St. James Ballroom
Data-Driven Programming Techniques in SAS and R (303890)
Jennifer Greer, ICF
Adam Lee, ICF
*Davia Moyse, ICF
Keywords: data-driven programming, SAS, R, solutioning
Many applied statisticians act as both data managers and analysts to complete technical tasks by first preparing the data, and then completing simple to complex statistical analyses. To address real-world problems and research questions, statisticians regularly repeat data preparation and analysis as new or updated data become available. To mitigate the need to update statistical programs as input datasets change, a data-driven approach to code development can be used. Data-driven programming can be thought of as specialized programming that reacts dynamically to the structure or content of the data and which minimizes manual coding for specific data features. This poster will present an overview of data-driven programming and discuss the benefits of this technical solution style. Example techniques, including specific coding language, for data management and statistical analysis in both SAS and R will be described in the context of survey data. We will also compare potential limitations of the data-driven approach in the two programming languages presented.