Making research reproducible: git, R, and org-mode

In this talk, I argue that to address the replication crisis in science our first aim needs to be to make our work reproducible. That is, reproducible for ourselves! Although replicability and reproducibility should be distinguished, the former depends on the latter. I identify three principles of making research reproducible: 1. Separating data from analysis; 2. using version control; 3. using code to analyze data. These three principles not only allow us to make our work reproducible; they also enable us to readily share our data and workflows with others. I introduce a working environment that is based on the editor emacs together with the powerful outlining tool org-mode, the version control system git, and the statistical software R, all of which have a strong and supportive online community and are widely used by computer scientists and researchers across the world. They are ideal tools to make research reproducible.