Reproducible research

"Conduct science with a particular level of reuse in mind." - Rule 3 of 10 Simple Rules for the Care and Feeding of Research data(1)

 

    Main points

In order to be able to repeat and verify research it is necessary to document the research process well. All material needed to reproduce the research (e.g. data, code and software) has to be publicly accessible. Reproducible research requires a thorough preparation in the planning phase. 

As a data supporter you will be able to advise a researcher on measures to take in order to start working reproducibly at an early stage. Expect the question "What is in it for me?" When asked, it may be useful to have testimonials at your disposal. How does reproducible research benefit researchers? The video below is a testimonial in which Patrick Vandewalle, senior researcher for Philips, shares his experiences with regard to reproducible research.


Set to HD quality for the best viewing experience.

 

Would you like to know more about Patrick Vandewalle and reproducible research?  

Five tips on reproducible research

In the book Reproducible research with R(6) the authors provide five practical tips on reproducible work:

  • Document everything!
  • Everything is a (text) file.
  • All files should be human readable.
  • Explicitly tie your files together.
  • Have a plan to organize, store, and make your files available.

  Case

The Reproducibility project(7) is an international project in which more than 250 researchers collaborate to repeat authoritative research. In fact, this project is a study to see how reproducible psychology experiments actually are. It drew a lot of attention and inspired reproducibility studies in other domains.

Click to open/close
  1. Goodman, A., e.a. (2014). 10 Simple Rules for the Care and Feeding of Scientific Data. Retrieved from http://arxiv.org/pdf/1401.2134v1.pdf
  2. VandeWalle, P. (2011). Reproducible Research in Signal Processing: How to Increase Impact. [video]. Download presentation from http://mediasite.mediagroup.ubc.ca/MediaGroup/Catalog/pages/catalog.aspx?catalogId=dec52c0c-13fd-4eb1-92b2-aa80a3a1d0d1
  3. Reproducible Research Repository. Retrieved from http://rr.epfl.ch/
  4. Runmycode. Retrieved from http://www.runmycode.org/
  5. Research Compendia, see notice on http://www.re3data.org/repository/r3d100010758
  6. Gandrud, C. (2013). Reproducible Research with R & RStudio. CRC Press. Retrieved from http://christophergandrud.github.io/RepResR-RStudio/
  7. The Reproducibility project, retreived from https://osf.io/ezcuj/wiki/home/

    Your additions

Do you know stories or anecdotes that prove that researchers benefitted from research with reproducibility in mind? Do you know about projects in which researchers replicated each other’s research? Or do you have something else you would like to add? Let us know and post a comment.


botMessage_toctoc_comments_9210