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。
===== 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
  * 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
  * 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
  * The degree is judged based on the probability distribution of the sample mean
  * 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
      - 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
      - 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
  * 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（The basic logics of hypothesis testing）  =====
  1.陈述假设：虚无假设与研究假设
  Describe the hypothesis: The null hypothesis(H0) and the research hypothesis(H1)
  2.设定决策标准：找出z分数界定的临界区域
  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.