ClarityKitFind a tool
Statistics

Multiple Testing Correction Calculator

Adjust a family of p-values using Bonferroni, Holm or Benjamini–Hochberg while preserving your input order.

Data stays on this deviceConvention explained below
Data explorer
Data explorer

Check the dataset and sample convention before interpreting results.

See the method

Add your observations

Runs on your device

Your statistical summary

Your result

Let’s calculate.

Use the Calculate button to see your result.

Review the method below for assumptions and conventions.

How to use this tool

  1. Enter unadjusted p-values, adjustment method, significance level.
  2. Select Calculate to view the result.
  3. Check the method and assumptions below before using the result.

Input reference and sample values

Use the units and choices shown for each field. Sample values are provided for practice, not as recommendations.

InputFormat and unitsSample
Unadjusted p-valuesKeep the structure and separators shown in the sample. Line breaks are retained in this input..001, .02, .03, .2
Adjustment methodChoose Holm: family-wise error / Bonferroni: family-wise error / Benjamini–Hochberg: false discovery rate.Holm: family-wise error
Significance levelNumeric value. Minimum: 0.0001. Maximum: 0.2.0.05

Result for these sample inputs

1
Tests in family
4
Adjusted p, test 1
0.004
Adjusted p, test 2
0.06
Adjusted p, test 3
0.06
Adjusted p, test 4
0.2

Input order is preserved. Define the complete family of tests before selecting an adjustment.

Select Try the example in the form to load these values and calculate them again. This replaces the current form inputs. Compare the result with the method and worked explanation below.

The method, explained

Bonferroni multiplies by the number of tests. Holm uses sorted step-down factors with a cumulative maximum. Benjamini–Hochberg uses sorted p × m/rank with a reverse cumulative minimum. All adjusted values are capped at 1.

A WORKED EXAMPLE

Using unadjusted p-values = .001, .02, .03, .2, adjustment method = holm, significance level = 0.05, the result is 1 probability. Change these example inputs to match your task; use the method above to check each step.

What to keep in mind

Bonferroni and Holm control family-wise error under arbitrary dependence. Benjamini–Hochberg controls FDR under independence or specified positive dependence; it is not a family-wise guarantee. A descriptive or probability calculation under the stated assumptions. Check the sample design and model before interpreting results.

Reference: SciPy: false discovery control

Common questions

Which error rate does each method control?

Bonferroni and Holm address the chance of any false rejection in the test family. Benjamini–Hochberg addresses the expected false-discovery proportion under its dependence assumptions. They answer different error-control questions.

Do I need to sort the p-values first?

No. Enter them in your preferred order. The calculation sorts internally where needed and maps every adjusted value back to its original test number.

Methodology maintained by ClarityKit. How these tools are built and checked.