Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

"Peer review" is not "have someone else re-do the experiment". That's just not feasible, especially since reviews are done without pay. It's not realistic to expect people to spend more than a few hours reviewing a paper. That amount of time is barely enough to check for overall conceptual issues and maybe flag some really glaring deficiencies. (And then conclude with 'accepted with minor revisions', those 'revisions' preferably being 'add these three citations to my paper, that'll push your paper into 'acceptable' territory'.)


But there were glaring deficiencies in the stats. Anyone reviewing the papers should have caught them.


Well yeah in the papers of the OP maybe, I don't know. I more meant to address several commentors in this thread that seem to think in general that peer review is 'redo the research' and/or 'validate that it's correct'. It's not.

Nowadays when you see articles results of new research of covid19 in the media, those articles often include 'hasn't been peer reviewed yet' or 'reviewed by other scientists' or any such verbiage, either as a disclaimer or as 'now it must be true'. But that's not how it works; it's not because something has been 'peer reviewed' that it's 'The Truth' or 'Real Science'. Peer review, in reality, just weeds out (most) quacks (although in the OP's case it seems it didn't even do that) and checks that the paper is not completely out of touch with what is happening in and known about the field. It's not QA of the work itself.

(I don't care to debate if it should be, and if more money should be spend on replication etc, just providing some real world context on something that is quite opaque to and often misunderstood by those not in academia)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: