Benjamini–HochbergHolm · Bonferroni · in the browser on-device

Correct your p-values — three ways, one table.

Paste a column of p-values and get them adjusted by Bonferroni, Holm (1979) and Benjamini–Hochberg (1995). Pick an α and see which hypotheses each method lets through. Nothing is uploaded; the arithmetic runs in this tab.

P-values · one per line, or separated by spaces/commas

—

α · threshold

Default 0.05. Lower α is stricter.

Correct Load the BH 1995 example (15 p-values) Clear

• Bonferroni: reject if raw p ≤ α/m. The oldest and strictest.
• Holm (1979): step-down on sorted p-values; controls the family-wise error rate like Bonferroni, but rejects at least as many. Scand. J. Statist. 6, 65–70.
• Benjamini–Hochberg (1995): step-up procedure that controls the false discovery rate (FDR) at α. More powerful when many tests are run. doi:10.1111/j.2517-6161.1995.tb02031.x