PsychStatsWiki

本页面只读。您可以查看源文件，但不能更改它。如果您觉得这是系统错误，请联系管理员。
====符号检验法（Sign test）====
  *符号检验法对应于相关样本的t检验，根据两个相关样本的每一对数据之差的符号（正负）来进行检验。
  *符号检验法可用于顺序型与类目型的数据。

* The symbolic test method corresponds to the t test of the relevant samples, and is tested according to the sign (positive or negative) of the difference between each pair of data of the two relevant samples.
* Symbolic testing can be used for sequential and categorical data.
----

**检验步骤（checking procedure）**
  *1.提出假设（H0：p（x>y）=0.5/差值d总体的中位数为0）；
  *2.根据H0所提供的前提条件确立统计模型
  *3.规定显著性水平α，随之确定否定域；
  *4.计算检验统计量的值；
  *5.做出决策。

*Hypothesis (H0: p (x>y) =0.5/ The median of the difference d population is 0);
*2. Establish the statistical model according to the preconditions provided by H0
*3. Specify the significance level α, and then determine the negative domain;
*4. Calculate the value of the test statistic;
*5. Make decisions.
----

**大样本情况下的符号检验法（Symbolic test method in case of large sample）**
  *大样本（n>25）情况下，可以用与二项分布近似的正态分布进行符号检验。
* For large samples (n>25), the sign test can be performed with a normal distribution approximating the binomial distribution.