Use screening designs when there are too many runs in a standard design and you are trying to understand six or more components. The goal of screening experiments is to simplify the follow up stages of the experiment. This is accomplished by learning what components can be held fixed (no significant linear effect), or combined with other components (similar linear effects). The model will always be linear. There are two flavors of screening design, simplex and non-simplex. If the single component ranges are all equal the simplex flavor is used. If the single component ranges are not equal and/or multi-component constraints have been added then the non-simplex flavor is used.
Mixture Components: How many components are involved in this experiment?
Fillers may be components too unless their proportion of the mixture is held constant.
Components that will remain a constant proportion of the total can be added to the design by setting the low and high limits to the same value.
Total: The sum of all the components must equal the total for all runs.
Units: (optional) All components must use the same units of measurement.
Common units of measure include direct amounts (mass or volume), weight %, % volume by weight, and molecular equivalents.
Name: Provide names for the components to help with later documentation
Low: The lowest value a component can take (often zero).
High: The highest value a component can take (no larger than the Total)
Edit Constraints button: Click this button to impose additional constraints on component combinations allowed. For more details use the help button on the Edit Constraints dialog.