[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y ] [Search | Home]
4Archive logo
>Casually hear a senior professor say...
If images are not shown try to refresh the page. If you like this website, please disable any AdBlock software!

You are currently reading a thread in /sci/ - Science & Math

Thread replies: 53
Thread images: 6
File: image.jpg (15 KB, 316x202) Image search: [iqdb] [SauceNao] [Google]
image.jpg
15 KB, 316x202
>Casually hear a senior professor say this to a post doc: "just change the units and add some data until it fits the regression line. No one will notice!"

Why does fraud in academia persist?
>>
>>7835392
Subpar researchers need to eat too.
>>
publish or perish
>>
>>7835392
Academic career these days is sheer murder. There are far more newly minted researchers with a PhD than there are rsalaried research positions and even fewer open positions for tenure.

Add to this that universities knowingly loos away when peole enter publishing cartels to pump up the statistics and you are approaching the rot that is academic state of today.

So you get Retraction Watch and people out for revenge like Francis Claire, a pseudonym for someone hunting down and exposing frauds.
>>
>>7836935
Honestly though lab rats don't really get tenure nowadays anyway so I hardly care if they cheat.
>>
>>7835392
when I accidentally met him duing his phd thesis, one of my former colleagues during my master thesis said to me "my supervisor had told me 'your graphs have to significant' so I made them significant". when I met a former fellow student on an open air rock concert, she told me "omg I hope noone ever finds out I copypasted three quarters of my master thesis"
even my phd thesis supervisor omitted controls and did very bad lab practice.
this fraud is fucking everywhere.
>>
Isn't science about peer review? Shouldn't it have been snapped up before publish?
>>
>>7836963
AHAHAHAHA
Listen bro, if something like climatology exists, you know peer review has gone down the drain
>>
>>7836963
peer review is horrible. nobody controls shit. the reviewers are pig lazy.
>>
>>7836963
>implying anyone actually reproduces experiments
>>
>>7836963

No. Peer review is just a giant jerk off between researchers.
>>
hey a little data here and there nobody hurt anybody
>>
>>7835392
Should've become a mathematician then. No data to fudge.

Although I imagine mathematicians probably lie in different ways.
>hey I need this lemma
>shit I can't prove it
>I'll just cite EGA without a page number
>problem solved
>>
>>7835392
Tenure
>>
huh. Just started a masters program. I would like to think my advisor doesn't do this. He's really involved and on multiple review committees. It's also medical related research which I feel would be more risky to fudge than something like physics or engineering.

I hear a lot of shit about research on /sci/. I'm interested to get started to see how much of it is true.
>>
File: 1454473459792.png (423 KB, 490x684) Image search: [iqdb] [SauceNao] [Google]
1454473459792.png
423 KB, 490x684
>>7836991
This is why computer science is best science. Want to reproduce an experimental result? Get their code off github and run it yourself.

This makes it more of a 'hard' science than physics desu. Such a shame CS students are all retarded plebs.
>>
>>7837457
until you realize their experiment was run on a Cray and your piddly little desktop will take like 40 years to replicate it
>>
>>7836988
>the reviewers are pig lazy

My mother in law is a geologist, with a PhD in physics. She┬┤s also an authority on geostatistics, has published books on the subject, given classes at universities for the last 30 years (grade, postgrade, private courses for mining and oil&gas companies, etc).

She's a reviewer and editor for a couple of journals, and she's made quite a few enemies in several different countries by detecting and exposing shitty research, faulty statistics, crypto-babble and mumbo-jumbo-saturated nonsense papers.

There is some hope out there. Just a little.
>>
>>7837414
Outright lying is more of an issue in applied sciences than pure ones, because data there is usually much harder or sometimes impossible to verify.

In theoretical disciplines like mathematics or philosophy, or in things like theoretical physics or economics (basically things without numerical data) you won't see much lying but you will see a ton of intentional obfuscation.

Creating non-trivial work is very demanding so a lot of people opt to create such an impenetrable wall of gibberish that no one will put in the work to notice it's meaningless.

That said, obviously not every researcher is like this. Lots do actual work.
>>
File: 1451258582752.jpg (115 KB, 891x960) Image search: [iqdb] [SauceNao] [Google]
1451258582752.jpg
115 KB, 891x960
>>7836963
Peer review is great in theory but here's what actually happens.
>send it out
>no one reads it
>get the ok back in like a month
OR
>send it out
>someone reads it
>take a fuck load of time to respond because they're preforming the same experiment as you
>scoop your work
>other guy sends out his shit for peer review without you
>his work gets published first and you're shit out luck
>>
>>7835392
>>7836935
>>7836957
>>7836988
>>7836991
>>7837318
>>7837360
>>7837535
>>7837545
How do we fix this?
>>
>>7837485
Until you realize moore's law and buy an equivalent processor next month. If physics advanced by moore's law we would almost have fusion energy now.
>>
>>7837650

You do know moore's law will plateau in our life time, right?
>>
>>7837659
Yes the law with plateau. Silicone chips will plateau. That doesn't meant that computers are going to stop becoming faster.
>>
>>7837485
Amazon ec2 instances are pretty fucking cheap these days.. Most deep learning papers can be replicated on one in less than a day for a few dollars. Heck, most deep learning papers can be replicated on a decent consumer GPU in a day. Still a lot cheaper/more reliable than having to buy an interferometer or w/e to replicate the average physics experiment
>>
>>7837643
>>7835392
When you encounter this kind of stuff, call them out on it.
Did you call him out, OP? No? Then don't complain about it. You had an opportunity to fight corruption and you didn't.
>>
>>7836935
Oh boo hoo.

Its survival of the fittest pure and simple, what do you expect.

Atleast the frauds always get weeded out eventually in STEM areas because if the research has any significance it will be investigated by many others to be validated. If its not significant, it'll just be forgotten and never read.

But in social sciences it can award you Nobel Prizes *cough*economics *cough*
>>
>>7837497
Aren't the reviewers generally anonymous?
>>
>>7836939
>lab rats don't really get tenure nowadays anyway
OK. Then who gets tenure?

>>7837643
>How do we fix this?
no easy fix. The risk of being found out must increase, sharply. It would be good if we could adjust the moral compass but realistically that takes more than mastering CRISPS. The emission fraud scandals show this is something which is part of our times.

And when fraiuds are caught the justice must be swift and brutal. Fate catches up with everyone, they say, but it can take too long time.

>>7837769
>Its survival of the fittest pure and simple, what do you expect.
Except not. It is survival of those willing to commit fraud.


>Atleast the frauds always get weeded out eventually in STEM areas
That can take way too long time. Schoen was going great guns for years before being found out. Retraction Watch shows many are well established before being caught and Sudbo was a professor heading an entire research group!
>because if the research has any significance it will be investigated by many others to be validated.
Wrong. Have you really not heard yet that reproducibility is already a huge problem??

>If its not significant, it'll just be forgotten and never read.
And it can still be enough to get that PhD, that tenure and that professorship. While the honest guy is stuck in poverty as a RA for years.
>>
GET HIM

GET HIM GOOD

THEY CHEAT THEIR WAY UP, AND YOU FAIL EXAMS BECOUSE OTHERS CHEAT

GET HIM
>>
>>7836963
It is an open secret that it does not work. Moreover there are known cases where the peers tried reproducing the results while delaying the reviewing process.

In China publishing is an industry and some of these things are pure fabrication. In some countries universities get money based on publication count and guess how that is going. Hint: academia no longer has any incentive to stop fraudsters amongst their own staff.

We truly have an enormous problem with ethics.
>>
>>7835392
>No one even considers the possibility that OP took this out of context.

Lets examine it:

>just change the units and add some data until it fits the regression line

The professor is supposed to aid the postdoc on his research. I'm sure that they've had tens of meetings about this and by now they both have a clear idea of what the final product will be.

The professor simply noticed that the postdoc made a mistake when representing the data so he just adviced him to make the changes so that his graphs will also represent their findings because the data naturally would look like that but the postdoc maybe fucked up.

>No one will notice
What if then, the professor made this sarcastic joke? The professor adviced the postdoc to make some changes to the data, and knowing how fraud is something that happens, decided to make a light hearted joke about it with his postdoc, which he likely knows for over half a decade by now so it is possible that they are close enough to make these pseudo dark jokes with each other.
>>
>>7837769
>at least frauds get weeded out
How do you know?
>>
>>7838009
>Aren't the reviewers generally anonymous?
No, and absolutely not in the case of book editors.
>>
>>7838220
>And it can still be enough to get that PhD, that tenure and that professorship. While the honest guy is stuck in poverty as a RA for years.

hate to break it to you buddy but the world is NOT FAIR.

the weak will always be exploited by the strong
>>
>>7838275

I found the undergrad!
>>
File: 1400729887878.gif (561 KB, 500x295) Image search: [iqdb] [SauceNao] [Google]
1400729887878.gif
561 KB, 500x295
>>7837643
There's no simple fix as I'm a bunch of other people said. The problem is that if I were to start raising hell then people would begin to look more closely at my work. Then they'd start to say this doesn't look right and would start to give me a ton of shit.

I think maybe you should have to actually show your work to a panel or something of that nature and have them reproduce the results. Then again that would be very difficult.
>>
>>7837666
not if people keep writing shittier algorithms because "lol just buy a faster cpu"
>>
>>7835392
Because publishing has stopped being about furthering knowledge and learning and more about strengthening your CV.

If a publication is inbetween you and your dream job, anyone is going to fudge the numbers a little bit if it garuantees publication.

Hell, I'm going to apply soon to a medical residency and I'm publishing some papers and going to work on an investigation. Like hell I'm going to let something stop me from publishing that and not letting me get my dream residency.
>>
>>7838462
is there a problem with publishing inconclusive?
>>
>>7838220
>OK. Then who gets tenure?
Applied theory/computational guys.
>>
>>7838469
Nothin inconclusive gets published in a reputable magazine, especially if you are a nobody who's just starting out.

Most of these crap papers can be avoided wirh a good selection of topic and data, though. There's that.
>>
Is this kinda shit legal?
>>
>>7838500
I doubt one gets jailed but a lot of old people with beards and committees of grumpy bureaucrats will do their best to not let you do research anymore
>>
>>7838476
What exactly do you mean with "applied theory"?
>>
>>7838589
Translating autism to actual useful applications in your field.
>>
>>7835392
>"just change the units and add some data until it fits the regression line. No one will notice!"
things that didn't happen and things that nobody but faggot OP said general?
>>
>>7836977
nice >>>/pol/ memes faggot
>>
File: 1453709790967.jpg (60 KB, 700x500) Image search: [iqdb] [SauceNao] [Google]
1453709790967.jpg
60 KB, 700x500
i'm interested in condensed matter physics. if i do a masters and try to RA it will i deal with similar situations? i always like going over new published theses regarding it, and i tend to write a ton on my own time, but can i get scooped if i'm not careful? is the demand for researchers in that area already saturated? what should i do?
>>
>>7838275
>hate to break it to you buddy but the world is NOT FAIR.
I am painfully aware of this already. I was just informing OP about this and, importantly, how it works. You seem one of the reasons I was steamrolled was that I was not fully aware of how the system works. I am very much informed now, thank you.

>the weak will always be exploited by the strong
There was very little Nietzsche about this, rather it was all about a willingness to drop all ethical norms to take the place of those who actually did the hard work.

This has direct consequences on health for frauds in pharma and medicine.

>>7838476
>Applied theory/computational guys.
Nice try. I did both my degrees in applied fields.

>>7838500
>legal?
Nope. there are lists on Wikipedia and the sanctions.
>>
Grow you dumb fucking autists, this is how real people without mental illnesses and the real world works.
They know what's important and what's not.
>>
>>7839040

Actually your thing is demonstrably bogus, and not even for moralfag or "intellectual reasons". Propagation and acceptance of untruth causes corruption (of all kinds), and civilizational rot over the long term.

So what is your thing, your argument? It's a very short post, but it's easy to ascribe some dumb normie version of your intent: "I'm listening to foolish, literal autists who care about honesty, when the winners are the people who market themselves effectively, tell a lie at the right time, etc. That's how fucking business works, these fucking people." And on an individual level, some version of this is true (they call it "game" these days), but (more importantly)...

When the /broader society/ accepts sufficiently flagrant lies, or the need to lie, particularly bald-faced lies about quantity, reality catches you downstream and reks everybody. The Great Leap Forward is the appropriate object lesson here: everyone lied about their quotas to avoid execution, so they thought they had more production than they really did, and millions starved. /For want of accurate data on which to act appropriately/. A culture of lying led to very real human suffering, compounded downstream by the earlier "necessity" of the lie.

Corruption, more common in Africa and Asia (India), also leads to objectively lower-tier civilizations, as is obvious.

https://en.wikipedia.org/wiki/Great_Leap_Forward

The whole subject of this thread recalls the above quota-hell: publish or perish - publish for publishing's sake, damn you, we have an institutional dick-length at stake.

There's a place for marketing and presentation, but anyone with any kind of business or administrative acumen knows that you can't lie about the numbers. Even the sociopathic CEOs who are actually competent in their roles know this perfectly well. Or if you do, and enough times, you may expect mission failure downstream.
>>
>>7839040
OK, so you would be happy to take Vioxx?
Thread replies: 53
Thread images: 6
Thread DB ID: 512880



[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y] [Search | Home]

[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y] [Search | Home]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the shown content originated from that site. This means that 4Archive shows their content, archived. If you need information for a Poster - contact them.
If a post contains personal/copyrighted/illegal content, then use the post's [Report] link! If a post is not removed within 24h contact me at [email protected] with the post's information.