|Macros for the Design and Analysis of Experiments
The SAS System provides the tools to give you a complete software solution
for constructing and analyzing experimental designs. Moreover, you can
use these tools in a variety of ways. If you are comfortable with SAS
programming, the FACTEX and OPTEX procedures, described in Part 3, "The FACTEX Procedure"
and Part 6, "The OPTEX Procedure" are fundamental tools for constructing many types of
experimental designs; and the analytic procedures of SAS/StatSoftware
enable you to compute the statistical results of your experiments.
Alternatively, you can use the point-and-click ease of the ADX Interface to
construct and analyze designs for your experiments. The macros discussed in this
appendix provide a third, intermediate alternative: you use them in SAS
programming, but they bundle much of the syntax required for the general
procedures into macro arguments. The macros draw on various SAS software tools:
- two-level factorial and fractional factorial designs for as
many as 128 runs and 11 factors. This includes designs with
and without blocking.
- two-level screening designs (Plackett-Burman designs) for as
many as 47 factors
- orthogonal and rotatable central composite designs (Box-Wilson
designs) for as many as 8 factors. This includes designs with
and without blocking.
- mixture designs for either constrained or unconstrained
components, with no limit on the number of factors. This includes
simplex-centroid, simplex-lattice, and McLean-Anderson designs.
You can also use the macros to
Note again that the macros in this collection are primarily intended to
provide a programming interface to common experimental design tasks,
as an alternative to using the underlying procedures directly. For users
who are not experienced with SAS programming, the ADX Interface may
be a more appropriate tool. The ADX Interface, which has been completely
revised in Version 7, 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. Information about the ADX Interface can be found at
. The ADX Interface is
documented in Getting Started with the ADX Interface for Design
- decode a design into units meaningful for your application
- randomize the design and display a data collection form
- perform a power-transformation analysis for the response variable
- analyze fractional factorial designs. The analysis includes a
listing of the alias structure and a normal plot for the effects.
Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.