Post #3237035

2024-03-27 23:34 UTC

@nasamuffin@jorts.horse Of course studies also get cited directly: from court judgments and government policies to Facebook rants. A recent example are studies on abortion-inducing drugs that were cited in legal proceedings in the US. They were later retracted after researchers pointed out their (ideologically motivated) flaws. That illustrates that pre-publication peer review can fail to weed out bad work while post-publication review can be very efficient in highlighting flaws. 4/4

Replies (3)

@nasamuffin@jorts.horse 2024-03-27 23:45

@johannes_lehmann@fediscience.org thanks, that's really interesting! I think I had some very naive impression that the journal editors were checking for sound methodology, bias, etc. But it sounds like that's never really been the case? In which case they totally were just gatekeeping for cash, I could see that. I appreciate the detailed explanation!

Open ##3237036
@WestCoastCliche@beige.party 2024-03-28 01:06

@johannes_lehmann@fediscience.org @nasamuffin@jorts.horse Johannes, I agree with most, but not all of what you wrote. My perspective is different. I am a mostly retired academic cardiologist who authored or co-authored >400 peer-reviewed papers and sat on several editorial boards. I have never seen an editor decide to publish a paper because of money (but yes, sometimes if a paper is sensationalistic). I have little faith in preprints and have seen the conclusions of some studies toned down or changed drastically after peer review and "real" publication. Post-publication peer-review might work for basic science better than it does for clinical medicine. I would rather trust vetted reviewers than some guy off the street with an axe to grind. Having said that, I agree that Elsevier and the other big publishing houses are a rip-off.

Open ##3237039
@lilafisch@chaos.social 2024-03-28 08:06

@johannes_lehmann@fediscience.org @nasamuffin@jorts.horse having reviewed a bunch of papers and seen them come back with significant (if not massive) improvements, how would this be implemented? Would an author upload a newer version and somehow archive the previous?

Open ##3237041