better training, slower and more careful science, shifting attention away from results, focusing on theory and models, etc. It's not about truth commissions or cleaning up science from bad scientists. It's about all of us learning to do things better, more carefully, gradually, incrementally. 4/N

From this perspective, it's not authoritarian oversight that's missing. In fact, that would be detrimental to scientific autonomy. Viewing autonomy as bad and the primary cause of fraud is a paranoid, pessimistic, and defensive frame I cannot get behind. 5/N

Completely agree, but I think it’s gonna be hard dismantling the incentives that drive the opposite. I’d agree science needs new processes, and renewed focus and investment on the small/slow, to get back on track. How to get them accepted is the problem.

I agree we need slower and more careful science that is theory driven. One problem is that our systems of promotion and tenure do not reward these behaviors. Nor do grant review panels. The incentive system is misaligned with the production of good science

For the particular kind of science I tend to do, preregistration works *iff* it's a practice that supports slowing down, being carefully, making incremental steps and thinking about the theory behind what I'm proposing to test experimentally.

Right - the institutional norms and incentive structures facilitate poor quality science and fraud. Getting rid of fraudsters just makes more space for more shoddy work to occupy