User:Rhansen/Getting Started with SAS
From BingWiki
What is SAS?
SAS is a popular statistical package for analyzing and managing data.
While some may find the SAS language difficult to learn (and some of SAS’s advanced features difficult to find and understand) SAS’s strengths are its: reliability and acceptance as a standard (for statistical reporting) in many large businesses, power at handling and flexibly generating any format of data (Plain text, a data base, other statistical packages) or output (HTML), and the well-documented comprehensiveness for many different analyses.
After becoming familiar with a small set of generally applicable procedures (those mentioned below) the user will commonly locate an example from the SAS Online Documentation to generate tabulated reports, charts, and plots, calculate descriptive statistics and do complex statistical analyses easily. Methods that are available include: hierarchical and multivariate regression, complex MANOVA on repeated measures, curve fitting (for Time series analysis, ETS), factor analysis, non-parametric tests. SAS works efficiently with any sized dataset.
| Getting Started with SAS/Part 1 | a simple program, debugging, basic data entry |
| Getting Started with SAS/Part 2 | printing data (PROC PRINT), inputting data from other files, output |
| Getting Started with SAS/Part 3 | basic analyses (PROC MEANS, UNIVARIATE, FREQ), categorical data, saving your work (as SAS Datasets and other) |
| Getting Started with SAS/Part 4 | basic statistical methods (PROCs TTEST, ANOVA, CORR, REG), more data manipulation (new variables, functions) graphs (bar charts, scatter plots) |
Campus Specific Notes
SAS is a large software package, and some complications usually arise on the wide variety of platforms that are commonly found on campus. As a result, the user should anticipate seeking the assistance from CSET staff to do an installation. SAS is licensed for all University-owned machines, though due to its size, and specialized use, it is not installed on all desktops. Windows 2000, an hour appointment and yearly renewals are minimal requirements for installation of the current version of SAS (9.1). An appointment for installation can be requested using the Software Database request form at http://computing.binghamton.edu/software (clicking on the "site licensed and volume purchased software packages" link in the paragraph Software For Your Computer, or by contacting the Help Desk at (7-6240, helpdesk@binghamton.edu).
Note: If you anticipate running SAS from your University provided UNIX account (Bingsun or StatMart) no installation is required. Other resources for SAS on Unix are found at http://statmart.cc.binghamton.edu/StatProg
Further Reading
Highlights of the SAS Online Documentation - Be sure to page through the Overview sections and Examples in the SAS Procedures Guide (esp. those mentioned in the Concise Glossary), Procedures like PROC REG are found in the SAS/STAT module. If you intend to do a complex Statistical Analysis there's lots of reading in the various sections Statements which are part of the Data step found in the SAS Dictionary Listing of Elements (esp. SAS Functions (e.g. by category), and the SAS System Options )
Concise Glossary for SAS (from the TTC Documentation Library, an older version is found (search for Concise)) – Most or nearly all of the SAS statements and PROCedure names are listed with links to the SAS Online Documentation kept locally on campus. After reading this narrative the glossary will be useful as a quick reference.
When running SAS click on Help Statistics – then the Search tab Help Graphs SAS/Graph, but note the /Analysis/Reporting/Graph-N-Go Saving work…Saving results/Exporting …HTML
Other narratives similar to this one are found at: UCLA's ATS (Academic Technical Services group), University of Texas or at other resources listed on the Statistical Programming Documentation page.
