Data Analysis Papers
You Can't Stop Statistics: SAS/STAT Software Keeps Rolling Along
Maura Stokes, Robert Rodriguez, and R & D Staff, SAS Institute, 2006
Abstract
Release 9.2 of SAS/STAT software delivers a variety of new procedures and enhancements, which are motivated by methodological advances in statistics, technological improvements in the SAS System, and user requests for extensions to existing software. In order to respond to this spectrum of requirements, the development for each new release of SAS/STAT is balanced across a combination of experimental procedures, procedures which are achieving production status for the first time, and incorporation of new features in standard procedures. SAS 9.2 got a head start with the release of additional SAS/STAT procedures via Web download. These procedures, GLIMMIX, QUANTREG, and GLMSELECT, were made available as experimental procedures only for SAS 9.1 for Windows, and the GLIMMIX procedure was made production in a subsequent release.
This paper describes the statistical development that will surface in SAS 9.2. First, the "post" SAS 9.1 download procedures will be reviewed, with examples from the GLIMMIX and GLMSELECT procedures. Attention turns next to a preview of SAS Stat Studio, the successor to SAS/INSIGHT software. ODS Graphics becomes production with SAS 9.2, and this paper reviews its many enhancements and describes the use of the new effect plot in PROC LOGISTIC.
One of the most exciting enhancements to SAS/STAT sofware is the first installment of facilities for Bayesian analyses in several procedures. These new capabilities are illustrated with examples from the PHREG and GENMOD procedures. The TTEST procedure emerges as one of the more updated procedures in SAS 9.2, and this paper includes an example of using PROC TTEST to perform a crossover analysis. New regressions diagnostics were introduced in the MIXED procedure in SAS 9.1, and this trend continues with the addition of diagnostics for GEE analysis in the GENMOD procedure. The usefulness of this new feature is illustrated with an example. Finally, the myriad of other SAS/STAT changes are highlighted. Please see the Papers section of www.sas.com/statistics for an updated version of this paper once SAS 9.2 has been released.
Statistics and Operations
Research Home Page