🧮 多重比较 P 值校正（Bonferroni / Holm / BH-FDR）Multiple Comparison P-value Adjustment (Bonferroni, Holm, BH FDR)
粘贴一组 P 值（可带名称，如基因名），一键计算 Bonferroni、Holm、Hochberg、Šidák、Holm–Šidák、Benjamini–Hochberg（FDR）和 Benjamini–Yekutieli 校正后的 P 值，标出在给定 α 或 FDR 下仍显著的项目，结果与 R 的 p.adjust() 和 statsmodels 完全一致，可导出 CSV。Paste p-values (optionally with labels such as gene names) to get Bonferroni, Holm, Hochberg, Šidák, Holm–Šidák, Benjamini–Hochberg (FDR) and Benjamini–Yekutieli adjusted p-values, with the items still significant at your α or FDR flagged. Identical to R p.adjust() and statsmodels; export to CSV.
结果与 R 的 p.adjust() 和 Python statsmodels.multipletests 一致（已逐一核对）。Bonferroni / Holm / Hochberg / Šidák 控制族错误率（FWER），Holm 总是不劣于 Bonferroni；Benjamini–Hochberg 控制错误发现率（FDR），适合多重比较较多的探索性分析（如基因、脑区、多指标）；BY 在任意相关结构下也控制 FDR，但更保守。Results match R p.adjust() and Python statsmodels multipletests (cross-checked). Bonferroni / Holm / Hochberg / Šidák control the family-wise error rate (Holm is never worse than Bonferroni); Benjamini–Hochberg controls the false discovery rate and suits many exploratory comparisons; BY controls FDR under any dependence but is more conservative.
使用说明与常见问题How-to & FAQ
「多重比较 P 值校正（Bonferroni / Holm / BH-FDR）」是什么？What is the Multiple Comparison P-value Adjustment (Bonferroni, Holm, BH FDR)?
粘贴一组 P 值（可带名称，如基因名），一键计算 Bonferroni、Holm、Hochberg、Šidák、Holm–Šidák、Benjamini–Hochberg（FDR）和 Benjamini–Yekutieli 校正后的 P 值，标出在给定 α 或 FDR 下仍显著的项目，结果与 R 的 p.adjust() 和 statsmodels 完全一致，可导出 CSV。Paste p-values (optionally with labels such as gene names) to get Bonferroni, Holm, Hochberg, Šidák, Holm–Šidák, Benjamini–Hochberg (FDR) and Benjamini–Yekutieli adjusted p-values, with the items still significant at your α or FDR flagged. Identical to R p.adjust() and statsmodels; export to CSV.
如何使用「多重比较 P 值校正（Bonferroni / Holm / BH-FDR）」？How do I use the Multiple Comparison P-value Adjustment (Bonferroni, Holm, BH FDR)?
每行粘贴一个 P 值，或“名称, P 值”（可直接从 Excel 复制两列）。选择校正方法和 α，表格显示原始 P、校正后 P 和是否显著；勾选“同时显示所有方法”可以并排比较。Paste one p-value per line, or “label, p” (two columns copied from Excel work too). Choose the method and α; the table shows raw p, adjusted p and significance. Tick “Show all methods” to compare them side by side.
Bonferroni 和 BH（FDR）该用哪个？Bonferroni or BH (FDR)?
需要严格控制“至少一个假阳性”的概率（如确证性试验的少数主要终点）时用 Holm（总是不劣于 Bonferroni）；比较很多（如基因表达、脑影像、多指标探索）且能接受一定比例假阳性时用 Benjamini–Hochberg 控制错误发现率。Use Holm (never worse than Bonferroni) when you must control the chance of any false positive, e.g. a few primary endpoints. Use Benjamini–Hochberg when there are many comparisons (genes, voxels, exploratory outcomes) and a controlled proportion of false discoveries is acceptable.
校正后的 P 值为什么会等于 1？Why are some adjusted p-values exactly 1?
校正后的 P 值上限为 1。例如 Bonferroni 是原始 P × 比较次数，超过 1 时记为 1。Adjusted p-values are capped at 1 — Bonferroni multiplies by the number of tests, and anything above 1 is reported as 1.
「多重比较 P 值校正（Bonferroni / Holm / BH-FDR）」免费吗？文件会上传吗？Is the Multiple Comparison P-value Adjustment (Bonferroni, Holm, BH FDR) free? Are my files uploaded?
「多重比较 P 值校正（Bonferroni / Holm / BH-FDR）」完全免费，无需注册。你输入的内容只在本机浏览器里处理（并保存在本机方便下次继续），不会上传到任何服务器，适合处理未发表的论文和数据。The Multiple Comparison P-value Adjustment (Bonferroni, Holm, BH FDR) is free with no sign-up. Everything you enter is processed in this browser only (and kept locally so you can continue later); nothing is uploaded, so it is safe for unpublished papers and data.