Oddbean new post about | logout
 Cool, when the journal does not deny something being being disproved by more studies, and retracts the paper, aye? I wish we all could be so science-, not ideology-based.

Also, that's why I trust meta-analyses more than single studies. 

If you want to hear sources with no history of fraud, I can send them. Personally though, I don't think that one shitty point 26 years ago proves something to be not trustworthy forever-from-now-on — Signal was hacked once, but we don't think it's now always vulnerable, or furthermore sponsored by CIA. Tor was traced once, same shit, etc etc. 
 it wasn't 26 years ago.  I'm talking about during covid, the Lancet published fraud on drugs that were competitors to the vaccine
 
 Ah! Pardon my mistake, I thought you talking about the 1998 fraud where they claimed false link between vaccines and autism.

Changes almost nothing about my message tho:

Still cool, when the journal does not deny something being being disproved by more studies, and retracts the paper without denialism and ideological protecting

Still that's why I trust meta-analyses more than single studies. 

Still, if you want to hear sources with no history of fraud, I can send them. Perhaps two mistakes in 25 years is actually too much for a journal to be trustable.