During this complimentary, 50-minute session, statistics professor John Bailer will explore ways to confirm that you have correctly constructed or recoded variables, illustrating surprises that might arise when SAS® does what you said instead of what you meant.
What You'll Learn
- Strategies for recoding variables, including Boolean "tricks" to define categorical variables.
- What should be present in a test data set for confirming that code works.
- Dangers of open-ended ranges when defining categorical variables in the presence of missing or out-of-bounds values of variables.
- The value of automatically constructing new variables using output data sets produced by PROCs and a little DATA step programming (time permitting).
Who Should Attend?
Anyone interested in issues that might arise when constructing or recoding variables in preparation of later analyses.
John Bailer, PhD, is Distinguished Professor and Chair in the Department of Statistics at Miami University in Oxford, OH, where he is also an affiliate member of the Department of Sociology and Gerontology and the Department of Zoology, an affiliate member at the Institute of Environmental Sciences and a Scripps Research Fellow at the Scripps Gerontology Center. He received his PhD in biostatistics from the University of North Carolina at Chapel Hill, and he worked at the National Institute of Environmental Health Sciences before joining the faculty at Miami in 1988. He has used SAS for approximately 30 years.