Workflow to study the effect of several factors.
Step #1:
Define the resolution, two levels of each factor, the number of factors, and the number of runs.
Determine the response.
Step #2:
Create and run a fractional factorial design using an R package. (use FrF2 package, or AlgDesign package)
always use (what is the right word?) the highest possible resolution.
Step #3:
View the output of the design, and finalize the design plan.
Step #4:
Run the experiment and collect data.
no replicate.
Step #5:
build a linear model for the main parameters and two-parameter interactions (or central composite face quadratic model).
(difficult term: aliased to each other. explain it in plain language)
visualize the effect of the main parameters and two-parameter interactions with Daniel plot. segregate parameters by their significance.
Remove insignificant parameters from the model and run the model again. Select significant main parameters and the two-parameter interactions.