Home▸Statistics▸Design of Experiments: Full Factorial vs One-Factor-at-a-Time

Design of Experiments: Full Factorial vs One-Factor-at-a-Time

Interactive 3D simulation of a 2^3 full factorial design of experiments: watch how a factorial plan estimates main effects and interactions from the same 8 runs a one-factor-at-a-time approach needs 4 runs to blindly miss.

Statistics3DModerate60 FPS

Design of Experiments (DoE) is the statistical discipline of choosing which combinations of input factors to actually test, so that a small number of runs yields reliable estimates of every effect that matters — including interactions between factors. This simulation renders a 2×2×2 full factorial design as a glowing cube: eight corners, one per combination of three two-level factors. Switch to one-factor-at-a-time (OFAT) and watch the design collapse to four runs radiating from a single baseline corner — cheaper, but structurally blind to any interaction between factors, which the estimated effects panel exposes directly.

⚙ Under the hood

Interactive 3D simulation of a 2^3 full factorial design of experiments: a glowing cube whose 8 corners are the factor combinations, showing how a factorial plan estimates every main effect and interaction from the same runs a one-factor-at-a-time design needs fewer runs to blindly miss.

Three.jsstatisticsdesign of experimentsfactorial designdata science

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install