4 ms·
> Retractions remain exceedingly rare for scientific papers: Just eight out of every 10,000 are retracted, according to a Retraction Watch database. This to me
by wand3r 3y ago
> Retractions remain exceedingly rare for scientific papers: Just eight out of every 10,000 are retracted, according to a Retraction Watch database.
This to me is a crucial statistic. Is this number much less than it should be? Or are many more papers dubious in their methodology?
- LordShredda 3y agoWhat's the rejection rate? Maybe it's low because peer review tends to block them from being published
- VHRanger 3y agoThe replication crisis and data colada showed there's tons of dubious science. About half of papers results don't hold up to replication.
- wolverine876 3y agoAbout 1/2 (1/3?) of papers in a specific field, when replicated, produced similar results but less strong statistical measures of them. It's fun to hear people complain about accuracy in this way.
- korginator 3y agoAdd this point from the article for context: "This retraction is Tessier-Lavigne’s fourth in as many months, a stunning turn of events for a researcher of his stature." That begs the question, how did at least four such cutting-edge research works with fabricated data and results go through the review process unscathed?
- ekianjo 3y agoYou just discovered that the peer review process is a big farce. Well done.
- wolverine876 3y agoThere's no evidence of fabricated data, afaik. (That's your own fabrication, apparently.)
- bryan0 3y agoHere’s some of the evidence of data manipulation: https://stanforddaily.com/2022/11/29/stanford-presidents-research-under-investigation-for-scientific-misconduct-university-admits-mistakes/ https://stanforddaily.com/2022/11/29/stanford-presidents-res...
- wolverine876 3y agoThanks. Clearly there are (or were at the time of the article) serious allegations that it is a possibility.
- wolverine876 3y ago> Is this number much less than it should be? Maybe it's more than it should be, or just about right. If we are just going to assume our biased preconceived notion is correct and the conflicting data is wrong, why have data? But if we use that method, what is the likelihood that our claim will eventually be retracted?
- ekianjo 3y agoSince most stuff is never replicated it should be much higher than that.