5.3. Choosing and running an experimental design
5.3.3. How do you select an experimental design?
188.8.131.52. Fractional factorial designs
|By ‘Screening Design’ it is
generally meant an experimental plan that is intended to find the few significant
factors from a list of many potential ones. Alternatively we refer to a
design as a screening design if its primary purpose is to identify significant
main effects, rather than interaction effects, the latter being assumed
an order of magnitude less important. Even when the experimental goal
is to eventually fit a response surface model (an RSM analysis), the first
experimant should be a screening design when there are many main effects
(5 or more).
Screening designs are typically of resolution III. The reason is that in designs of resolution III, some main effects are confounded with two-factor interactions; the two-factor interaction are assumed of minor importance to the main effects (this might not be true, but we assume it anyway).
Sometimes designs of resolution IV are also used for screening designs. In these, main effects are confounded with, at worst, three factor interactions. This is better from the confounding viewpoint, but they require more runs than a resolution three design.
Another common screening-type family of designs are the Plackett-Burman designs, so named after their inventors. These designs are of resolution III and will be described later.
In short, screening designs are economical experimental plans which
focus on determining the relative significance of many main effects.