GENERAL PROCEDURES FOR RADIOASSAY 



173 



To express the deviation in terms of coiiiiting rate, the scjuare root of the 

 total counts is divided by time. Thus if 1000 counts were measured in 

 5 min, the deviation would be \/l000/5 = 6 counts/min, and the count- 

 ing rate would be expressed as 200 ± 6 counts/min. This means that 

 there is a 68.3 per cent probability that the counting-rate error is less 

 than 6 counts/min in this sample. Other error terms may be used. 

 However, in this text the standard deviation will be employed unless 

 otherwise noted. 



The counting rate when there is no radioactive source present is called 

 the background and is due to cosmic radiation and radioactivity in the air 

 and surrounding materials. The counting rate of a sample is eciual to the 

 observed rate minus the background rate. When the sample count is 

 large compared with the background, the standard deviation can be esti- 

 mated with little inaccuracy by taking the square root of the total counts 

 as indicated. However, when the background and sample counts are not 

 greatly different, it becomes necessary to take into account the deviation 

 in the background as well as in the sample count. This can be done by 

 use of the following formula: 



Standard deviation of sample counting rate = 





H 



(5-1) 



where Cs = total counts for sample in time Ts 



Cb = total counts for background in time Tb 

 A typical illustration is presented in Table 5-1. In this case it will be 

 noted that the standard deviation as calculated from the sample plus 

 background was 3.5/60 X 100 = 5.8 per cent, whereas the actual stand- 

 ard deviation was fio X 100 = 10 per cent. With weak samples the 

 background contribution to the counting error becomes of increasing 

 importance. In practice this may be overcome somewhat by the reduc- 

 tion of background by shielding or other methods, and bj^ utihzing spare 

 time to give long background counting periods. Of course, the weak 

 samples will also have to be counted for long periods of time. 



Table 5-1. Calculation of Standard Deviation for Sample Containing 



Low Activity 



