Bayes Factor
Updated 2026-08-17
INTRODUCTION
English translation pending.
CORE DEFINITION
The Bayes factor is the ratio of the probability of the data under one hypothesis to its probability under another. It expresses evidence as a strength rather than a verdict: a factor of ten means the data are ten times more likely under the first hypothesis, and a factor near one means the data cannot distinguish them at all. Unlike a p-value, it compares the hypotheses directly, accumulates across experiments, and never depends on a fixed sample size, which is why it is often proposed as the replacement for threshold-based significance testing.
SCAFFOLDING EFFECT
Reduce cognitive load
- Ask what the data favor: compare two hypotheses directly instead of testing one against a null. - Accumulate the evidence: multiply factors across replications rather than restarting each time. - Report the strength: give the factor and its interpretation, not just a pass or a fail.
Anchor fast decisions
A p-value answers a question about the data given one hypothesis, while the question people actually want answered is which hypothesis the data support. The Bayes factor addresses that directly, by taking the ratio of how likely the data are under each contender, so the same number carries the direction and the strength of the evidence. Because ratios multiply, evidence from separate studies composes into a running total, which is why the framework accumulates knowledge where threshold testing resets it with each experiment.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- en.wikipedia.orghttps://en.wikipedia.org/wiki/Bayes_factorverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS