Split-Plot Factorial Designs
Split-Plot Regular Two-Level Factorial Designs
2 - 15 Factors
The Split-Plot Regular Two-Level Factorial Design builder offers two-level full factorial and fractional factorial designs with restricted randomization. The number of unique runs will always be a power of 2 between 4 and 512 to estimate the main effects and interactions for between 2 and 15 two-level factors. This collection of designs provides an effective means for screening through many factors to find the critical few.
If there is only one hard-to-change factor, the design must be replicated to estimate p-values and test the hard-to-change factor effect.
The split-plot structure sorts the hard-to-change factor(s) into groups of like settings and randomizes their run order. This reduces the number of times hard-to-change factors are changed during the experiment. When the group changes, a complete reset and restart of the new group will provide the best estimates. This reset is necessary even if the HTC factor level remains unchanged from group to group.
Split-Plot Multilevel Categoric Design
2- 12 Factors
The multilevel categoric (general factorial) design allows you to have factors that each have a different number of levels. It will create an experiment that includes all possible combinations of your factor levels. All factors should be categoric (i.e. batch type, tool type, process method) rather than numeric.
The split-plot structure sorts the hard-to-change factor(s) into groups of like settings and randomizes their run order. This reduces the number of times hard-to-change factors are changed during the experiment. When the group changes, a complete reset and restart of the new group will provide the best estimates. This reset is necessary even if the HTC factor level remains unchanged from group to group.
If you have many factors and/or many categoric levels select the Split-Plot Optimal (custom) design on the Factorial tab. This algorithm selects a smaller number of design points, depending on the model chosen.
Name: (defaults to alphabetically ascending letters) Enter a descriptive name for each factor.
Units: (optional) Enter the units of measure for each factor.
Change: A factor can be set as Easy or Hard to change.
Easy: indicates this factor will be completely randomized and can potentially change from run to run.
Hard: indicates this factor will be changed as little as possible, restricting the randomization.
Type:
Nominal: (default) This type of factor is one that simply uses names or classes to describe the levels, for instance peanut butter types (Creamy, Chunky, SuperChunk).
Ordinal: This type of factor uses numbers that are ordered to show the natural progression, for instance temperature (200, 250, 300 Kelvin), where the baseline is the first level. These will be analyzed using orthogonal polynomial contrasts, which can be broken down into linear, quadratic, cubic, etc. components. All levels and combinations of levels of categoric factors will be included in the design.
Note
Instead of using ordinal contrasts you may be better off building a response surface design with discrete numeric factors.
Levels: Enter the number of levels (N) for each factor.
L[i]: Specifies the setting to use in the experiment. Specify the exact spelling and punctuation for each of the levels (a.k.a. treatments).
Note
If you change an Ordinal factor level manually, you will also need to appropriately edit the Contrast for that level. The software will keep its default values, which are based on the original spacing of the factor levels. (Honestly, it is easier to just re-build the design with the new factor levels!)
Optimal (custom)
2 - 30 Factors
This option creates a custom fractional factorial design that includes hard-to-change factors requiring restrictions to randomization. The customization includes defining a designed for model, and providing better control over the number of blocks, groups and runs than the multilevel categoric options.