论文部分内容阅读
1971年,Ralph B.D’Agostino 提出使用 D检验法进行正态性检验,由于原法系用原始数据作计算,当样本含量较大时,甚为麻烦。其后国内学者提出了用频数表计算 D 值的改进方法,大大简化了大样本作 D 检验的手续。因此近年来,国内广泛地使用了该法。但我们对十份实际测量材料(共1,239人)同时用动差法(矩法)和 D 检验法进行(指用频数表计算)正态性检验,发现在动差法判定为非正态资料时,D 检验的结果相同,而判为正态检验时则D 法反而判为非正态。仅仅根据以上资料下此
In 1971, Ralph B. D’Agostino proposed using the D-test to perform normality tests. Because the original law was calculated using raw data, it was troublesome when the sample content was large. Later, domestic scholars proposed an improved method for calculating the D value using a frequency table, which greatly simplifies the procedure for large-scale D-tests. Therefore, in recent years, the law has been widely used in China. However, we tested the normality of ten actual measurement materials (1,239 persons in total) by using the kinematic method (moment method) and the D-test (referring to the frequency table calculation) and found that it was determined as non-normal data by the kinematic method. At the same time, the result of the D-test is the same, but when the test is judged as normal, the D-method is judged as non-normal. Just based on the above information