liveBook · Manning

Likelihood ratio test

As explained in Timeless Algorithms, The Foundational Ideas

A likelihood ratio test compares nested statistical models using their maximized likelihoods to assess whether the more complex model provides a significant improvement.

A likelihood ratio test (LRT) compares nested models by their maximized likelihoods. It asks whether a more complex model improves likelihood enough to justify its added parameters; under standard regularity conditions, the comparison yields a test statistic with an approximate chi-square distribution. section 3.7.1

Adding predictors to a regression

Compare a linear regression with five predictors to one with those same predictors plus two more. A modest increase in maximized likelihood favors the simpler model; a sufficiently large increase suggests the added parameters improve fit beyond chance. section 3.7.1

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Explains maximized-likelihood comparison and approximate chi-square assessment under standard regularity conditions, without detailing the statistic’s derivation.
Examples
Uses a five-predictor linear regression compared with a nested model adding two predictors.
What it assumes
Requires fitting nested models and comparing their maximized likelihoods.
What you can do after
Evaluate whether added model parameters improve fit enough to be unlikely due to chance alone.
How it connects
Connects maximum likelihood, nested models, linear regression, and the chi-square distribution.

Related concepts

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.