For testing one simple hypothesis against another, the Neyman–Pearson lemma establishes that the likelihood-ratio test is most powerful among procedures controlling Type I error at a chosen significance level. It gives a principled basis for selecting a test: favor the valid procedure best able to detect a real effect, and fix the decision rule before analyzing the data. section 4.1.3
Neyman–Pearson lemma
The Neyman–Pearson lemma states that a likelihood-ratio rejection rule is most powerful for testing a simple null hypothesis against a simple alternative at a specified significance level.
How Timeless Algorithms, The Foundational Ideas covers it
- How deep it goes
- States the result for two simple hypotheses and explains its power rationale, without deriving the likelihood-ratio rule.
- What it assumes
- Uses hypotheses, significance levels, Type I error, and power without defining them in this treatment.
- What you can do after
- Understand why likelihood-ratio tests are preferred for distinguishing two simple hypotheses under controlled Type I error.
- How it connects
- Connects likelihood-ratio tests with statistical power, Type I error, significance levels, and hypothesis testing.
This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.
