/sci/ - Science & Math

>>12769898
easy, just abolish social “““science”””. leave science to the real men

>>12769898
Isn't the whole point that studies should be reproducible?

>>12769983
yes

>>12769983
yes

>>12770046
>>12770055
So it's not actual science
What is it then?

government grants to replicate studies?

>>12770059
Phenomenology.
https://en.m.wikipedia.org/wiki/Phenomenology_(philosophy)

>>12769898
>how do we fix this?
We need another paradigm shift in the field of medicine, especially psychology

>>12770068
>Psychology has a paradigm to shift
Dumbass. There is no paradigm in psychology, unless you mean bullshit to validate whatever you feel is correct.

1) Why are so many studies not reproducible?
a) is it cheating? p-hacking?
b) mistakes?
c) deep-rooted protocol problems?

2) Can peer review fix it?

>>12770091
the answer to all these questions is that the perpetrators are social scientists and they just suck

>>12769898
Nuke india
Execute all social “scientists”

>social science BAD
brainlet take

>>12770137
India has little to do with corruption of the American Empire.

>>12770184
Only antisocial people hate social science.
Spergs BTFO'd, will they ever recover?

Because how science is practiced an exercise in fraud.

>>12770184
>defends social science
brainlet take

>>12770184
post a pic of your prettiest dress

>>12770183
It’s called over saturation you obsessed retard

>>12770183
Anyone in medical research will tell you India and China are like 80% of the replication crisis. I read a study out of fucking Sudan that was better designed than half the shit I see out of Mumbai.

>>12770316
There are a multitude of misaligned incentives in the United States. Publish or publish, corporate interests, and regulatory capture have only been getting worse. Deflecting blame is a convenient distraction from the real problems.

>>12770063
>en.m

Stop treating studies seriously until an experiment is replicated

>>12770184
>has to attach "science" at the end of the field name

I wonder why, since I always held nutritional, political and social sciences in high regard

>>12769898
this is not a problem in MINT
only social SOIENCES, stuff like gender studies, psych, sociology, med, etc. which are not real science anyways
when will (((media))) pretend that there is any credibility to claims of these pseuds?

>>12770082
Well you're kind of an asshole but yes, I agree with you. I was trying to be polite, but psychology is retarded.

>>12770601
Proof? Sounds like a massive cope.

In machine learning, it's more of a cultural problem I think. Most grad students just want to publish a paper at a top conference before they graduate. The idea and reproducibility is secondary. The primary goal for them is to beat state of the art model on one data set.

>>12769898
kick all nigger and women out of academia. Then 80% of admins. Channel all the equality and admin funds into geniuses. Make academia repulsive for smoothbrains (dont tell everyone education is the solution to life).

Basically scale back to a based, merit-based academe of the pre-WW1 Germany or post-WW2 America. Works every time.

>>12771597
But what about midget space niggers?

>>12771604
There are 1000 jobs that make more money

>>12771613
But the memes, think about the memes

>>12769898
It was inevitable when you have retarded arbitrary standards with the expectation of standard development and infinite growth i.e. in order to get a degree you MUST publish a paper oh btw you NEED a degree. Whoops now you have thousands of people who have nothing interesting to propose but have to anyways because bureaucracy.

>>12769898

It is because it is most important for success in academia to publish a study in a high-impact factor journal. It doesn't actually have to work in other labs.

So there is no negative feedback for BS science. Most important is it has to be a hot story.

So a paper called "single cell RNA-sequencing of sarscov2-infected femal transgender african american aborted fetuses" will be published at Nature without review because it's "hot" right now.

>>12769898
1) Abolish the soft sciences (anything that relies on surveys as primary data collection).
2) Ban China from academia.
There. You've cut down on this by 99%.

>claim the civil war happened
>Scientist tries to replicate it
>Can't replicate the civil war
>The civil war didn't happen.

>>12771795
The amount of stupidty in that argument

>>12771803
>Can't refute the argument

I hate these headlines.

>more than two-thirds of researchers have tried and failed to reproduce another scientist's experiments
There's surely a problem, but it's sensationalist bullshit if you count "not being able to reproduce at least one result in the entire lifetime of a scientist".

>>12771809
>*points to the african and asian wars of idependence*
They speak englih too

>>12769898
imagine showing a pic of chemistry / bio when EVERYONE knows that the replication crysis is caused by pseudoscientific fields like sociology and psychology.

>>12770063
kill yourself dirty phoneposter

>>12771795
>he doesn't know the Historical Method, nor how it is different from the Scientific Method
https://en.wikipedia.org/wiki/Historical_method

>>12770184
show tits roastie

Divide studies that haven't been replicated at least twice as "Unreplicated" and "Partially replicated"
At least 3 of the studies need to have different and independent sources of funding for it to be fully replicated.
>>12771795
>CIA please fund my american civil war replication
clownpilled

>>12769944
I don't think you understand what social science is

>>12770063
phoneposting sub 50 IQ retard

>>12770128
basically anything with greater complexity than physics there is trouble..
people don't understand statistics so they end out accidentally fucking themselves with frequentist bullshit..

>>12773257
My smoothbrain only reacts to things that make sense if you look at it and not if you have to interpret it

"man is the measure of all things" "no two men are the same"
Statistically only one of these statements is correct
it seems we have identified the replication problem

>>12769944
It used to be done by men, until feminazis took over. Ever realize how the psychology and biology of the 80s was based, and gradually shifted toward shit like "gender" being socially constructed, and now even sex being socially constructed?

>>12772062
Please don't embarrass yourself by defending that clusterfuck.

>>12773306
>now even sex being socially constructed
wat, I'm literally in a socialist progressive eceleb's tranny Discord and I've never heard of that

>>12769898
Not a problem in real sciences such as Biology, Chemistry, Physics and Astronomy.

>>12773388
kys.
https://arcdigital.media/is-sex-socially-constructed-81cf3ef79f07
https://thesocietypages.org/trot/2018/11/26/the-social-construction-of-gender-and-sex/
Of course, a based philosopher already tore all that shit apart.
https://philpapers.org/rec/ALEISS
Nonetheless, it persists.

>>12770128
>>12770601
>>12771779
>>12772007
>>12773498
>Not a problem in real sciences such as Biology, Chemistry, Physics and Astronomy.
Where the fuck is this "it's only a problem in psych/social sciences" meme coming from?
Replication is an issue in almost all sciences! Especially in medicine and biomedical research. Some pharmaceuticals scientists have said they could not reproduce results of the best cancer research papers published in top journals when they tried to confirm results so they can work on producing therapeutics.

>A survey on cancer researchers found that half of them had been unable to reproduce a published result.[63] A similar survey by Nature on 1,576 researchers who took a brief online questionnaire on reproducibility showed that more than 70% of researchers have tried and failed to reproduce another scientist's experiments, and more than half have failed to reproduce their own experiments. "Although 52% of those surveyed agree there is a significant 'crisis' of reproducibility, less than 31% think failure to reproduce published results means the result is probably wrong, and most say they still trust the published literature."[64]

>A 2016 article by John Ioannidis, Professor of Medicine and of Health Research and Policy at Stanford University School of Medicine and a Professor of Statistics at Stanford University School of Humanities and Sciences, elaborated on "Why Most Clinical Research Is Not Useful".[65] In the article Ioannidis laid out some of the problems and called for reform, characterizing certain points for medical research to be useful again; one example he made was the need for medicine to be "patient centered" (e.g. in the form of the Patient-Centered Outcomes Research Institute) instead of the current practice to mainly take care of "the needs of physicians, investigators, or sponsors".

Problem is fucking wide-spread. Quit repeating nonsense that it's only a problem for soft-scienses.

>>12773306
>that beautiful time in the early 80's where almost half of your fellow CS nerds were wearing skintight lycra
WE HAVE TO GO BACK

>>12773535
>a failure of some graduate students to replicate their own work once is equivalent to the replication crisis wherein over half of the studies in a field couldn't be reproduced

>>12773535
>Where the fuck is this "it's only a problem in psych/social sciences" meme coming from?
It's a cope from the IFLS! crowd as the replication crisis undermines their religion.

>>12770137
>he can’t into cliodynamics

>>12773535
Also, Mochizuki’s cult has shit up math

>>12769898
Most scientists can't even replicate themselves. All day they just focus on replicating 'studies' instead.

>>12769898
>picture suggest biology
>most of replication crisis is in humanities
Fucking jurnos

>>12770316
There is nothing wrong with sudan scientists

>>12773999
>Most scientists can't even replicate themselves

>>12770184
What the social sciences study are perfectly valid. The issue is that most social scientists are bad scientists.

>>12771597
Social sciences are now saying meritocracy and objectivity and components of white supremacy.
These are the people who are influencing public policy.
>>12771604
What about the gay niggers from outer space.

>>12773306
I'm starting to understand why engineers are so based.

>>12773513
>https://philpapers.org/rec/ALEISS
>Citations: "Are women adult human females?"
kek, thanks for sharing

>>12769898
>How do we fix this?
I get the validity of something relies on the reproducibility of the experiment or theory, but, doesn't that just dilute the entire thing? if it were that fucking easy it wouldn't have taken a scientist to figure it out? etc.?

>>12769898

there is no replication problem, there is a problem with people calling "science" what actually isn't and using the scientific method in an inappropriate way so they can bend the truth as they want.

>>12769898
Politically identity issue at play. Something like 90% of academia identify as liberal. Biases in paper is what cannot be replicated when those papers are put to test.

>>12769898
defund anything that isnt MINT (and maybe med.)

>>12770059
Jewish power

>>12769898
>Get funding for study
>Study shows results that are boring or not what your funder wanted
>Choices are to go hungry, change careers, or bullshit the data in various ways

>>12776723
>people
You mean universities across the planet?

>>12776784
wtf is a MINT

>>12776984
STEM in german

>>12776984
MINT is an acronym referring to the economies of Mexico, Indonesia, Nigeria, and Turkey.

>>12776960
>universities

I mean journalists and humanities spewing crap.

this is what happened

1) STEM is fine, people graduating make money, research goes ahead

2)Enter the "publish or perish dogma"

3)Need to inflate the amount of publications, regardless of quality, you need more people

4)In an effort to shove people into STEM there's a two sided approach : call STEM what is not STEM and lower the quality of output

5)Proceed then to complain that research quality has declined, blame the scientific method instead of retarded policy making and quantity over quality.

The replication problem is non-existant because you should not even publish a study which is so poorly made that is not replicable, it is one of the fundamental points of hte scientific method.

How do you know if a study is a good study?
You check if the math holds, if there's data manipulation, if the statistical analysis is proper, if the sample size is large enough, etc.
It is a meticolous process.

Ofc if somebody has peer-reviewed a badly done study somebody then in the future will use that study as a basis for another study and will get wonky results.

>>12777035
Your argument is basically
>everything is fine
>event happens
>now everything is shit
which isn't very scientific, to be frank.

You should've explained WHY '2)' — the event that hypothetically caused everything — happened and HOW.
Your post is based on presumed knowledge by the reader, I wouldn't publish it if I was a reviewer.

>>12777126

If a certain process A gives the expected result B and you induce a change in the process A leading to produce C it is obvious that the modification is the problem if C is not the desired output.

the problem is that the point 2) in my argument was not introduced in a "scientific" way, there is no metric which suggests that a net increase in bulk publications leads to an increase in quality.

One could make the reasoning that "since 95% of studies is not cash effective if we increase the bulk number of publications we will get more of that sweet 5% of good things which we value higly"

This implies that the quality and the monetarization of publications follows a knowable probability distributions and makes gross oversimplifications on what actually makes a good study "good".

>Your post is based on presumed knowledge by the reader, I wouldn't publish it if I was a reviewer.

Yeah of course, it is impossible to make an effective point which would pass peer review on the so called "replication crisis", the phenomena is not accurately described, the data is not separated by research fields and even if there's a separation by fields we lack a causal mechanism to explain what you have to tweak to improve/worsen the quality of scientific productivity.

What we do know is that the bulk increase of publications due to bad research ethics pushed by the "publish or die" has produced an increase of crap.

>>12777174
>no replicated evidence of the replication crisis.
Damn. Science wins again.