用户工具

站点工具


假设检验的基本逻辑

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
假设检验的基本逻辑 [2023/03/13 09:51] wisture假设检验的基本逻辑 [2024/03/19 02:32] (当前版本) 李佳熠2
行 1: 行 1:
-===== 假设检验  =====+===== 假设检验 (Hypothesis testing)  =====
   * 定义:用样本数据来评价有关总体的某一假设的可置信性的推论程序。以样本为依据,对总体作出判断   * 定义:用样本数据来评价有关总体的某一假设的可置信性的推论程序。以样本为依据,对总体作出判断
 +  * Definition: A process of inference using sample data to evaluate the confidence of a hypothesis about a population. Based on the sample, the judgment is made on the population
   * 可置信性:在多大程度上可以相信这个假设是正确的   * 可置信性:在多大程度上可以相信这个假设是正确的
 +  * Confidence: The extent to which the hypothesis can be believed to be correct
   * 程度的判断基于样本均值的概率分布   * 程度的判断基于样本均值的概率分布
 +  * The degree is judged based on the probability distribution of the sample mean
   * 研究设计的两种结果:   * 研究设计的两种结果:
 +  * Two results of research design:
       - 研究假设:样本均值与总体均值间有显著差异       - 研究假设:样本均值与总体均值间有显著差异
 +      - Research hypothesis(H1): There is a significant difference between the sample mean and the population mean
       - 虚无假设:样本均值与总体均值间无显著差异       - 虚无假设:样本均值与总体均值间无显著差异
 +      - Null hypothesis(H0): There has no significant difference between the sample mean and the population mean
   * 显著性水平(α):在检验过程中,我们假设虚无假设是真实的,同时计算出观测到的差异完全是由随机误差所致的概率,称为观测概率p。若p≤α,则拒绝虚无假设;如果p>α,则不能拒绝虚无假设   * 显著性水平(α):在检验过程中,我们假设虚无假设是真实的,同时计算出观测到的差异完全是由随机误差所致的概率,称为观测概率p。若p≤α,则拒绝虚无假设;如果p>α,则不能拒绝虚无假设
 +  * Significance level (α): During the test, we assume that the null hypothesis is true, and the difference we observed is entirely due to random error, which is called the observed probability p. If p ≤α, the null hypothesis is rejected, and if p is >α, the null hypothesis cannot be rejected.
 ---- ----
-===== 假设检验的基本逻辑  ===== +===== 假设检验的基本逻辑(The basic logics of hypothesis testing)  ===== 
-  陈述假设:虚无假设与研究假设 +  1.陈述假设:虚无假设与研究假设 
-  设定决策标准:找出z分数界定的临界区域 +  Describe the hypothesis: The null hypothesis(H0) and the research hypothesis(H1) 
-  - 收集数据,计算样本统计量 +  2.设定决策标准:找出z分数界定的临界区域 
-  做出决策:如果得出的z分数落在拒绝区域中,就拒绝H0+  Set the decision standard: Find out the critical value by z-score 
 +  3.收集数据,计算样本统计量 
 +  Collect the data and calculate the sample statistics 
 +  4.做出决策:如果得出的z分数落在拒绝区域中,就拒绝H0 
 +  Decide: If the z-score falls in the reject area, reject H0.
假设检验的基本逻辑.1678701069.txt.gz · 最后更改: 2023/03/13 09:51 由 wisture