liveBook · Manning

Frequentist paradigm

As explained in Grokking Bayes

Frequentist probability interprets probability in terms of long-run event frequencies under repeated trials.

The frequentist paradigm interprets probability as the long-run frequency of an event across infinitely many repeatable trials. An unknown quantity is treated as fixed; randomness enters through the process that generates observed data. section 1.6

This viewpoint makes probability questions depend on repeatability. A frequentist may reject “What is the probability it rains today?” because today cannot be repeated, and instead calculate how often comparable observations produce rain. Frequentist methods are computationally lightweight, ready to use, and require fewer diagnostics than Bayesian models. section 1.6

The paradigm is commonly associated with:

  • Fixed unknown quantities whose observed data vary through chance. section 1.6
  • Established procedures such as classical hypothesis testing and p-values. section 11.1 section 1.6
  • Large-dataset applications that do not require constructing a prior distribution. section 1.6

Weather as a boundary case

A Bayesian analysis can assign a probability of rain using humidity and cloud cover as evidence. A frequentist analysis instead asks about rain frequencies across repeatable observations or trials. section 1.6

Compare situations favoring frequentist methods, including abundant data, no required prior knowledge, and established objectives such as A/B testing. section 1.6

How Grokking Bayes covers it

How deep it goes
Defines long-run frequency, fixed unknowns, random data generation, and p-values without developing their calculations.
Examples
Uses weather probabilities, infinitely repeated trials, classical hypothesis testing, p-values, and A/B testing.
What you can do after
Evaluate probability questions through repeatability and recognize when frequentist testing methods fit the objective.
How it connects
Connects probability with Bayesian probability, hypothesis testing, p-values, prior distributions, and A/B testing.

Also explained in

Compare how 3 books explain frequentist paradigm →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Grokking Bayes. The section links point to the parts of the book it draws on.