general pattern. The normal distribution grossly underestimates the probability of large deviations from the true values and the probability is well described by exponential functions. By analogy with physical measurements, the results indicate that the 95% confidence intervals in epidemiology and environmental studies should be expanded to account for unsuspected systematic errors