Statistical criteria to establish bioassay programs

A statistical study of fitting log-normal distributions to air sampling concentrations and random intakes is reported in this paper. An improved method on the approximation of a sum of log-normal distributions to a log-normal distribution is used to forecast bioassay results and their uncertainties. These methods can be used in the design and implementation of air control monitoring and bioassay programs for introducing statistical criteria to determine when a bioassay is required as well as the frequency.