What McNemar’s test does
McNemar’s test is used on paired categorical data — for example, the same group of people surveyed before and after an event, each classified as ‘yes’ or ‘no’ both times. It checks whether the proportion changing from one category to the other differs from what you’d expect by chance, using only the pairs that actually changed (the discordant pairs).
How this calculator works
Enter the two discordant pair counts: how many cases switched from ‘yes’ to ‘no’, and how many switched from ‘no’ to ‘yes’ (pairs that stayed the same in both categories aren’t used in this test). The calculator computes a chi-square statistic, applying a continuity correction by default since that is standard for this test, and converts it into a p-value using the chi-square distribution with one degree of freedom.
A small p-value (typically below 0.05) suggests the change between the two categories was not symmetric, indicating a real shift rather than random noise. This tool is for general educational and planning purposes.