| Overview |
The FACTEX procedure constructs orthogonal factorial experimental designs. These designs can be either full or fractional factorial designs, and they can be with or without blocks. You can also construct designs for experiments with multiple stages, such as split-plot (Huang, Chen, and Voelkel 1998) and split-lot designs (Butler 2004). After you have constructed a design by using the FACTEX procedure and run the experiment, you can analyze the results with a variety of SAS procedures including the GLM and REG procedures.
Factorial experiments are useful for studying the effects of various factors on a response. Texts that discuss experimental design include Box, Hunter, and Hunter (1978), Cochran and Cox (1957), Montgomery (1991), and Wu and Hamada (2000). For details about the general mathematical theory of orthogonal factorial designs, refer to Bose (1947).
Note: For two-level designs, instead of using PROC FACTEX directly, a more appropriate tool might be the ADX Interface for Design of Experiments. The ADX Interface is designed primarily for engineers and researchers who require a point-and-click solution for the entire experimental process, from building the designs through determining significant effects to optimization and reporting. ADX gives you most of the two-level designs provided by the FACTEX procedure in a system that integrates construction and analysis of designs, without the need for programming. In addition to two-level designs for standard models (with and without blocking), ADX makes it easy to use PROC FACTEX to construct designs for estimating particular effects of interest. Moreover, ADX also uses the OPTEX procedure to construct two-level designs of nonstandard sizes. For more information, see Getting Started with the SAS ADX Interface for Design of Experiments.