Replicate a Design
Augment a design with replication for a direct check on the assumption that the error variance is constant. Replication also reduces the variability of the regression coefficients in the presence of large process or measurement variability.
This example replicates an 8 run design.
1. Select Help > Sample Data Folder and open Design Experiment/Reactor 8 Runs.jmp.
2. Select DOE > Augment Design.
3. Select Percent Reacted and click Y, Response.
4. Select Feed Rate, Catalyst, Stir Rate, Temperature, and Concentration and click X, Factor.
5. Click OK.
Figure 6.12 Choose an Augmentation Type
6. Select Group new runs into separate block to add a blocking variable to the design.
7. Click Replicate, enter 2, and then click OK.
Note: The replicate dialog box asks for the number of times each run is to be performed. Entering 2 specifies that you want each run to appear twice in the resulting design. This results in a single replicate of your original design.
Figure 6.13 Reactor Data Design Augmented with a Replicate
The original 8 runs are in Block 1 and the replicated runs are in Block 2.
Note: The full design can be evaluated from the Augment Design Platform. For more information about design evaluation, see Evaluate Designs.
8. Click Make Table.
Figure 6.14 The Replicated Design
The table includes the original design and results in Block 1. The augmented runs are in Block 2. The percent reacted values can be added to the table after the augmented runs have been measured.