Robust Parameter Designs

Robust parameter designs are based on the work of Genichi Taguchi. The purpose of robust designs is to find the settings of controllable factors that will minimize variation in the response due to uncontrollable noise factors. This is done by combining an inner array of control factors with an outer array of noise factors. All control factor setting combinations specified in the inner array are tested at each noise factor settings combination specified in the outer array.

Available Robust Parameter Designs

The available designs for the inner and outer arrays of a robust parameter design depend on the design type selected for each array.

For the inner array, the following design types may be used. Click a design type to view the available designs.

• Two level factorial designs
• Plackett-Burman designs
• General full factorial designs
• Taguchi OA designs

Note that most Taguchi orthogonal arrays are saturated and therefore not effective for examining factorial interactions. If interactions of the control factors are likely, using a two level factorial design for the inner array may be desirable.

For the outer array, all of the above design types are available except Taguchi OA designs. When configuring the outer array, note the following:

• You can use a two level factorial design, a Placket-Burman design or a general full factorial design.
• Each of these design types can have 1 to 15 factors when used for the outer array.
• In addition to the settings described for configuring the inner array, you can also specify a number of replicates in the Additional Settings for the outer array.
• You cannot define blocks, center points, responses or repeated measurements for the outer array, regardless of the design type.

For more information about how to use the design types, please consult the documentation on design folios.