liveBook · Manning

Evidence

As explained in Timeless Algorithms, The Foundational Ideas

Evidence is information or observations considered in support of a claim or conclusion.

Evidence is information relevant to an unresolved question. It can narrow uncertainty by making some hypotheses more plausible, but it does not guarantee a conclusion. In Bayesian reasoning, prior probabilities combine with the likelihood of observed evidence to produce posterior probabilities. section 2.1 section 13.4.1

In hypothesis testing, evidence is assessed to decide whether it is sufficiently inconsistent with the null hypothesis to reject it. Failure to reject does not prove the null true, and rejection does not prove the alternative true. Conclusions should describe what the data supports, what remains uncertain, and the limits of the analysis. section 4.1.7 section 4.3

A positive medical test

When a disease affects one person in 1,000, a test with 99% sensitivity and a 1% false-positive rate raises the probability of disease after a positive result to about 9%. The result is informative, not conclusive; further evidence is needed for a definitive diagnosis. section 10.2.1

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Explains Bayesian updating and hypothesis-test conclusions while stopping short of treating either outcome as certainty.
Examples
Uses medical diagnosis and testing, plus a courtroom acquittal analogy, to illustrate evidential limits.
What it assumes
Builds Bayesian reasoning on probability rules and treats test assumptions and data quality as limits on conclusions.
What you can do after
Evaluate conclusions in proportion to evidence, state uncertainty, and avoid claiming more than a test establishes.
How it connects
Connects evidence to Bayesian inference, probability, hypothesis testing, statistical inference, and machine learning.

Also explained in

Compare how 8 books explain evidence →

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.