Why you should care about reproducibility
Markowetz highlights five reasons to work reproducibly (Markowetz, 2015):
- Avoiding disaster: By working reproducibly, you can trust your own research results and will not have to retract published results or keep publications back because you cannot reproduce your results.
- Writing papers easier: Well documented analyses ensure that you have easy access to the latest results, your work can easily be written up, and collaborators can easily get on board as additional authors. Furthermore, you can be sure that you easily comply with the highest-level journal guidelines.
- Convincing reviewers: Making code and data available to the reviewers means their review comments will be constructive as they are able to develop an in-depth understanding of your work and can even try changes to your analysis themselves and see the impact.
- Facilitating continuity of work: Well documented work means your work can easily be picked up and continued - either by others in your laboratory, or yourself if you want to build on your own work after a longer period.
- Building your reputation: Putting in effort to make your research reproducible shows that you are a careful researcher and makes your research results more robust.
Papers whose underlying data is available get cited more often (Piwowar, Day & Fridsma, 2007; Piwowar & Vision, 2013). All research outputs that are shared can be built upon more easily by others and in some cases, others following up on your work might lead to new collaborations. Other benefits of working openly are covered in our Open Research chapter.