← Back to Glossary

Likelihood Ratio

Likelihood ratio measures how strongly evidence shifts odds. Values above 1 increase odds, below 1 decrease odds.

Definition

A likelihood ratio describes evidence strength. It tells you how much to multiply prior odds to get posterior odds.

Why it matters

Likelihood ratio is the clean input for Bayes updates. It separates evidence strength from base rates.

Common pitfalls

Mixing with odds ratio: Likelihood ratio is about evidence, not baseline odds.

Using LR without context: LR depends on how evidence is defined and measured.