Similar to testing one population mean (as in the previous chapter), there are methods to test equality of two population means or more than two population means. For more than two population means on the same outcome/response variable, one-way analysis of variance method is used. However, for two population means there are two situations, and they require different tests. The test depends on if the two samples are dependent or not. If the two samples are dependent, then a paired sample t-test is appropriate. Often the two dependent samples are also known as paired samples. If the two samples are not dependent or paired, then they are two independent samples. In case of two independent samples, we use independent samples t-test. Although the t test is used for both independent an dependent samples, the formulation of the two t statistics are very different.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Hypothesis Tests for Two and More Means

  • Shahjahan Khan,
  • Md. Shafiur Rahman

摘要

Similar to testing one population mean (as in the previous chapter), there are methods to test equality of two population means or more than two population means. For more than two population means on the same outcome/response variable, one-way analysis of variance method is used. However, for two population means there are two situations, and they require different tests. The test depends on if the two samples are dependent or not. If the two samples are dependent, then a paired sample t-test is appropriate. Often the two dependent samples are also known as paired samples. If the two samples are not dependent or paired, then they are two independent samples. In case of two independent samples, we use independent samples t-test. Although the t test is used for both independent an dependent samples, the formulation of the two t statistics are very different.