Hacker News new | comments | show | ask | jobs | submit login

Well the one nice thing about scientific research is that it's open (cue conspiracy theories), and that you can usually read the paper/draft. No one is asking you to "just believe" research; that would be stupid, and it's not how the scientific process works. The fact that you are dismissing all research outright is ridiculous, and willfully ignorant. Your sources are shabby, and you talk of "science" and "scientists" as some great collective, which is frankly impossible to do. The idea that "Something is rotten in the core of science" is a wild extrapolation; there are problems with certain fields' practices (and may always be), medicine is probably a good example. However, saying that you give research no weight is stupid, and you discredit yourself further when you say that "Science has been supplanted by money and politics"; you sound like a conspiracy theorist, and I cannot believe you actually know many academics who will back you up on that claim. For whatever the system's flaws, there is _no_ reason to say that all research cannot be trusted.

> No one is asking you to "just believe" research;

Did you read the study the parent post is talking about? A well funded laboratory, that was trying to not "just believe" research (as everyone else apparently does) was trying to replicate these results. If the science was good, all it should have taken is time and money (both of which they had enough of). And yet, 47 out of 53 celebrated results published in peer reviewed papers of the highest caliber could not be replicated . Let that sink for a minute before you reply.

> there is _no_ reason to say that all research cannot be trusted.

Ok. Your reason to state that research can be trusted is that it is eventually replicated (thus confirmed), or thrown out (thus shown false), is that right? (You didn't state that as your reason, so perhaps you have other ideas -- but that's a common one, so I'll reply to it).

Assuming that's the case -- do you have any idea what percentage of results are replicated? And how much time after official publication?

Because if it takes e.g. 30 years until a bad publication is discredited, and (as the data point given by the parent shows) there are areas in which 90% of the data apparently can be discredited when you try to replicate it -- then, there actually might be reason to distrust research in general, because at any given point in time, more than 90% of non-discredited published results are wrong.

See also http://saveyourself.ca/articles/ioannidis.php (and the paper it references). This situation is not science fiction. 90% un-replicatable publications is probably limited to very few subjects. But 50% overall in medicine and biology is totally believable.

Which is not to say science (the abstract idea / discipline / method) is wrong - it's right. It's just that the things we human practice and often call "science" is very, very far from the ideal of science. Ignore that at your own peril.

Applications are open for YC Summer 2018

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