A family of statistical distributions for modelling occupational radiation doses in low dose occupations

New statistical distributions have been defined to describe occupational exposures to ionising radiation. These distributions are particularly useful in modelling occupations where most doses are low. The maximum likelihood method was used for parameter estimation and has been adapted to allow doses that are recorded as zero to be included in the calculations. The method can then be applied to estimate true doses from the complete set of recorded dose values when the a priori dose distribution and the dose measurement distributions have been derived previously. This application is important in epidemiological cohort studies where it can improve the accuracy of excess relative risk estimates. (authors)