Hypothesis-testing workflow and common errors

Learning objectives

• Form H0 and H1 before touching data.
• Interpret p-value, Type I/II errors and power.
• Report effect size with the test result.

Hypothesis workflow

Question → parameter and minimum effect → test choice → assumption checks → calculation → decision → practical assessment.

A low p-value is evidence against H0. It is not the probability that H0 is true.

Assumptions and limitations

• Independence, distribution, variance and sample size.
• Pre-registered question; no data-driven hypothesis switching.
• Report estimate, CI and effect alongside the p.

Garment-factory example

H0: mean chest measurement equals target. H1: it differs. p = 0.02 is evidence against H0; the estimated bias and its CI decide whether to act.

Method

1. Fix the question and the minimum effect.
2. Choose the test and check assumptions.
3. Calculate with a free spreadsheet.
4. Decide and communicate risks and limitations.

Common mistakes

• Repeated testing until significant.
• Accepting H0 as proven when p is large.