Rapid methods for ascertaining whether the activity of a weak radioactive sample exceeds a predetermined level

The paper considers methods of determining whether the mean rate of occurrence of random events, such as radioactive disintegrations, is greater or less than some given tolerance rate. The most efficient method is one which makes this determination in the shortest time and with an acceptably low probability of error. Sequential test procedures minimize the sample size by observing the random process until some terminating condition is satisfied. The performance of several sequential tests is investigated, with particular reference to conditions encountered in monitoring radioactive contamination on the hands. The most efficient of these tests is shown to be one which measures the difference between the number of events actually observed and the number expected at the tolerance rate. Under conditions likely to be encountered in routine monitoring for radioactivity on the hands this method gives a fivefold saving in time over the method now in use.