/sci/ - Science & Math

>>15936987
not true at all, the strongest scientific evidence is "i made it up"

>>15936987
Much like the universe, evidence comes from nothing.

>>15936987
There is no formula for when a non-deductive argument is best. it all depends.

>>15936987
>How true is this?
Not entirely false, but badly missing the point.

In science as practiced by actual scientists rather than larpers, the high tier of scientific understanding is "coherent theory that explains a large number of observations with few degrees of freedom", and the low tier is "any number of studies, meta-analyses of studies, reviews, or whatever it may be". That is to say, everything on this pyramid is low tier, and the entire edifice of collected evidence is instantly outranked by any decently good theory that explains observations.

If you don't have anything like that, and you have to make do with observations alone without any actual understanding to guide you, the situation depicted in the pyramid is roughly correct. But that is a small part of science where little actual understanding lies.

>>15936987

There is no evidence of evidence.

>>15937044
/thread

>>15936987
this is could possibly be relevant in social sciences and fields like medicines, or nutritions where the standard of experiments is so sloppy or it is impossible to control for all variables, you need to average out the results from multiple research.
for hard sciences and formal sciences like physics, chemistry or mathematics, a single well done experiment or a novel framework in a paper trumps all.

>>15937044
also this

>>15936987
I'm yet to see any solid RCTs and meta analyses confirming the validity of RCTs and meta analyses. In fact, according to this meme pyramid itself, the scientific method itself lies at the bottom.

>>15937044
90 IQ take. The high tier of scientific understanding is "coherent theory that explains a large number of observations with few degrees of freedom", VALIDATED BY SOLID EVIDENCE. This chart pretains to what evidence counts as solid.

This pyramid is about medicine and psychology, not science.

>>15937044
so you mean the reasoning that intelligence is in part genetic so dumb people having more kids than smart people will result in dysgenics, outranks the claims of evidence that intelligence is rising? Good to hear :D

>>15936987
A carefully controlled experiment in non-human primates replicated by several labs is stronger than a p-hacked epidemiological dataset in free living humans.

>>15936987
For current state of academia is true.
I made it up gets all the money.

>>15937044
You’re a larper yourself. See >>15937885

>>15937885
>>15937935
Idiots.
Solid evidence has nothing to do with that graph unless your soience is a low tier data torturing hack.
High tier quality soience like physics one single experiment get reproduced by a couple of group and we done (for example the double slit exp or the aether exp). One Einstein or Newton eclipse the whole "scientific community".
Meanwhile, in low tier soience like medicine, you get the obvious fake alzheimer research the that the took decades to dig out. Or the flip flop every other year whether eating too many eggs or meat are harmful or not. Once upon a time, majority of medical "doctors" agree that smoking cigarette and chugging cocaine are good for your health.
Appeal to authority of the mass are behavior of low quality tier soience because they dont actually have any scientific theories except for concessuses.
I would go a step further and say those aint actually soience, they are just data torturing enterprises to achieve whatever conclusion that community want to sell.

If this pyramid is so great, why do the social sciences, where this pyramid is used the most, resulted in literally nothing useful for society nor advancing the sciences?
They also possible have the most frauds.
If your science looks like this pyramid, it's a dogshit field.

>>15937968
>a theory is better evidence than the evidence itself
You’re a pseud (as is evident)

>>15937988
nta, but evidence/experiments are basically worthless without a theory
see neuroscience for a good example, the amount of money/resources that has gone into the "connectome" is complete fraud

>>15937988
please see >>15937907
Peoples always harp on about how “evidence” shows were not getting dumber, despite the completely sound reasoning that dumb people outbreeding smart people must eventually lower the population’s intelligence.

>>15936987
>evidence
Who cares? You come up with an experiment and it's either replicable or not. If you want to abolish some theory, you come up with an experiment that proves it wrong. There is no "degree of evidence", it either holds up and models the reality as we know it, or it doesn't.

>>15936987
>Meta-Analyses as the highest
>"I read all the papers and it says my conclusion is right"
>"None of the 3 people, who were too lazy to read my paper, disagreed with me"
>"You have to accept this as fact now"

>>15938022
Lol do you think peer review works like that? Maybe in karachi university

>>15938259
>Lol do you think peer review works like that?
Lol do you think it doesn't?

>>15938259
Have you not done any significant peer review/publish papers?
Sometime it's exactly as that anon described.
Otherwise, how the fuck did many papers that copy paste their own figures get published in the most prestigeous journals?

>metaanalysis
Authors read through 200 papers that get published in this topic.
No shit, you ran it through a tokenizer and collect keywords, you didnt read anything.

>>15936987
>intellectual masturbation ranked higher than actually collecting real data from the world around you

>>15936987
>meta-analyses
>useful in any way
LMAO

>>15936987
Isn't the top just the bottom to the x power?