1MD - Statistics - Correlation doesn't imply causation

It doesn't take more than a minute to debunk a bad claim.

[C...]

🎶 I don't wanna close my eyes
I don't wanna fall asleep cause I'll miss babe
And I don't wanna miss a thing 🎶

[felixstuber]

Normally when I say "causation does not mean causation" it is short for: "Just because A correlates with B doesn't mean A causes B."

On the other hand, I usually don't simply reply with "causation doesn't mean causation" but try to deliver some points why in this case no A=>B causation exists.

[KnighteMinistriez]

You made some good points in this video. It forced me to think a little bit harder, stop doing that. How do you accomplish so much in a minute. It's really hard for me to make a video under ten minutes in length. You're really good at this. You're awesome. Merry Xmas.

[sebidotorg]

“I used to think correlation implied causation, but then I took a statistics class.”
“So the class helped?”
“Maybe.”

[ronaldQdobbs]

Correlation doesn't mean causation because shadows that you don't personally cast aren't real shadows! - Sleepy Warrior, 2019

[farmersboy]

Hmm, who could this be aimed at, I wonder?

[adshwartz]

I think this phrase is meant to make us think deeper and not find the answer with the most obvious thing, because the obvious things are happening in the background in most situations.

We humans have built ourselfes a very designed environment to full fill our needs, that created a surroundings full with repetitive patterns that is making us subconsciously biasd towards comfortable patterns

k bye


That cracked me up! Yes, I'm a simple guy that way...

[JohnnyDrivebye]

Caught in the cønfusøn øf causation øf pøst høc ergø prømpter høc. Thats the proper way to say it. Cheers!

[GaryGraham]

You caused me to watch this video by a bell therefore when the bell rings I must watch your video! simples👍

[fecu]

Correlation can just be an artifact of the size of the dataset. If you select by correlation then you might find the price index of pigfeed in Idaho corrates with a subset of Graham's Number. Given that the total available set of datasets is effectively infinite then there will be cases where increasing the sample size to the maximum will perfectly preserve the correlation. We will find cases of perfect statistical significance and no meaningful connection between them. We can find these things if we look for them, our own agency being a causal link of sorts in that case, but we can also happen upon them. Microstates are equally likely after all.

So while correlation does _imply_ causation (the majority of correlations will be causal), it does not necesitate it.

EDIT: I'm now doubting if it's true to say that the majority of correlating datasets are causal. Graham's Number likes _does_ contain the prices of pigfeed in Idaho. It's probably only safe to say we can factor in statistical significance but never sure it's not pure coincidence. That's science's approach for good reason.

[arcanics]

The phrase IS wrong, it should be "Correlation doesn't MEAN causation." Because correlation DOES imply causation, but because something is implied does not make it true.

[tubbysnowman]

Interesting. I've never heard "correlation does not imply causation" I've only ever heard it as "correlation does not equal causation.

Which is technically more true, but you still have the problem of there needs to be some linked causation somewhere.

[captphobos]

Awesome! People are obsessed with knowing why? This seems to be the way the human brain works... Why is always the wrong question, 'how' makes a lot more sense. Brown and Newton are heroes, they spotted the correlation, but never double guessed the mechanism... Ironically Einstien suggested the how on both Brownian Motion and Gravity.

[TheSeverian]

There most certainly doesn't have to be any causal relation to a correlation. You left out the entire category of chance and coincidence.

[bobby_greene]

But a lack of pirates still causes atmospheric CO2 levels to rise, right?

[Zpostate]

Jeez, take away all my cool kid debunking phrases, why don't you?

[sonaruo]

The problem is that today people learn this sentences and they use them as spells thinking that if they use them somehow magically will win the argument. No it does only show their ignorance on the topic talked nothing more.


While in general the quote can be true. IT begs the question how often it actual happens. How many times people put 2 things together that are no connected?


If you are uneducated you have partial info you may be more ease to incline in drawing wrong conclusion like flat earther the word horizon seems similar for horizontal earth is flat lol.

[miff]

One Minute Debunk videos have a strong correlation to @Janey being first to comment, therefore, @Janey is the cause of these videos. No credit for OMD😋