Thread by @lteytelman

I don't know about 50 years ago, but there's plenty of evidence that our methods sections today are typically insufficient for reproducing the published work. "We used a modified version of paper XYZ" and "contact author for details" are common. 1/

Key early lesson from the @elizabethiorns @BrianNosek Cancer Reproducibility project is that the protocols are often missing. 2/ vox.com/science-and-he… nature.com/news/cancer-re… theatlantic.com/science/archiv…

In his 2015 blog post, @tpoi nailed the state of methods sections: medium.com/@tpoi/do-the-r… 3/

Terrific post in 2014 by David Crotty (@scholarlykitchn) scholarlykitchen.sspnet.org/2014/11/18/nev… makes it clear that this is an old & serious problem. (David can be trusted on this - he worked for years on protocols for @CSHLPress & is now editorial director for @OUPAcademic) 4/

There's a reason tweets like this by @themorgantrail resonate.

External Tweet loading...
If nothing shows, it may have been deleted
by @themorgantrail view original on Twitter

5/

This is the frustrating norm today:

External Tweet loading...
If nothing shows, it may have been deleted
by @mbeisen view original on Twitter

6/

Poor methods sections aren't just an inconvenience; they waste time and money. Missing details delay the progress of science. We have the internet and can do a lot better. 7/7