Benjamini-Hochberg False Discovery Rate: Step-Up Rule versus Bonferroni and Uncorrected Testing
Test hundreds of hypotheses at once and compare three decision rules on the same sorted p-values: uncorrected, Bonferroni and the Benjamini-Hochberg step-up line. Change the share of true nulls, the effect size and the level, and watch false discoveries and power trade off. The exact identity FDR equals pi0 times alpha is checked against simulation. A Blender-built plotting easel with a tiered cake prop, numerical checks built in.
Test hundreds of hypotheses at once and compare three decision rules on the same sorted p-values: uncorrected, Bonferroni and the Benjamini-Hochberg step-up line. Change the share of true nulls, the effect size and the level, and watch false discoveries and power trade off. The exact identity FDR equals pi0 times alpha is checked against simulation. A Blender-built plotting easel with a tiered cake prop, numerical checks built in.
Test hundreds of hypotheses at once and compare three decision rules on the same sorted p-values: uncorrected, Bonferroni and the Benjamini-Hochberg step-up line. Change the share of true nulls, the effect size and the level, and watch false discoveries and power trade off. The exact identity FDR equals pi0 times alpha is checked against simulation. A Blender-built plotting easel with a tiered cake prop, numerical checks built in.
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install