用户工具

站点工具


事后检验

差别

这里会显示出您选择的修订版和当前版本之间的差别。

到此差别页面的链接

两侧同时换到之前的修订记录前一修订版
后一修订版
前一修订版
事后检验 [2024/03/29 05:33] hant_g._cavendish事后检验 [2024/04/06 09:20] (当前版本) caomingsu
行 3: 行 3:
      * The ANOVA cannot indicate which alternative hypothesis is supported.      * The ANOVA cannot indicate which alternative hypothesis is supported.
   - 要进行事后检验以说明各组差异在什么地方;   - 要进行事后检验以说明各组差异在什么地方;
-     * Post hoc tests are needed to show where the groups differ.+     * Post-hoc tests are needed to show where the groups differ.
   - 事后检验是比较每一个组和另一个处理组,一次比较两个,这称为成对比较。   - 事后检验是比较每一个组和另一个处理组,一次比较两个,这称为成对比较。
      * Post-hoc test compares each group to another treatment group, two at a time; this is called a pairwise comparison.      * Post-hoc test compares each group to another treatment group, two at a time; this is called a pairwise comparison.
行 9: 行 9:
 ---- ----
 === 族系(familywise)误差 === === 族系(familywise)误差 ===
-  每一个比较都是一个单独的假设检验,每一个都有犯I类错误的风险。所以,比较对数越多,作结论的风险越大。即容易发现实际不存在的差异。这称为实验导致的(experimentwise)alpha水平或族系(familywise)误差 +  每一个比较都是一个单独的假设检验,每一个都有犯I类错误的风险。所以,比较对数越多,作结论的风险越大。即容易发现实际不存在的差异。这称为实验导致的(experimentwise)alpha水平或族系(familywise)误差 
-  * 多次比较后犯I类错误的概率为{{::多次比较后一类错误概率.png?200|}},其中c=比较次数+     Each comparison is a separate hypothesis test, each with the risk of making a Type I error. So the more pairs of comparisons there are, the greater the risk of making a conclusion. That is, it is easy to find differences that do not actually exist. This is called the **experimentwise** alpha level or **familywise** error. 
 +  - 多次比较后犯I类错误的概率(Probability of making a Type I error after multiple comparisons)**α<sub>EW</sub>=1-(1-α)<sup>c</sup>**,其中c比较次数(c=Number of comparisons)。
  
 ---- ----
-=== Tukey's HSD检验 === +=== Tukey's HSD检验 (Tukey's HSD test) === 
-  此检验各组要有相同的样本容量 +  此检验各组要有相同的样本容量; 
-  * 可以计算出单一的值确定处理均值间的最小差异,考查此差异在统计上是否显著 +    This test requires the same sample size for each group. 
-  * {{::hsd公式.png?200|}} +  - 可以计算出单一的值确定处理均值间的最小差异,考查此差异在统计上是否显著; 
-  q可以查表得到+    A single value can be calculated to determine the smallest difference between treatment means, to see if the difference is statistically significant. 
 +  - {{::hsd公式.png?200|}} 
 +  q可以查表得到。 
 +    * q score can be obtained by looking up the table.
  
 ---- ----
-=== Scheffe检验 === +=== Scheffe检验 (Scheffe test) === 
-  适用于n不相等的情况 +  适用于n不相等的情况; 
-  * 降低I类错误的风险,提高II类错误的风险 +    Applies when n is not equal. 
- +  - 降低I类错误的风险,提高II类错误的风险 
 +    * Reduces the risk of Type I errors and increases the risk of Type II errors.
事后检验.1711690414.txt.gz · 最后更改: 2024/03/29 05:33 由 hant_g._cavendish