Perform a mixtures experiment

The design and subsequent analysis of a mixture experiment might have the following steps. Depending on your experiment, you might do some of the steps in a different order, perform a step more than one time, or eliminate a step.

  1. Choose a mixture design for the experiment. Before you start using Minitab, you need to determine what design is appropriate for your problem.
    • Use Stat > DOE > Mixture > Create Mixture Design to generate a simplex centroid, simplex lattice, or extreme vertices mixture design. You can also include amounts or process variables in your design to create mixture-amounts designs and mixture-process variable designs.
    • Use Stat > DOE > Mixture > Define Custom Mixture Design to create a design from data you already have in the worksheet. Define Custom Mixture Design lets you specify which columns contain your components and other design characteristics. You can then easily fit a model to the design.
  2. Use Stat > DOE > Modify Design for any of the following tasks:
    • Rename the components.
    • Replicate the design.
    • Randomize the design.
    • Renumber the design.
    • Rename process variables.
    • Change the values of the factor levels for the process variables.
  3. Use Stat > DOE > Display Design to change the display order of the runs and to change the units in which Minitab expresses the components or process variables in the worksheet.
  4. Conduct the mixture experiment and collect the response data. Then, enter the data in your Minitab worksheet.
  5. Use Stat > DOE > Mixture > Analyze Mixture Design to fit a model to the experimental data.
  6. Use plots to visualize the design space or response surface patterns. Use Simplex Design Plot to view the design space, or Response Trace Plot and Contour/Surface Plots to visualize response surface patterns.
  7. If you are trying to optimize responses, use Stat > DOE > Mixture > Response Optimizer or Stat > DOE > Mixture > Overlaid Contour Plot to obtain numerical and graphical analyses.
By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy