Statistics Kingdom

Statistical Test Decision Tree

Answer a few quick questions about what you're checking and your data, and follow the diagram to the correct statistical test, regression or survival analysis.

Interactive diagram 40+ tests covered No signup
Question Click your answer Your selected path Recommended test — click to open
How to use this tool

Click a bubble in the diagram to answer each question. The tool then reveals the next relevant question, building a path towards your test. Keep answering until you reach a recommended test, then click its blue link to open the matching calculator.

What is a statistical test?

A statistical test checks whether you can reject a null assumption about a population, based on sample data. Common uses include checking whether a measurement matches an expected value, comparing two or more measurements, and deciding whether two variables are significantly related.

How does a statistical test work?
  1. Calculate a statistic that captures the assumption you're testing.
  2. Estimate the distribution of that statistic assuming the null assumption is correct.
  3. Find the probability (p-value) of getting a statistic this extreme under that assumption.
  4. If the p-value is small — usually below 0.05 — it's reasonable to reject the null assumption.
Quick pointers
  • Comparing a central tendency → mean, T or rank tests
  • Comparing proportions or counts → proportion or Chi-squared tests
  • Relationship between variables → correlation or regression
  • Time-to-event data → Kaplan-Meier survival analysis