Does pear review really work? How many researchers actually have the time and the resources to attempt to reproduce other researchers' results? I imagine most of the pear reviewers just read the papers and if they appear plausible, they are accepted.
>does peer review really work
Depends. When different labs around the work beforehand agree to all perform the same experiments for the sake of truth then it works.
If you do something on your own and publish then I doubt anyone would waste their time checking your results, much less replicating the experiment.
There is no glory if you are not the first, so if you are the only 'first' no one will give a shit about reviewing your thing. But if various labs agree to all share 'first' status beforehand then progress will follow.
Just ask Shininja Mochizooka, yokozuna of math of the japanese imperial academy of pure mathematics.
>Does pear review really work?
Idk man. I always thoroughly review my pears before I buy them at the grocery store, but sometimes I still end up with ones with mushy brown spots on them under the skin. Not a great procedure, 5/10.
I think that its mainly a problem in social sciences and maybe medicine.
I'm a PhD graduate student in chemistry and I reproduce other people's research constantly. For example: I read a paper with a cool molecule that I feel like attaching to one of my molecules. I follow their procedure as they wrote it, and if I don't get what they got, then something is really wrong and immediately noticeable.
Basically yes, if nobody can reproduce it then the authors either publish a correction to the publication after being notified, or it comes out that they just bullshitted the whole thing and they get fucking ruined. In the USA, Europe, etc. you don't come back from any kind of intentional bullshitting.
making one bullshit paper is not worth the risk. Sure, maybe nobody will reproduce your experiment at first to verify it, but if it has anything to do with a worthwhile science, someone is bound to base their work off of yours. Then, when they realize that the shit they're trying to achieve isn't working, it will clue them into thinking that maybe your paper was bullshit, and then recreate your experiment to see if it was.
I can only assume that if you are caught making up a bullshit paper, you'll be pretty much exiled from the scientific community. Is it really worth risking your entire career on something like that?
Dude, the peer reviewers arent the ones who reproduce the work. They just read it and look for obvious problems and mistakes.
Once it is vetted, other researchers can then attempt to reproduce the work. Maybe because they want to expand on the work, maybe because they doubt the findings.
Reproducibility in itself is becoming an issue as fields become more and more advanced, but in general science has not collapsed and peer review, while far from perfect, offers a real benefit to the scientific process and the accumulation of knowledge as a whole.
Peer reviewing is a joke. They think its magically %200 science proven just because your friend reviewed it. If thats true, it basically proves quantum, flat earth and consciousness elevator dimensions.
But anon how do you know it's not something wrong on your end? Maybe you're not keeping it anhydrous enough. Maybe your reagents are bad. Maybe it got stuck in your column, or decomposed. Maybe your technique simply sucked and you measured the acid wrong and decomposed everything/didn't add enough catalyst.
Also, I feel like people fudge their yields a lot. I refuse to believe anything but the shortest, most favourable one pot syntheses can give you 97% yield. Maybe I just need to git gud, but there's no way you could extract + column + recrystallise and still have yields that high. Yet they're reported all the time. And no one calls them out. Chemistry is definitely reproduced a lot more often than other work, but the field still has it's problems with that regard.
>Implying OP didn't mean derrière review.
If you don't get the same result then generally you repeat the experiment. If you consistently get a different result then you can start to build a claim that their procedure is wrong
And yield and conversion can generally be confirmed pretty conclusively via different analytical techniques via NMR or whatever. Even if you can't get 100% of your product out of your column, you can still easily see the % conversion via NMR.
I rarely see chemists claim yields as anything other than "quantitative", meaning the yield was as good as it's ever going to be feasibly
Combine the fact that researchers are more inclined to submit statistically significant results for publication, and that publishers are more likely to publish statistically significant results. You will end up with an over representation of "statistically significant" results. At a P-value of <0.05 approximately 5% of statistically significant data sets will be due to chance alone. With the bias in submission and publication there is a major over representation of results that are due to chance alone.
Studies of reproducibility in the social sciences have shown sometimes as many as half of the papers claiming statistical significance are not actually reproducible. This is due in part to the bias's in publication.
It is less of an issue in the sciences that are capable of using laboratory conditions and producing more robust data sets.
People can put out false results without ever realizing it, or doing anything intentional. It can still readily slip past the publisher and the peer review process. Independent reproduction of work is needed before it can be deemed reliable.
I'm sick of normies educated by pop-sci youtube videos acting like quantum mechanics is some newly discovered, untested theory when it's been verified and used in technology for the last century.