The replication issue continues to roil…

My friend, science writer George Johnson’s latest piece in the NYT, here. I think the most interesting notion is the idea that scientists strive for ‘exciting’ novel results (to move their professional ambitions forward) and hence bias what they publish towards what in fact isn’t robust.

NIH on the science reproducibility problem…

Story here [that’s a pun for those in the know about this issue]. Original comment in Nature here. I think some of the ideas are pretty good. I’m participating in an NSF workshop on this issue next month.

Cryptic explanation of the pun is here.

More on the replication problem in Science…from The Economist

The article is here. The figure above demonstrates the non-fraud side of the problem. Central to the problem is the concept of statistical power. A lot of published science studies are statistically underpowered. Kudos to The Economist for publishing a serious and understandable piece about this problem.

The "Replication Issue" as written up in the Economist….

The article is here. Their money quote:

“Peer review should be tightened—or perhaps dispensed with altogether, in favour [sic] of post-publication evaluation in the form of appended comments. That system has worked well in recent years in physics and mathematics.”

The Replication Issue is a complicated one and is something that I am actively working on with colleagues around the country. I am most definitely not for getting rid of peer review. BUt the post-publications comments might be a good idea to add on.