ABSTRACT
Reproducibility of repetitive counting of the same sample by all liquid-scintillation methods is determined by the statistics of counting. Because there is no problem of sample geometry reproducibility in most liquid-scintillation detectors, the reproducibility of counts holds even when a sample is removed from the detector and replaced between counts. The reproducibility of a series of samples prepared separately and counted by liquid scintillation methods will depend primarily on the reproducibility of the chemical and volumetric steps used in the sample preparation if the count taken is large enough to give good counting statistics. The absolute accuracy of an alpha count can only be approached, since radioactive decay is random with time. The beta/alpha hybrid instruments can be adjusted so that the electronic-pulse separation rejects nearly all of the ambient background and primarily alpha is counted. In all cases, the desired accuracy of counting depends on the statistics of counting and determines the total count that must be accumulated.
