StuckOut

ResourcesStatistics

Choosing a Statistical Test Your Data Actually Supports

The wrong test on the right data is worth no marks. A short decision path.

7 min read

Start with your data type

Continuous, ordinal or categorical outcome. That single answer removes most of the options before you consider anything else.

Then how many groups

One sample, two groups, or more than two. Paired or independent. These two questions plus the data type get you to a shortlist of three.

Check the assumptions before running it

Normality, equal variance, independence. Running the test first and checking after is how students end up reporting a result their data cannot support.

Report the effect size too

A p value tells you whether there is an effect. It does not tell you whether it matters. Markers increasingly want both.

Still not moving?

If this did not unstick you, the problem is probably specific to your code or your data. Send it over and we will look at the actual thing.

Send my requirements →

Related subjects