Screening designs are among the most popular designs for industrial experimentation. They are typically used in the initial stages of experimentation. They examine many factors to identify those that have the greatest effect on the response or responses. The factors identified are then studied using more sensitive designs.
Screening designs generally require fewer experimental runs than other designs. They are attractive because they are a relatively inexpensive and efficient way to begin improving a process.
In a screening design, each continuous factor is usually set at two levels to economize on the number of runs needed. The design consists of only a fraction of the possible combinations of factor levels. Tables of standard designs that involve continuous and categorical factors have been cataloged.
When standard screening designs exist, the Screening Design platform supplies a list of popular screening designs for two or more factors. These factors can be two-level continuous factors or three-level categorical or discrete continuous factors.
However, screening situations occur for which standard designs are not available. For these situations, JMP generates main effects screening designs. There are designs that are either orthogonal or near orthogonal. They focus on estimating main effects in the presence of negligible interactions.
In the Screening Design platform, you enter your factors and their levels. If a standard screening design exists, you are given the option to choose such a design from a list of standard screening designs that includes blocking designs. Alternatively, you can generate a main effects screening design, which assures you of an orthogonal or near-orthogonal design. If a standard screening design does not exist, a main effects screening design is automatically generated for you.