Design of experiments is the structured way to learn how inputs drive outputs: choose factors and levels, pick a design such as a full or fractional factorial, randomize the run order, run it, then analyze with ANOVA or regression. Varying factors together, rather than one at a time, gives the same precision in fewer runs and reveals interactions. Insiders use a DOE as a noun for any planned experiment.
Let's run a DOE instead of more one-offs: pressure, power and O2 flow, two levels each, plus three center points.
Also heard as
- design of experiments
- designed experiment
- DOX
Do not mix it up with
- OFAT: OFAT changes one factor while holding the rest fixed, so it misses the interactions a DOE is built to find.
Related terms
Factorial design
Experiment that runs every combination of the chosen factor levels, so each main effect and interaction can be estimated.
Fractional factorial design
Experiment that runs a chosen fraction of all factor combinations, trading some aliased effects for far fewer runs.
Factor
Input variable deliberately set at chosen levels in an experiment, such as temperature, pressure, time or supplier.
Randomization
Running experimental trials in a random order so unknown time trends and nuisance effects do not bias the results.
OFAT
One factor at a time: changing a single input while holding the rest fixed, the habit designed experiments replace.