1977 Volume 27 Issue 4 Pages 380-383
The precision (relative value) is commonly used in determining the sample size which should be observed to establish the time standard. In this paper, the relationship between the sample size and the precision is examined, on the assumption that the population is normally distributed with unknown variance. Comparing the precision of Niebel's method with the proposed method, it becomes apparent that Niebel provides smaller precision value, depending on the initial sample size, than the required one. On the contrary, the proposed method presents much the same as the required one.