Reproducible Research in Daily Practice

• Activity Number: 289
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Consulting
• Author(s): JoAnn Alvarez*+ and Matthew Shotwell
• Companies: Vanderbilt University School of Medicine and Vanderbilt University School of Medicine
• Keywords: reproducible research ; good practice ; consulting statistics

Abstract

Reproducible methods can reduce errors and improve the quality of research. However, concerns persist regarding the practicality of adopting these methods on a routine basis or institution-wide. The Department of Biostatistics at Vanderbilt University has 27 master's level staff biostatisticians working on collaborative research with clinicians. The great majority of them use R, Sweave, and Latex every day to prepare data, run analyses, and make graphics, tables, and written reports. The department promotes the use of reproducible methods by providing computer support staff that are knowledgeable in these methods, developing a wiki code repository, offering opportunities for continuing education in R, Sweave, and Latex at weekly R/Sweave clinics, peer continuing education talks, and monthly computing talks. This presentation addresses how statisticians can personally implement reproducible methods by explaining the workflow involved, offering solutions for situations when a final report is requested in word-processor formats, and showing examples of reports. Finally, we describe how institutions can provide support for these practices.