〓酒について (第2報) : 優劣判定に関する推計學的考察

概要

論文の詳細を見る
It is, sometimes, not significant but also reasonable that the samples in the sensorial test for Sake are distinguished only by the total of orders given by judges. So the present sensorial test for Sake must have been objectively treated on the basis of statistical methods and aoso the variation in the test must have been positively adjusted. In this report, the differences between samples in the sensorial test for Sake are statistically distinguished by using the order statistics. The method of the test for signiflcant difference between two samples is as follows : [numerical formula] t : the total of differences between orders of two samples by each judge. u^2 : the unbiased estimate of variance of the differences between orders of two samples by each judge. n : the number of judges. And the values of the 5% level in the F-distribution table, of which the degrees of freedom are n_1=1 and n_2=n-1,are expressed as follows : F^1_<n-1>(0.05) And F_0>F^1_<n-1>(0.05) The two samples are said to be distinguished by the orders of judges. F_0⩽F^1_<n-1>(0.05) The two samples are not said to be so clearly distinguished.
1953-08-15