/sci/ - Science & Math

All climatology is founded on correlations.

>>9655040
>correlation does not imply causation

But doesn't correlation actually IMPLY causation? Shouldn't it be,

>correlation does not confirm causation

?

>>9655040
That was pretty good.

>>9655048
That was just a little obvious, but I would bet it still works.

It's not a fallacy, but sometimes the argument itself is in the data and you need a degree of knowledge to fully understand it. The problem if you link to the article to anyone, it's probably that he will avoid reading it thoroughly, and just use that as some sort of fool proof argument
>>9655053
Mathematically it isn't even necessary for causation, that's why you need proper analysis in your data.

>>9655053
No, because correlation does not give you the first clue as to which way the arrow of causality points, if it in fact exists.

It would work better the other way -- lack of correlation implies a lack of causation.

>>9655059
https://theincidentaleconomist.com/wordpress/causation-without-correlation-is-possible/

>>9655048
>climatology
not science or math

>>9655063
Like anybody on /sci/ is going to look at a source with "economist" in it.

>>9655040
Correlation makes causation more plausible

>>9655092
Check any stats book then retard.

>>9655040
the logical response is "correlation is not causation" and an agreement that statistical data does not provide functional evidence of theory. However, it does clue an investigator of areas to continue their research because correlations imply some link.

>>9655054
>I would bet
there is no wagering at 4chan, Grandpa

>>9655053
>But doesn't correlation actually IMPLY causation?
No.
There's a correlation between cold weather and increased spending on consumer goods in a lot of places.
Did the cold weather cause increased spending? Did the increased spending cause the cold weather?
Answer is neither, it's just coincidental to Thanksgiving and Christmas happening when the weather is colder and being holidays where people traditionally by each other gifts or food / goods to entertain family coming to visit.

>>9655040
>correlation does not imply causation
It's not a fallacy m8

>>9655040
https://www.lesswrong.com/posts/NhQju3htS9W6p6wE6/stuff-that-makes-stuff-happen

>>9655048
>All climatology is founded on correlations.
How?

>>9655040
No, but it is more probable as both factor and consequences keep being repeated
Inductive method
learn2science, bitch

>>9655167
>it's just coincidental to Thanksgiving and Christmas happening when the weather is colder

But Christmas originated in celebrating the winter solstice, which is a direct causal link.

>>9655040
>What's the logical response to the "correlation does not imply causation" fallacy?
Correlation does in fact imply causation. Just not necessarily in the way that seems obvious.

If X is persistently correlated with Y, this implies there is a causal link between X and Y -- that it to say, it implies that there is an event Z (which might be equal to X or Y) that causes both X and Y.

It does not imply that X causes Y, or that Y causes X, which is the meaningful intuition between the idea that "correlation does not imply causation". A correlation between X and Y certainly does not imply that X causes Y. But it DOES imply that X and Y are caused by the same factor, either indirectly, or by the special case of X causing Y or Y causing X.

>>9655474
At that level of indirection, most non-autistic people would probably give the point up to their opponent, but not you, bud.

>>9655040
Fallacist's fallacy?

>>9655040
"no u"

>>9655048
>>9655087
>letting your beliefs and ideologies get in the way of rational thinking on scientific issues.
disgusting

>>9655486
That isn't true. You just turned it into an even worse fallacy. K and Y, ergo S.

"Correlation does not imply causation" is only true for the strict logical meaning of "implies." All scientific evidence is correlative. The primary role of the scientist is to design experiments so that causation is the only reasonable explanation for observed correlation.

>>9655040
Math doesn't feature any causation.
"n is a positive integer and n+n=n*n=n^n implies n=2" is not a causal relationship.

>>9655048
>All climatology is founded on correlations.
Wait...
Is this a /pol/ containment thread?

>>9657740
Where's your baseline?

>>9657771
>Where's your baseline?
Well, there's nobody here denying the existence of free will, time or gravity.
Nobody'd claiming we're alone in the universe and/or that it's all just a simulation.
There's no serious math involved, we're not talking about the college experience, and we're not shitting on women, psychology or engineers.
That alone narrows it down, but add in that OP understands logical fallacies far less than he thinks he does, and the first response brings up climatology.
I rest my case.

>>9655040
>What's the logical response to the "correlation does not imply causation" fallacy?
causes and effects are always correlated

>>9655053
I always say this. The whole principle of empiricism is founded on the idea that correlation suggests a causative link.
>>9655040
So: correlation does not imply causation, but it does suggest a causative link.

>>9655167
>There's a correlation between cold weather and increased spending on consumer goods in a lot of places.

Therefore, it can be assumed that the cold weather MAY be the cause of the increased spending on consumer goods, or that spending on consumer goods caused the weather to turn cold.

>Did the cold weather cause increased spending? Did the increased spending cause the cold weather?

No, of course not; but until this is proven, the implication exists.

Correlation does not confirm causation, but it does imply.

>>9657736
>all math is arithmetic

>>9655040
Showing a causal link.

>>9655040

Mention the fallacy fallacy.

>>9655040
correlation does not imply that one of the correlated things is causing the other but it does imply something is causing something
so correlation does imply causation

>>9655040
It is a fallacy.
It litterally does not imply causation and there are 1000s of cases where it was just random parreal events, not even a cause Z.
I swear math people are the biggest brainlets around and would ruin society if ever put in charge.
>>9655048
This is a good one, like the dilbart bait comic.
Note Adams here is disputing the economic models, not the climate models.

>>9655040
There is no fallacy. Correlation does not imply causation, however, causation and correlation can coexist.

>>9655040
Good ol' correlation and causation.

>>9655092
lmao you are beyond hope, as a statement within mathematics it is true that causation does not require correlation

>>9655040
You don't know what a fallacy is, you fucking moron.

>>9659163
>fucking moron.
Do you need to swear?

>>9655048
>teenager going through babby's first rebellion detected

>>9655040
"You are right."

>>9655053

Only if you can understand what
If and Only If means as a mathematical argument can you comprehend causation

>>9658845
Yeah, climate models being wrong is meme. Mainstream models have been pretty accurate. The ones that have failed are generally those made by "skeptics."

https://www.youtube.com/watch?v=tPSIvu0gQ90

As for economic models. It shouldn't be hard to measure how much damage sea level rise and temperature rise could cause.

>>9659969
>It shouldn't be hard to measure how much damage sea level rise and temperature rise could cause.
That’s the point though. Long term economic models have never been right, if for nothing else becuase they get the political response game theory wrong, but also just because the economic market adapts to future events in a way that is unknown and unknowable

>>9658034
>until this is proven, the implication exists
That's not how implication works you idiot.
There is no reason why you should ever expect a causal relationship between two correlated events. It's not uncommon for correlated events to have no direct relationship between each other.
Mostly though you should think a little harder before claiming the exact opposite of the fact they make a point of ramming down your throat in any statistics course. Like maybe you're wrong and the people teaching this topic for a living at universities aren't wrong. You idiot.

>>9660781
>Long term economic models have never been right, if for nothing else becuase they get the political response game theory wrong, but also just because the economic market adapts to future events in a way that is unknown and unknowable
400 million per city
http://www.miaminewtimes.com/news/miami-beachs-400-million-sea-level-rise-plan-is-unprecedented-but-not-everyone-is-sold-8398989

>>9660793
>It's not uncommon for correlated events to have no direct relationship between each other.

Right, because correlation does not confirm causation. You're just agreeing with me Anon.

>M-maybe you're wrong Anon-sama! B-baka!

Okay man. Whatever you say.

>>9655040
It doesn’t, though.

>>9660781
> Long term economic models have never been right

According to who? Some cartoonist? Estimating the cost and damages from sea level rise shouldn't be any harder than estimating the cost of hurricanes and tornadoes. The same goes for changes in temperature. We already know how temperature change can effect crops and farming.

Also, the act of adapting to these effects still cost money.

>>9655040
just because it can be a tired argument, that doesn't make it a fallacy you retard

>>9655040
Leftists tend to use this fallacy as a fallacy itself. The fallacy of overextending a fallacy.
In the 90's (and even to an extent today) there was a belief that that HIV did not cause AIDS. Which if you strictly take the definition of "correlation does not imply causation" then no evidence is ever good enough, all evidence is simply correlation. We hadn't proved the mechanism yet. Well obviously given the "correlation" between HIV and AIDS, we would assume that HIV does indeed cause AIDS.
Not many scientists take that phrase seriously, because all science is literally correlation implies causation. You just need to find the right correlation.
I only ever hear /reddit/ tier people say it anyway.

>>9660991
No, there is zero basis for your claim that correlation implies causation just because there's a possibility for causation.
There's always a possibility for causation, that's not the same as implying a causation.
By analogy there's a possibility when you drive that you will be killed in a car accident. That doesn't mean driving implies death.
Can I ask you honestly why you're arguing this when you know the exact opposite claim is what is drilled into you repeatedly in any formal university statistics course? Why do you believe they're all wrong and you're right exactly? What extraordinary evidence do you have that this horrible mistake has been made, and worse, has been aggressively hammered into every statistics student's head?

>>9663376
It's easy to say HIV causing AIDS was obvious when the mechanism has long since been established. That's drawing the target after the bullets have already been fired (Texas sharpshooter fallacy).
The reason HIV / AIDS denial was considered pseud shit at the time was because these people didn't merely point out we didn't know for sure, they went further and asserted conspiracies and/or /x/-tier alternative "explanations."

>>9663376
>Not many scientists take that phrase seriously, because all science is literally correlation implies causation. You just need to find the right correlation.

That's not how scientists I know use the word correlation. I don't know anything about AIDS controversy, but there's a difference between having gaps in a mechanism, and only having some data about "such and such patient had this symptom".

In Alzheimer's, there's a whole bunch of hypotheses about causes because the mechanism isn't clear yet. New drugs are soon to come along, but treatment hasn't been successful because there isn't a clear idea of what to target. If you're targeting symptoms and not causes you're not going to cure it.

You can sort causes into numbers too, but if you have tested theory backing your decisions then you're not relying on just correlations anymore. That's what science is about.

temporal correlation implies empirical causation. if event 1 is consistently followed by event 2 at some consistent window, induction suggests that will continue to happen in the future, barring any contradictory information.

>>9663855
So correlation implies correlation.
Congratulations.

>>9663855
>that will continue to happen in the future
That's not causation, that's continued correlation.
Go look at this picture until you understand the difference:
>>9659066