Online Program

Return to main conference page
Friday, February 16
PS2 Poster Session 2 and Refreshments Fri, Feb 16, 5:15 PM - 6:30 PM
Salons F-I

Reproducible Research Implemented Through Version Control Systems (303678)

View Presentation View Presentation

Christopher Barbour, Montana State University 
Kenneth Flagg, Montana State University 
*Lillian S. Lin, Montana State University 
Andrea Mack, Montana State University 
Michaela Powell, Montana State University 
Tan Tran, Montana State University 
Stephen J. Walsh, Montana State University 

Keywords: reproducible research, version control systems (git)

Statistical consultants must share and stay current with the latest versions of project files. Version control software, coupled with tools for reproducible analyses, can support a networked system to ensure accurate reporting. Such systems fulfill this role in the software development community as standard practice for programmer collaborations. A user checks out the project from a central server, works locally, then sends their changes back to the server for others to sync with. At MSU Statistical Consulting and Research Services (SCRS), Gitlab server software is used to manage projects. Reports and client documents are stored securely on the University's server, allowing team members to easily find, follow, and reproduce each other's work. This system is particularly useful when collaborations with clients intermittently span multiple months or years or projects are reassigned within SCRS. By presenting challenges and advantages of our standard operating procedures, we demonstrate how consulting groups can adopt version control software to improve workflow and a file structure that lends itself to reproducible analyses.