[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 1.38 MB, 720x720, 1594520128469.webm [View same] [iqdb] [saucenao] [google]
12154060 No.12154060 [Reply] [Original]

>https://en.wikipedia.org/wiki/Replication_crisis
Do scientards really?

>> No.12154104

>>12154060
Great, jews are wondering how to make the next big scam after Evidence Based success.

>> No.12156041

>>12154060
There is a moral rot going on. Real scientists don't. Then again honest people don't get tenure. Therein lies the moral hazard.

t.Former researcher

>> No.12156048

>>12154060
>Do scientards really?
Why do you expect all scientific research to be replicable?

>> No.12156058

>he doesn't even have the brain capacity to read the fucking abstract of a Wikipedia article
the only legitimate one is medicine and it isn't something new. it's because of how many variables there is and how much shilling there is, especially in diet and sports. same could be said about really all biology that is related to humans. this becomes stronger and more apparent when there is a lot of new studies coming out

>> No.12156061

>>12156048
not all, replicability should be achieved in the expected values. Not publishing negative results is a huge problem, so is incompetence. But even worse is the frauding scum that simply lies to get published.

>> No.12156073

>>12156058
psychology is just as bad. And I can guarantee you the problem doesn’t lie with humans being too compex to make predictions. If you do your statistical analysis correctly and describe your experimental setup properly, you can expect successful replications near the confidence interval.
The problem is anyone can make up some load of bullshit that might sound plausible and claim they just happened to get an outlier sample when caught cheating.

>> No.12156097

>>12156061
Any human system is inherently going to not work not as entirely intended. There is no perfect system, but this one works quite well and I don't see any reason to change, other than to make it a little more strict. I see a failure of replicability as a good thing for the most part, not a bad thing. Confirming a negative is just as valuable as confirming a positive.
>replicability should be achieved in the expected values.
What are the expected values?

>> No.12156314

>>12156097
most of the time confidence is required to be 95% or 98%
how the fuck can you say the replication crisis is no problem when you have no clue about the basics? So fucking arrogant without the slightest hint of competence.

>> No.12156595

>>12156073
>social sciences
>science
kinda cringe bro

>> No.12156600

>>12156073
The replication crisis wouldn't even exist in medicine and nutrition if locking people up and observing and controlling every aspect of their lives during a study was actually possible.

>> No.12156619

>>12156314
>most of the time confidence is required to be 95% or 98%
According to who? According to what field?

>> No.12156629

>>12156097
> There is no perfect system
But the internet gives us the possibility of a better system than "peer review" bullshit.

>>12156314
How the fuck is confidence measured?

>> No.12156905

>>12156629
>How the fuck is confidence measured?
Very rigorously of course, infallibly.

>> No.12156917

>>12156905
How, not what is it like.

>> No.12156921

>>12156629
>But the internet gives us the possibility of a better system than "peer review" bullshit.
How?

>> No.12156929

>>12156917
>How
With lots of caution and effort to ensure its validity

>> No.12156930

>>12156921
Make a forum with upvotes or something. It's not infallible either but you get less bullshit than publisher cabal approved shovelwork.

>> No.12156941

>>12156930

The internet isn't a democracy you know, it's susceptible to the same kind of shilling bullshit but instead of the reputation of an academic institution pushing for the validity of a study you just got random internet strangers voting on papers. Instead of a replication crisis you'd just have a shitty popularity contest.

>> No.12156955

>>12156595
sure they are. You can’t get as precise as with physics, but you can get insight into various areas via experiemts. How thefuck do you think every medicine you take is tested?

>> No.12156970

>>12156600
oh sure, but go around asking for funds on studies where you lock people up to control for variables.
Ethics commissions already block you if you simply scare people and recommend to ask them to recollect a scary experience instead.
You obviously have a higher uncertainty in social sciences, but they are still scientific if not completely overrun by ideology. And that is not a problem with the science itself now, is it?

>> No.12156980

>>12156619
>>12156629
>how the fuck is confidence measured
https://en.wikipedia.org/wiki/Confidence_interval
holy shit you know minors are not allowed in this site, right?

>> No.12156983

>>12156941
you can confirm for degrees in the field, even specific research areas

>> No.12156990

>>12154060
>Clicked and accepted through the scope section of what you posted
>no actual sciences listed

>> No.12156993

>>12156905
if done correctly it follows from sample size and significance of the variable. The problem is that it is not done correctly, as we see failed replications at a higher rate than expected.
This is literally the first thing you learn in every introduction course in statistics. What kind of 3rd world shithole do you live in to not even have heard of that?

>> No.12157001

>>12156990
what is a science in your opinion?

>> No.12157010

>>12156930
Science is a not democracy you absolute retard.
Imagine allowing idiots on the internet to vote on science, lmfao. What the fuck are you on.

>> No.12157011

>>12156980
And this is applied in all peer review?

>> No.12157017

>>12156993
If it's done correctly, it'll get the job done correctly! That's all I know! You eggheads make it happen!

>> No.12157022

>>12157011
if not it’s on peer reviewers failing to check elementary properties of the data set. And if they are too retarded copy paste the fucking table and read the value, what are they even there for?

>> No.12157029

>>12157017
I study physics, you blown up moron. It’s not that hard to get a basic grasp of social sciences, and you even seem to be incapable of that much.

>> No.12157044

>>12157029
The only thing blown up about me are my wallet and my bratwurst down stairs! Now get back to work, number cruncher!

>> No.12157062

>>12157044
I honestly hope you’re not employed anywhere where you need middle school statistics.

>> No.12157075

>>12157062
I'm employed by myself and the stock market, as long as it goes up, I go up, and so do you, AS LONG AS YOU KEEP DOING YOUR JOB AND PRESSING THE FUCKING MATH BUTTONS, MATH JOCKEY

>> No.12157313

>>12157075
ah, so you’re a skillless parasite. Weird to be proud of that, but whatever keeps you happy and living I guess.

>> No.12157337

>>12157313
You aren't even half the man needed to run this company, Sarah! Get back to work before I snap your little bra!

>> No.12157461

>>12157337
huh? what skills do you need to leech other’s achievements and work? What the fuck are you even doing on a science and math board?
You seriously are proud enough to claim superiority over social sciences being fully aware that you’re too retarded to grasp even basic math?

>> No.12157477

>>12157461
I've got guts, a six pack, and confidence, short stuff, I played football back in college! You're just a maggot crawling through my shit! Now this is your final warning, get your ass back to work!

>> No.12157585

>>12156048
isn't that literally the point of science

>> No.12157589

>>12154060
>Affects social sciences and medicine mostly
Nice nothingburger

>> No.12157615

>>12156941
Oh, by no means. Science is a monarchy, isn't it? And the noblemen are dicksuckers at pearson, mcmillan etc. All hail the pure blooded peer reviewers and their circle of friends, gatekeepers and absolute judges of science, if it weren't for them the world would be in darkness.