Welcome to Statistics and Operations Research

SAS has long developed software for data analysis, econometrics, operations research, and quality improvement. The purpose of these pages is to provide our users with technical information about using this software, including details about software capabilities, examples, papers, e-newsletter, and communities.

Featured News


JSM 2019

Denver, CO
July 27 – August 1, 2019

More than 50 SAS employees from R&D, Technical Support, Publications, and Education, as well as from JMP, will be attending JSM. They will be teaching short courses and Computer Technology Workshops in addition to delivering papers, presenting posters, and attending roundtables.

Computer Technology Workshops (Added fee)

Click here to see a complete list of SAS and JMP presentations. Be sure to stop by the SAS booth at the EXPO and enter our drawing to win an Apple watch!

Predicting Inside the Dead Zone of Complete Separation in Logistic Regression

A logistic regression program can fail for a number of reasons, but even when it does not fail, other problems can appear. This paper focuses on one of those problems: complete separation and the existence of a dead zone in your data. Proceeding with analyses under complete separation is not necessarily fatal; the results can be successfully used as a practical tool, but you should handle them appropriately.

Read more.

tech support tip Estimating Attributable Risks When Covariates Are Present Using Logistic and Other Models This note discusses the model-based approach, which makes use of the NLEST macro for the final estimation of the attributable risk statistics. An example from the PROC STDRATE documentation is used to illustrate the model-based approach and to compare it with the stratification method in PROC STDRATE. Estimation based on the odds ratio is also shown for retrospective studies and when the risk of the outcome is small.

Read more.