# Screening designs

In many applications, the number of factors that potentially affect process quality is too great to study all factors in detail. The usual goal of a screening design is to identify the most important factors that affect process quality. After screening experiments, you usually do optimization experiments that provide more detail on the relationships among the most important factors and the response variables.

The following designs are often used for screening:
• 2-level fractional factorial designs
• Plackett-Burman designs
• Definitive screening designs

Your choice of screening design depends on many different considerations. Many considerations are specific to the process. The following are some considerations in Minitab:
The number of factors to study
The following are the maximum numbers of factors in Minitab for common screening designs:
• Definitive screening designs: 48
• Plackett-Burman designs: 47
• Fractional factorial designs: 15
Types of terms to study
Fractional factorial and Plackett-Burman designs are meant to screen linear terms. Definitive screening designs provide information about square terms and about more 2-way interactions.
Often, fractional factorial and Plackett-Burman designs have the lowest number of runs in a single replicate for a given number of factors. However, if a model needs quadratic terms, you must add runs to the fractional factorial and Plackett-Burman designs. A definitive screening design already includes runs to model square terms. If a model will include square terms, the definitive screening design can have the fewest runs per replicate.
Number of levels for the factor
Plackett-Burman designs and fractional factorial designs can have 2 levels per continuous factor. If you add center points to detect curvature in either type of design, each continuous factor has 3 levels.
Whether you can increase the order of the design sequentially
Sometimes, you begin with a smaller design to check whether square terms or interactions are present before you add more runs to estimate those terms. Folds and axial runs are two strategies in sequential experimentation.
In Minitab, you can fold Plackett-Burman and fractional factorial designs. A fold gives you runs to estimate interactions that are confounded by the initial fraction of the design.
With a fractional factorial design, you can also add axial runs. Axial runs let you estimate the square terms for the factors in the design.
By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy