Data considerations for Define Custom Definitive Screening Design

Consider the following guidelines when you define the design.

The design is to screen linear, quadratic, and interaction effects
A definitive screening design is usually for the case when you have many factors, but expect that only a small number of their linear, quadratic, and interaction effects are important.
Text factors have two levels
If your design has text factors with more than two levels, you can analyze it with Fit Regression Model without defining the design.
The size of the design should be adequate
The smaller the effect that you want to detect, the more data you need to collect. To increase the amount of data, you can replicate the runs in the experiment. Use Modify Design to add replicates.
The design should have 2 to 48 factors
If you have only one factor, you can use One-Way ANOVA without defining the design.
The run order should be randomized for data collection
If the worksheet contains a column that contains the run order, you can specify that column when you define the design. Otherwise, you can randomize the design after you define the design with Modify Design.
By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy