Introduction SAS is probably the most widely used statistical software package. It is despised by some and used by many. It's actually one of the easiest such packages to learn, and a variety of analytic and scientific fields require it. About 25 years ago, the bulky SAS manuals provided most of the documentation; often the most helpful information was learned from a friend (still the best method). In recent years, however, there's been a small boom in SAS books at all skill levels. Potential buyers should probably consider their experience with SAS or any other packaged programming, their preferred learning style (e.g., long explanations vs. concise), their need for statistical instruction vs. SAS instruction (many SAS guides give minimal explanation of the statistics involved) and, unfortunately, their budget.
My Top Three SAS Books Just to show you my biases, I'll begin by listing my favorite SAS books. They all have clear writing and organization, cover useful topics, and have been reviewed favorably at Halfvalue.com. These books are 'The Little SAS Book: A Primer, Third Edition','Applied Statistics and the SAS Programming Language', and 'Handbook of Statistical Analyses Using SAS, Second Edition'. Hint: Search these authors' names to discover other excellent books that may be more helpful for your specific needs. I Have to Learn WHAT? Whether you're an eager beaver or a reluctant recruit, there are a number of excellent introductory SAS books. The authors' dilemma is to strike a balance between writing about the basics and not oversimplifying so that you feel like you've wasted your money. 'The Little SAS Book: A Primer, Third Edition' does this best (recently updated from the popular 2nd edition), although there are fans of the more statistically oriented 'Quick Start to Data Analysis with SAS (Statistics Software)'. As stated above, the 1995 DiIorio and Hardy seems like an excellent beginner/intermediate transition book, but it is not as quite as good as the Delwiche and Slaughter for the true beginner. The Intermediate Level User Although I haven't used 'Quick Start to Data Analysis with SAS (Statistics Software)' Halfvalue.com allows provides a view of the Table of Contents, some sample writing, and the index--this looks like an excellent bridge between the beginner and intermediate levels, with some material for relatively advanced programmers as well. The recently revised 'SAS Programming for Researchers and Social Scientists' is also excellent, and has a website link. Books that cover more thoroughly the PROCS you may have learned as a beginner include 'Proc Tabulate by Example' Making Sure You've Really Got It At this stage, you'll also want to get one of the "Debugging" books, since you'll probably have discovered how easily programming results in immediate mistakes (not as bad, however, as the "not-know-til-later" mistakes). Ron Cody once again provides very useful information in 'Cody's Data Cleaning Techniques Using SAS Software'. 'The SAS Workbook and Solutions Set' really tests your knowledge, with numerous problem--some of them very difficult and/or tricky--and a separate answer book. (You may find that some of the problems ask for greater detail (e.g., very specific formatting requirements) than is required by your school or work setting. Many people, including myself, tend towards the "get it done correctly and efficiently" philosophy, and de-emphasize the elegance of the programming or the output.) 'Debugging SAS Programs: A Handbook of Tools and Techinques' has some good techniques, but is more oriented (compared with Cody) towards using SAS in a Windows environment (SAS PC), and those with MACRO experience. MACROs, SQL, and Other Specialized Techniques These SAS procedures are designed to save you time, increase your flexibility, and perform more narrowly used analyses. I suggest that one not invest the money or time in these topics until he or she is sure that they are worth it. SQL is particularly useful for atypical dataset merges, and I found that 'SAS SQL Procedure User's Guide, Version 8' is clearer and better organized than 'Getting Started With SQL Procedure: Version 6'; 'The Practical SQL Handbook: Using SQL Variants (4th Edition)' is highly rated but neither geared towards SAS nor the social sciences. Popular MACRO books include 'SAS (R) Macro Language Course Notes' and 'SAS Macro Programming Made Easy'. Finally, almost any book by Ron Cody will be clear and helpful. Some of his more specialized books include 'Longitudinal Data and SAS: A Programmer's Guide' and 'SAS Functions by Example'. For the Advanced User, and a Couple of Websites for All 'Handbook of Statistical Analyses Using SAS, Second Edition' and 'Data Mining Using SAS Applications' get high reviews at Halfvalue.com for those experienced enough to require advanced emphasis on statistics and on "data mining (validity checking, etc."). IF you're at this level of expertise, you'll probably benefit from any SAS Institute published book, and will benefit from advanced tutorials and articles. Online sites for the advanced include http://www.lexjansen.com/sugi and http://support.sas.com. While these latter two have information for the non-advanced, UCLA's http://www.ats.ucla.edu/stat and the University of Georgia's http://listserv.uga.edu/archives/sas seem more inclusive of all skill levels. The latter site is a message board for discussing SAS problems, and users seem friendly and helpful. Conclusion As stated earlier, users have different learning styles and come to SAS with varying levels of experience; therefore, this can only be a rough guide. I learned SAS initially with a voluminous SAS manual, a small handbook, and a lot of help from friends and colleagues. These days, the Net can aid your search for SAS help as well (e.g., searching for ready-made MACROs, tutorials on SAS and Job Control Language). It is a tradition for users to help each other, especially beginners, because everyone has been there, and everyone has more that they can learn. :) |