Now Playing: Generating Experimental Designs (2:47)
There are several procedures in PASS that can be used to generate a variety of types of experimental designs. These procedures are different than most other procedures in PASS in that they are not used for sample size determination, but instead for creating columns of factor assignments.
The Design Generator procedure is the simplest of the experimental design tools. It produces columns with the combination of each of the entered levels of the factors.
As another example, screening designs are used to find the important factors from a large number of two-level factors. Suppose the desire is to obtain a six-factor screening design using 16 runs. The levels of each factor are produced on the Output Window as well as the Output Spreadsheet.
Next, we will consider the Fractional Factorial Design tool. Suppose we wish to generate a six-factor design using sixteen runs separated in blocks of four runs each. The result is a one-quarter replication design. The output gives details about the design, blocking, and aliases. The factor levels are given in both the regular output window and the output spreadsheet.
Other designs for which design columns can be generated in PASS are Latin Square Designs, D-Optimal Designs, Response Surface Designs, Taguchi Designs, Two-Level Designs, and Balanced Incomplete Block Designs.