==== 事后检验（Post hoc tests） ====
ANOVA得到一个两点决策，但并未知道差异存在在哪些组之间。这时我们需要进行事后检验，这使得我们能够比较各个组，找到差异存在在哪些组之间。事后检验就是比较每一个处理组和另一个处理组，一次比较两个，称为成对比较。

ANOVA obtains a two-point decision, but does not know the differences are in which groups. This is where we need to do a post hoc test, which allows us to compare groups and find out where the differences lie. A post hoc test is a comparison of each treatment group with another treatment group, two at a time, called a pairwise comparison.

----
=== 重复测量方差的Tukey's HSD检验 (Tukey's HSD test for repeated measurements of variance)  ===
重复测量方差HSD值的计算和单因素方差分析的HSD值计算有所不同。

The calculation of the HSD value of repeated measurement variance is different from that of the HSD value of one-way ANOVA.

重复测量方差HSD值的计算如下：
{{ ::hsd.jpg?nolink&300 |}}

How the HSD value of repeated measurement varience calculated is listed above.

公式中用误差的和方，自由度，均方代替原来分母项的组内和方，自由度，均方。
其中q值可以从表中查出，需要用到K（处理数），df<sub>误差</sub>以及显著性水平α。

In the formula, the sum of the error, the degree of freedom, and the mean square replace the original denominator term in the sum of the group, the degree of freedom, and the mean square.
The q value can be found in the table, using K (processing number), df<sub>error</sub>, and significance level α.

得到HSD值之后，将样本两两均值之差与HSD值作比较，大于HSD值则认为这两组之间有显著差异。

After the HSD value is obtained, compare the difference between the pair means of the sample with the HSD value, and if the difference is greater than the HSD value, it is considered that there is a significant difference between the two groups.