Skip to Data Skip to Controls Skip to Chart Skip to Results

Randomization Test for Difference in ProportionsStatLens

Choose a dataset:

Edit or paste two columns (CSV with headers, group and outcome):

Open a CSV, TSV, or JSON file:

Group 1 Group label: Successes: Sample size (n₁):

Group 2 Group label: Successes: Sample size (n₂):

Count as success:

H₀: p_{Group 1} − p_{Group 2} = 0

Hₐ: p_{Group 1} − p_{Group 2}

Shuffles

Original Groups

→

Shuffled Groups

Select a dataset, then run a simulation to see results.

Help

New to this tool? Walk through it step by step.

Open guided walkthrough ↗

Keyboard Shortcuts

Shuffle the group labels many times, recording the difference in proportions each time. The histogram is the null distribution under “no difference”; the p-value is the share of shuffles at least as extreme as what you observed. This is a test — there is no confidence interval.

1: +1 shuffle
2: +10 shuffles
3: +100 shuffles
4: +1000 shuffles
0: Reset
Space: Play / Pause
?: Show this help
Esc: Close this dialog

Randomization Test for Difference in ProportionsStatLens ? ⚙

Original Groups

Shuffled Groups

Randomization Test for Difference in ProportionsStatLens