I've heard about this before as a strategy to attack scientific studies/results that you find unlikely, unpluasible or just inconvenient: theerrorbar.com?e=39 (podcast)

@theerrorbar.bsky.social

Thank you for your review! I just learned about the @prereview.bsky.social platform. Creating an open peer review ecosystem makes sense to me. It fosters reception and interaction over the publish-and-forget culture...

Feyerabend is always a great and entertaining read. But beware of his own errors and ‚perspectives‘. I heard that many of his historical analyses are not very well regarded among historians.

This is a great piece addressing an important question. I‘d imagine that peer review for software might be different from one for (shorter) code as purposes are slightly different. Software must be functional and reliable, code must be correct with respect to a particular purpose. Though, it‘s fuzzy

That’s a good article but introducing decades of research about the problems of JIFs with „Critics say that…“ is a bit disappointing.

For example, check the work on knowledge channels that shows how seemingly irrelevant knowledge shapes the way how researchers and inventors think. Put otherwise, even some firms hire basic scientists to build absorptive capacity.

Sociology of science has shown us how the idea of a linear innovation (research x has impact y on society) represents only a fraction of human intellectual activity. Thus, SoS tells us that the very question for „any advances“ is misleading.

This will be a surprising turn in the course of natural evolution!

In manchen Ländern werden Algorithmen menschlicher, in anderen werden Menschen algorithmischer.

Overdeterminism!

At least you don't have to bring your own cookies!

Yes, we need more questions!

We all have such anecdotes. But I am still curious about the causal relationship. Accepting or rejecting a paper can depend on so many different factors and impressions.

Ah, das klingt nach einer legendären Prokraistnationsmethode: Man hat soviele Aufgaben auf dem Tisch, dass man die eine mit der anderen verprokrastinieren kann!

I just learned that there the R package "wakefield" to generate random data. We could make a RFK package to randomly retract studies?!

I'd say it's better to encourage people to look for the complexities themselves than to directly point at the issues and problems. This might sound very optimistic 🙃