When we announced the SAS Global Forum 2013 Best Contributed Paper winners on our blog, the response was huge! I asked the authors to comment on whether or not their paper topics had any broader applications. The response was overwhelmingly positive. Seems like these winning SAS findings are Swiss army knives of the anayltics world! Check out the authors' responses:
Art Carpenter: “These techniques apply wherever the macro language is used.”
Macro Quoting to the Rescue: Passing Special Characters, with co-author Mary Rosenbloom
Beinan Zhao: “Most of my work involves doing data management and statistical analysis for clinical/public health research. Palo Alto Medical Foundation is one of the pioneers in the U.S. who adopted the Electronic Health Records (EHR) system in clinical practice and put data from it into meaningful use. As the number of healthcare facilities that use EHR system has been increasing worldwide, the methods we developed could be adopted by anyone who’s interested in using EHR data for purposes of quality assurance, studies of cost effectiveness or comparative effectiveness of different treatments or evaluations of patient behaviors, etc… The SAS algorithms can be adopted by other people in academia, pharmaceutical companies, consulting firms who have healthcare arms, health insurance companies or government agencies.”
Estimating Patient Adherence to Medication with Electronic Health Records Data and Pharmacy Claims Combined
Patrick Thorton: “The technique I presented might be considered whenever there are indicator variables where two or more may be ‘Yes’ at the same time. I would venture that every industry has this situation, and therefore, might benefit from combining the multiple indicator variables into a single variable for sub-setting and reporting.”
A Concise Display of Multiple Response Items
Stephen Overton: “The ideas presented in my paper are cross-industry and can be applied in any industry seeking to produce aggregate reporting based on ‘big data’.”
Escape from Big Data Restrictions by Leveraging Advanced OLAP Cube Techniques
David Scocca: “Code review could be done in any area where programming is going on.”
Communicating Standards: A Code Review Experience
Lucy D’Agostino McGowan: “We applied our method to diabetes, but other applications include other chronic diseases (e.g. asthma, hypertension, obesity, and cancer), and other outcomes (e.g. health behaviors and risk factors). Outside of public health, our general approach can be applied to multiple areas and disciplines.… Statistical analysis addressing disparities in health is an exciting area of research. While small area analysis is not a new technique, it can be applied in countless new areas, including disparities research.”
Multilevel Reweighted Regression Models to Estimate County-Level Racial Health Disparities Using PROC GLIMMIX with co-author Melody S. Goodman
Alec Lin: “Even though I used online sales data to develop the process, it can definitely be applied in other industries as well, especially in the era of big data.”
Variable Reduction in SAS® by Using Weight of Evidence and Information Value
Leonard Gordon: “Classification and regression trees (CART) are employed in a wide variety of areas. They are used as a non-parametric decision making tool.”
Using Classification and Regression Trees (CART) in SAS® Enterprise Miner™ for Applications in Public Health
Dylan Ellis: “Being able to write your own SAS functions with PROC FCMP is incredibly useful, with or without the special Run_Macro command (which allows that function to call an underlying macro). I think once more people become aware of this capability, user-written functions will be as ubiquitous as formats and macros. As for Run_Macro , my presentation is just an introduction to the technique – for more advanced examples of what PROC FCMP can do see the following papers by staff at SAS Institute:
- Stacey Christian (326-2010)– implements recursive partitioning of time series, an iteratively reweighted least squares algorithm, and an automated routine to fit a proportional hazards model.
- Jason Secosky (227-2012) – executes an SQL query from a DATA step, and also demonstrates how to retrieve images and driving distances from the Google Maps API.
- Andrew Henrick (129-2013) – uses hash objects inside FCMP functions – and as arguments to FCMP functions – to help streamline where statements and optimize expensive computations.”
Thanks once again to our paper winners who were kind enough to share their insights with us. If you’re hankering for more of their tried-and-true methods, check out: “Don’t trust the data: Advice from SAS Global Forum Paper Winners.” Their advice ranges from the nitty-gritty practicalities of how to select your paper topic to humorous quips like "don't forget the punctuation."
