Hi there,
I just gave a short introduction to Reproducible Research and Open Science this morning to new PhD students at Inria Rhone Alpes. Slides are available here:
https://github.com/alegrand/SMPE/raw/master/lectures/talk_21_10_13_Grenoble.pdf
No solution provided, just horror stories, explanations about what’s happening, and pointers to learn how to improve methodology.
I thought it may be the occasion to share a few links here as well:
- The Reproducible research: Methodological principles for a transparent science MOOC is still open with more than 10,000 registered people. It’s still time ! We’re more than a year beyond schedule for Episode 2 but Christophe, Konrad and I are making progress.
- The Turing Way: A Handbook for Reproducible Data Science (a nice community work)
-
Software Carpentry: their turorials are very good, lightweight, and pragmatic.
- A forum to share such resources. At some point, it will be probably be replaced by a French version of https://www.ukrn.org/.
Now, most of the previous references address computational aspects and leave statistics and experimental aspects aside. We listed a few interesting references in the Vers une recherche reproductible: Faire évoluer ses pratiques bibliography but I’d recommend thes two ones:
-
The Irreproducibility Crisis of Modern Science: Causes, Consequences, and the Road to Reform by Randall and Welser, National Association of Scholars. 2018. Super fun to read!
-
Reproducibility and Replicability in Science, A Consensus Study Report of The National Academies of Sciences-Engineering-Medicine, 2019. Maybe people will finally agree on the replicable/reproducible/repeatable terminology!
Or more precisely, I think I already had been pointed to it but hadn’t found the time to read it. Anyway, this was really a crazy brainstorming experience.