5. Multi-level plans

5.1 General

Two-level designs are widely used because they are economical in terms of the number of trials. But there's no reason not to consider designs with factors taking up more than two levels. Each factor must be given the number of levels necessary to meet the requirements of the study. Here too, there are complete designs and fractional designs, which enable the number of tests to be reduced despite the increase in the number of levels.

Interpretation of the results will depend on the nature of the variables.

If the factors are continuous, a first- or second-degree model can be used, and the calculations performed with the