liveBook · Manning

Two-tailed test

As explained in Timeless Algorithms, The Foundational Ideas

A two-tailed test is a hypothesis test whose rejection region includes extreme values in both directions of the test-statistic distribution.

A two-tailed test is appropriate when meaningful departures from a target could occur in either direction. Its alternative hypothesis includes values both below and above the null value, so extreme test-statistic values in either direction can support rejection. section 4.2.1

Beverage fill volumes

To test whether a batch averages 12 ounces, set H₀: μ = 12 and H₁: μ ≠ 12. The two-tailed choice fits concern about both underfilling and overfilling. Define the hypotheses before examining sample data so the test reflects the quality-control question, not the observed result. section 4.2.1

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Defines the choice through hypotheses and a practical example, while deferring the decision to reject the null based on sample evidence.
Examples
Uses a beverage batch with a 12-ounce target to contrast concern about underfilling and overfilling.
What you can do after
Apply it by stating hypotheses and choosing a two-tailed test before examining sample data when departures in either direction matter.
How it connects
Connects alternative and null hypotheses with one-tailed tests, significance levels, and Type I errors.

Also explained in

Compare how 2 books explain two-tailed test →

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.