Multiple-Testing Adjuster
Apply Holm or Benjamini–Hochberg adjustments to a complete family of exact p-values.
Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.
Adjust the exact p-values from one explicitly defined hypothesis family. Include every test in that family, not only the smallest p-values. Defaults are fictional.
Calculation results
4 tests; results retain the original input order. Compare adjusted p-values with your prespecified threshold.
| Test | Original p | Adjusted p |
|---|---|---|
| 1 | 0.01 | 0.03 |
| 2 | 0.04 | 0.06 |
| 3 | 0.03 | 0.06 |
| 4 | 0.002 | 0.008 |
Choose the error criterion first
Holm uses ordered p-values, multipliers m−i+1, and cumulative maxima; it controls family-wise error for valid p-values under arbitrary dependence. BH uses m/i and reverse cumulative minima; its usual false-discovery-rate guarantee assumes independence or suitable positive dependence. Values are capped at 1.
FDR is an expected proportion among rejected hypotheses, not the probability that a particular finding is false. These methods do not repair selective reporting, an incomplete test family, invalid p-values, or unplanned repeated looks at the data. Rounded entries such as “p < 0.001” are not accepted as exact numbers.
R documentation: p.adjust methods · Read the practical guide · Download blank worksheet (TXT)