A factor is any controllable input whose effect on a response is under study. Each factor runs at two or more levels, the specific settings used, such as 150 and 200 C, or supplier A and supplier B. Factors can be continuous or categorical, and the response is the measured output the experiment tries to explain, such as thickness, yield or strength. Picking the right factors and level ranges matters more than any analysis that follows.
Five factors at two levels is 32 runs full factorial and we only have wafers for 16. Make it a half fraction.
Set levels far enough apart to see an effect over the noise: a timid range is the most common reason a DOE comes back with nothing significant.
Also heard as
- input
- X
- control factor
Related terms
DOE
Design of experiments: planning a set of runs that vary several factors at once so their effects can be separated.
Factorial design
Experiment that runs every combination of the chosen factor levels, so each main effect and interaction can be estimated.
Interaction
Situation where the effect of one factor depends on the level of another, so their combined effect is not simply additive.
Knob
Any adjustable process parameter on a tool, such as RF power, pressure, gas flow, temperature or time.