Two factors interact when changing one shifts the response by a different amount depending on where the other is set: higher temperature might raise yield at low pressure and lower it at high pressure. On an interaction plot the lines for each level are not parallel, and crossed lines mean the best setting of one factor flips with the other. Finding interactions is the main reason to use factorial designs instead of one-factor-at-a-time testing.
There's a strong power-by-pressure interaction. High power only helps when you're at low pressure.
When a large interaction is present, main effects on their own mislead. Read the interaction plot before the main effects plot.
Also heard as
- two-factor interaction
- 2FI
- interaction effect
Related terms
Factorial design
Experiment that runs every combination of the chosen factor levels, so each main effect and interaction can be estimated.
Aliasing
Situation in a fractional design where two or more effects leave identical patterns in the data and cannot be told apart.
OFAT
One factor at a time: changing a single input while holding the rest fixed, the habit designed experiments replace.
ANOVA
Analysis of variance: splitting total variation into parts from each factor and from error, then testing which parts are real.