lrwyatt
commented
7 years ago Jeff's Points:
The question about scientific reproducibility is a matter of how not if. The salient points of this chapter about data sharing are (i) that data handlers have a lack of training about sharing and (ii) that our cultural and economic climates do not incentivize data sharing. The citation/publication process is slow and doesn't allow for complete protections and other currencies to acknowledge contributions to science over this timeline.
My own thoughts:
1) The notion of a lack of training on handling data resonates strongly with me. Too much "learning on the job" required. Figuring out all of this 5+ years into research career is too late in the game for success in data sharing leading to reproducibility.
2) I agree with Jeff Leek: give data producers a break! Open is good (some inherent problems with privilege there too) but the system seems to have researchers set up to fail and now using sticks/policies to "encourage" them to do better. Very few data producers actively work against data sharing or are trying to be bad guys.
2) Much of our problems would be alleviated by economics. USA needs to value and invest more in scientific training and research activities. Grant application process is problematic, academic scientists are underpaid and over-worked, and the resources that they require are not easily attainable. I experienced this from grad school, through my postdoc, and even now at the faculty level.
laderast
Leave your comments and concerns about data sharing.