Experimental Resolution

Experimental resolution is a measure of how clearly a fractional factorial design separates effects. It tells you which effects are confounded and how much confidence you can place in your estimates. Selecting the right resolution is essential for designing an experiment that balances efficiency with interpretability. 

Resolutions are labeled with Roman numerals: 

• Resolution III: Main effects are confounded with two‑way interactions. 

• Resolution IV: Main effects are clear; two‑way interactions are confounded with each other. 

• Resolution V: Main effects and two‑way interactions are clear; three‑way interactions are confounded. 

Higher resolutions provide clearer separation but require more runs. Lower resolutions are more efficient but blend more effects together. 

In screening experiments, Resolution III or IV designs are often sufficient because the goal is to identify key factors quickly. For characterization or optimization, higher‑resolution designs may be needed to ensure that interactions are estimated clearly. 

In the Improve phase, selecting the right experimental resolution ensures that your DOE provides the clarity you need without unnecessary effort. It helps you design experiments that are efficient, interpretable, and aligned with your improvement goals.