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

I like this.

Requiring reproduction is several bridges to far.

But requiring reproducibility seems complete reasonable!



If it’s not worth reproducing it’s not worth publishing imo


Until grants explicitly grant money for an independent lab to reproduce the research they are paying for (effectively almost doubling the grant amount), no one is going to spend their hard-won funds on reproducing someone else's research. At least not in a timely manner.

What such a requirement would do is dramatically scale back research, and highly incentivize researchers to lie, or the "independent labs" to fudge in order to keep drawing funding from their collaborators.


> Until grants explicitly grant money for an independent lab to reproduce the research they are paying for (effectively almost doubling the grant amount)

I think a lot of money would be saved by avoiding researchers to waste resources in dead ends, by trying to build upon results that can't be reproduced. So not exactly double.


I completely agree, but unfortunately it's not the same money that is in the initial grant, it's someone else's money. So by granting more money granting agencies can theoretically save some other granting agency's money, but not necessarily their own. Meanwhile that other granting agency will get a lot more bang for their buck (and quantitatively more prestigious results) than the first granting agency.

Bad incentive structures.


I agree we need to allocate at least 1/4 maybe up to 1/3 of all grant money to reproduction of research. While that seems wasteful it’s actually a huge savings and prevents people from wasting energy on false research


Is it a jobs program or are we investing money to get to the truth of things?


Granting agencies, whether government or NGO, typically want certain kinds of applied science results. I think "truth" is more assumed than the explicit goal.


So you just take their word for it? Pretend that reading about it is equivalent to seeing it?


Many scientists reading an article will seek to extend the work in the article, not duplicate it. If the extension fails then they may spend the effort to replicate the original results.

Even reproduction doesn't solve the problem of theoretical errors: https://news.ycombinator.com/item?id=36230450


Reproduction strikes me as the first significant step in extending and debugging. But that's just me.


Engineering and Science go hand in hand, but one is one and the other is the other.


That's the current system.

I suggest an improvement.




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

Search: