Not Even $1 Million of Evidence
Celebrated as a good thing, it might actually represent bad news
The recent news that the Center for Scientific Integrity, the non-profit home of Retraction Watch, received a $900,000 grant from Open Philanthropy to stop flawed and fake papers from getting into health guidelines has been greeted as “good news” by various and sundry.
In many ways it is — or would be in normal times. One big boon is that it allows anonymous tips and keeps some things out of the public eye. Science has been disrupted largely because it has become a plaything to interlopers from inside and outside scientific disciplines who have learned to exploit its tech-infected and porous “open” aspects.
Called “the Medical Evidence Project,” it will be run by James Heathers, a popular Substacker and self-described science integrity consultant and “data thug” based in Boston. I like Heathers’ work generally. He has good instincts and skills. He also has that “Substack energy,” which may need to be curtailed for this project to be taken seriously where it needs to be taken seriously — not that many in those places are raising their heads in these dark days.
- The entire roster of vocal science integrity experts of stature might be sufficiently large to mount a decent afternoon foosball tournament.
Heathers likens this “metascience” project to “white hat hacking,” a reflexive appropriation of tech terminology for something far more intellectually substantive than traipsing through computer systems looking for flaws to report. Maybe it would be better to talk about science using clear language.
For instance, maybe the “Medical Evidence Project” is more about assessing the credibility, design, and interpretation of scientific claims affecting human health, and reporting on any that seem to lack the necessary scientific rigor to guide health practice? That sounds better to these ears — not as sexy, but this isn’t supposed to be sexy work.
And while there are good aspects to this, I think this is mostly bad news in disguise.