False Discovery Rate (FDR) Online Calculator
Paste your p-values below to run a multiple-testing correction using Benjamini-Hochberg (BH) or Benjamini-Yekutieli (BY).
| Rank | Original Test # | P-value | Critical Value | Adjusted P (q-value) | Significant |
|---|
What this FDR online calculator does
This fdr online calculator helps you correct for multiple hypothesis testing. When you run many tests at once, some small p-values happen by chance. False Discovery Rate (FDR) methods control the expected proportion of false positives among the findings you call significant.
Instead of asking, “What is the chance I make even one false positive?” (very strict), FDR asks, “Among the significant results, what fraction are likely false?” That makes FDR a practical default in genomics, psychology, A/B testing, epidemiology, and exploratory data science.
How to use this calculator
- Enter your target FDR level, usually 0.05 or 0.10.
- Choose a correction method:
- BH for standard independent or positively dependent tests.
- BY for arbitrary dependence (more conservative).
- Paste p-values separated by commas, spaces, tabs, or line breaks.
- Click Calculate FDR.
- Review:
- How many discoveries are significant at your chosen level
- The largest significant p-value (if any)
- Adjusted p-values (q-values) for each test
Benjamini-Hochberg in plain English
Step-by-step idea
Suppose you have m p-values. Sort them from smallest to largest. For each rank i, compute a critical value:
BH critical value = (i / m) × α
Find the largest rank where p(i) ≤ critical(i). All p-values up to that rank are called significant.
The calculator also reports adjusted p-values (q-values), which tell you the smallest FDR level at which each test would be significant.
BH vs BY: which one should you choose?
Use BH when:
- Your tests are independent or weakly/positively dependent.
- You want good statistical power.
- You are doing typical high-throughput screening or broad exploratory work.
Use BY when:
- Dependence structure is unknown or potentially complex.
- You prefer a safer, stricter correction.
- False positives are especially costly.
Interpreting the results table
- Rank: Position after sorting p-values ascending.
- Original Test #: Where that p-value came from in your input list.
- P-value: Raw p-value.
- Critical Value: BH/BY threshold at that rank.
- Adjusted P (q-value): Multiple-testing-adjusted value.
- Significant: Whether it passes your chosen FDR target.
Common mistakes to avoid
- Mixing one-tailed and two-tailed p-values in the same correction set.
- Correcting only “interesting” tests instead of the full family of tests performed.
- Treating q-values as effect sizes (they are not).
- Ignoring practical significance after statistical significance.
- Using FDR when your objective is strict family-wise error control (consider Bonferroni/Holm instead).
Quick FAQ
Is this the same as Bonferroni?
No. Bonferroni controls the probability of any false positive and is usually more conservative. FDR controls the expected fraction of false positives among declared discoveries.
What FDR level should I use?
Common choices are 0.05 and 0.10. Lower values are stricter; higher values are more permissive.
Can I paste values with line breaks and commas?
Yes. This calculator accepts commas, spaces, tabs, and new lines.
Final note
A good fdr online calculator is most useful when paired with strong study design, clear hypotheses, and transparent reporting. Use these corrections to improve reliability—not as a substitute for thoughtful analysis.