Randomize design

Stat > DOE > Modify Design > Randomize design > Specify

This topic applies to screening designs, factorial designs, response surface designs, and mixture designs.

Randomize the design to reduce bias from extraneous or uncontrollable conditions. For more information, go to What is randomization?.

  • Randomize entire design: Randomize all the runs within each block, then randomize the order of the runs. Usually, you do not select this option if you have response data in the worksheet.
    Note

    For split plot designs, runs are randomized within whole plots, which may be within blocks.

  • Randomize just block: Randomize only the runs in 1 block and do not randomize the order of the blocks. Usually, you select this option for sequential experimentation. For example, use when you have one block with response data and then add a replicate for which you have not yet collected response data.
Base for random data generator
Enter a positive integer. If you use the same number when you modify the same base design again, the random order is the same. The random order is not the same if the base design order is not the same.
Display the design in run order
Reorder the rows of the design in run order. If you do not select this option, Minitab updates the RunOrder column, but does not reorder the other design columns.
Columns not to reorder
By default, Minitab sorts a column if the number of rows is less than or equal to the number of rows in the design. For columns with fewer rows than the design, Minitab enters missing value symbols (*) in the empty cells. Minitab does not reorder columns that have more rows than the design. You can only enter columns that are not included in the design.