How to use
- Enter the four paired outcome cells.
- Choose whether you need a two-sided or directional read.
- Read the exact p-value first, then compare the asymptotic read only as supporting context.
Wave 34 statistics entry
Read paired change from one 2x2 table
The share URL stores only alternative and decimal places.
| Positive → negative | Negative → positive | Exact probability |
|---|
How to choose between nearby pages
Use McNemar test when the same items are measured twice and the question is directional change inside paired binary outcomes. Use Fisher exact test when the 2x2 table is unpaired and small, and move to risk ratio or odds ratio when the next question is effect-size framing rather than paired change significance.
FAQ
What does this page calculate?
It runs McNemar test on one paired 2x2 table and reports exact one-sided or two-sided p-values together with asymptotic chi-square reads.
When should I choose McNemar test instead of Fisher exact or chi-square?
Choose McNemar test when the same items are measured twice or paired into a matched 2x2 table and the question is whether discordant pairs tilt in one direction.
Does the share URL include my table counts?
No. The share URL stores only light settings such as alternative and decimal places. The four cell counts stay in your browser.
Why are exact and asymptotic results both shown?
The exact p-value is safer when discordant pairs are few. The asymptotic values help you compare with the textbook McNemar chi-square read and its continuity-corrected version.