8.12: 独立性检验的假设检验

Hypothesis Test for Test of Independence
JoVE Core
Statistics
A subscription to JoVE is required to view this content.  Sign in or start your free trial.
JoVE Core Statistics
Hypothesis Test for Test of Independence
Please note that all translations are automatically generated. Click here for the English version.

3,598 Views

01:16 min
April 30, 2023

Overview

独立性检验是一种基于卡方的检验,用于确定两个变量或因子是独立还是相关。此假设检验用于检查变量的独立性。可以根据列联表中的变量构建两个定性调查问题或实验。目标是查看这两个变量是不相关 (independent) 还是相关 (dependent)。此检验的 null 假设和备择假设为:

H0:两个变量(因子)是独立的。

H1:两个变量(因子)是相关的

首先,确定观察到的频率并计算预期的频率。每个条目的预期频率是通过将行总计和列总计相乘并除以所有频率之和来获得的。然后,使用列联表中观察到的频率值和计算的预期频率来计算检验统计量。然后,在卡方表的帮助下,计算出具有适当置信度的单尾检验中的临界值。如果检验统计量大于临界值并落在临界区域,则拒绝原假设;否则,它将被接受。

本文改编自 Openstax, 统计学导论, 第 11.5 节, 卡方检验的比较。

Transcript

考虑一个关于酒精消费和事故死亡率的数据集。执行假设检验以确定两个变量是否独立。换句话说,饮酒与较高的事故死亡率之间是否存在关系?

原假设指出饮酒和事故死亡是独立的事件,而备择假设则相反。

行总计和列总计的乘积除以所有频率的总和,即可得出每个表条目的预期频率。

使用预期值和观测值,计算卡方检验统计量。

接下来,在卡方表的帮助下,确定用一个自由度在右尾部分隔 0.05 面积的临界值。

由于检验统计量大于临界值并且位于临界区域内,因此拒绝原假设 – 即饮酒量与道路事故死亡率之间没有关系。

因此,在 5% 的显著性水平上,有足够的证据得出结论,饮酒和事故死亡是因变量。

Key Terms and definitions​

  • Test of Independence - A chi-square-based test to check if two variables are independent or dependent.
  • Null Hypothesis - The proposition that suggests no statistical relationship in a set of observed data.
  • Alternative Hypothesis - The proposition that suggests a statistical relationship exists in the data.
  • Test Statistic - A metric to decide if a null hypothesis should be rejected or not.
  • Chi-Square Test for Independence - A hypothesis test to determine the independence of two qualitative variables.

Learning Objectives

  • Define Test of Independence - Explain when and how it is used (e.g., hypothesis testing).
  • Contrast Null Hypothesis vs Alternative Hypothesis - Explain their role in a chi-square test for independence (e.g., hypothesis formation).
  • Explore how Test Statistics influence hypothesis testing decisions (e.g., chi-square test).
  • Explain the significance of Chi-Square Test for Independence in statistics.
  • Apply these concepts in analyzing contingency tables and testing for variable independence.

Questions that this video will help you answer

  • [Question 1] How is a Test of Independence conducted using Chi-Square Test?
  • [Question 2] What is the role of Null and Alternative Hypothesis in this test?
  • [Question 3] How does the Test Statistic influence the decision of a Hypothesis Test?

This video is also useful for

  • Students - Understand how chi-square test of independence aids in learning hypothesis testing.
  • Educators - Provides a clear framework for teaching hypothesis testing and inference.
  • Researchers - Crucial in scientific study, testing relationships between variables.
  • Statistics Enthusiasts - Offers valuable insights into chi-square testing and statistical hypothesis testing.