SUGI27 - An Overview of the SAS User's Group International Meeting
April 14-17, 2002, Orlando, Florida
by John A. Wass
SUGI is an annual event for many industrial statisticians and programmers as well as business oriented users of mass data manipulation and storage techniques. SASŪ is a high level programming language originally developed to do applied statistics that today finds a multitude of users in many fields. This meeting attracts a diverse audience from all over the world, and is the major event of the year for many users, being their main avenue of hearing papers on topics of interest, attending tutorials and networking with colleagues. There are usually between 3,000-4,000 attendees.
The applied statistical tracks of presentations have much to recommend them to the mathematical modeler and biologist. Although many of the papers come from diverse areas, including psychology to economics, there are always papers on new procedures (called PROCS in the SAS language) and applications to problems in the natural sciences. Examples of presentations with applicability to biological modeling include such titles as: "Structural Equation Modeling for Health, Business, and Education", "Expanding Modeling Power in SAS/ETS Software", "Discriminant Analysis: A Powerful Classification Technique in Data Mining", "Use of the ROC Curve and the Bootstrap in Comparing Weighted Logistic Regression Models", "A SAS IML Macro for Computation of Confidence Intervals for Variance Components of Mixed Models", "Individual Growth Analysis Using PROC MIXED", "How to Use SAS for Logistic Regression with Correlated Data", "Robust Regression and Outlier Detection with the ROBUSTREG Procedure", and "Generating Randomization Using SAS Programming". This gives but a flavor of a single series of presentations that goes on simultaneously with many programming and applications tracks.
In addition, there are many demo areas where users can interact with SAS staff and receive information on upcoming versions of software and have questions answered on the use of procedures and programming issues. Although the language has many predefined statistical routines and diagnostics, it is exceedingly flexible and subroutines can be written in "C" to accomplish many specialized analyses within the system.
Prior to the meeting there are several 1-3 day courses that may be taken to extend a user's knowledge in specific areas of programming and statistics. The types of analysis that may be done within this language range from the simplest descriptive statistics to the more demanding clinical and theoretic studies. The reviewer uses the system both for industrial product development and research purposes.
As if the above were not enough to keep one occupied, there are numerous hands-on tutorial sessions to acquaint the researcher with specific software, beginners' seminars on programming topics, short statistical tutorials, and even a session on effective paper presentations. In addition there are specialized groups that are determined by the attendees themselves called BOF's (Birds of a Feather) and these involve areas of interest to a particular group, e.g., pharmaceuticals, business, data mining, etc. Here is where the researcher has the maximal opportunity to meet with colleagues of similar interests and find out what is going on in another laboratory or scientific group. Of course the best method is to attend the presentations of greatest interest. A short version of the Proceedings is published on-line and readily available.
The SAS Institute maintains a website where further information is available. Interested parties may point their browsers to http://www.sas.com/sugi.