Reproducibility is the process of repeating analyses or experiments to confirm results, and is an essential component of the scientific process. As data and computation have become increasingly more sophisticated, reproducibility is even more important because algorithmic research is often impossible to verify manually or intuitively.
One important aspect of reproducibility is writing good documentation. Documenting the workflow, data, and code is important not only for the replication of the results, but also the communication of scholarly methods. It can also be challenging to remember all the details of one’s own work without documentation.
Version control is another way to facilitate replication of your results, by allowing others to keep track of changes that are made to data and files. There are several tools that can help automate the process of documenting the changes made, as well as when and by whom they were made, from tracking changes in Google Docs to Git, an open source software version control system.
Council of Library and Information Resources (CLIR) postdoc Yasmin Alnoamany created a Guide to Reproducible Research as a way for researchers to get started making their work reproducible. It teaches researchers how to manage the main entities of scientific research and effectively provide documentation. It also lists best practices and suggests the tools for managing data and software.