Mixture Designs |
In a mixture design, all main effects are considered to be significant. Second- and higher-order interactions are selected through ANOVA or stepwise regression.
The prediction profiler and contour optimizer provide two graphical methods for optimizing the response.
You can incorporate additional linear constraints on the factors in a mixture design, and you can incorporate process variables. This will be discussed in Chapter 11, "More on Mixture Designs."
Copyright © 2008 by SAS Institute Inc., Cary, NC, USA. All rights reserved.