Non-Parametric Test Concept Guide

Use

Use this when data is skewed, ordinal, small, or full of outliers.

Beginner Idea

Non-parametric thinking often looks at order or ranks instead of assuming clean bell-shaped numeric data.

When To Be Careful

• Small sample.
• Strong skew.
• Extreme outliers.
• Ordinal ratings.
• Counts with very small cells.
• Averages that do not describe a typical case.

Safer Wording

Because the data is ..., a rank-based or simpler descriptive comparison may be safer than a standard mean comparison.