论文部分内容阅读
(一) 多样本差异比较按传统统计分析法是采用费歇氏的F测验法后,如F值显著则再采用学生氏t测验法来比较每两个样本差异的显著。这种方法既繁琐又带有一定程度的偏袒,即比较每对样本差异的显著性时只有同一个标准,而不是用不同标准来比较不同对样本的差异。从统计学来说,这就产生了偏袒,因为没有应用邓肯氏的方法来纠正这种偏袒。现在我们根据
(A) Comparison of multi-sample differences According to the traditional statistical analysis method is the use of Fisher’s F test method, such as the F value is significant then the Student’s t test to compare the significant difference between each two samples. This method is tedious and somewhat biased, ie, comparing the significance of each pair of samples with only one standard, rather than using different criteria to compare different sample pairs. From a statistical point of view, this gives rise to partiality because there is no application of Duncan’s method to correct such favoritism. Now we are based