Making science work

I wanted to kick off with a little bit of idle bloviating. Many folks have noted that academic science is in a bit of a crisis: concerns range from budget woes to postdocpocalypse to big-pharma malfeasance to rank scientific fraud. It seems to me two steps would greatly improve this situation.

First, provide a stable funding mechanism for academic researchers. I see a lot of folks whose science suffers because they spend too much time worrying about the next grant; conversely, I’ve seen several scientists’ labs grow beyond their desire because avoiding collapse often results in accidental and over-exuberant growth.

Second, require pre-registration of experiments. This can be implemented like the new federal (formerly just NIH) public-access policy, in which you must document following this procedure for work paid by government grants, or could go farther by requiring that all federal CVs list only publications with followed the policy. This seems to be the single biggest move for fighting the corporate-sponsor malfeasance, file drawer effect, and the widespread practice of publishing exploratory work as if it were a hypothesis-driven experiment. And it’s cheap and quick. Let’s do it.