Previous Page

|

Next Page

Design Augmentation

In some situations, it is statistically useful and more economical to augment an existing design rather than to perform a completely new experiment. You should consider design augmentation in the following situations:

• The optimum factor settings are within the original experimental region, so that a screening design can be augmented to a response surface.
• The initial fractional factorial experiment cannot distinguish between aliased significant effects, and so it must be augmented to break certain alias strings.
• Conducting an additional experiment would be too expensive, so you want to reuse information from an initial screening experiment.

Covered in This Section

Augmenting Designs

Folding Over a Design

Adding Optimal Runs

Augmenting to a Response Surface Design

Deleting Inactive Factors

Points to Remember