If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(Slate)   Next time some pseudo-intellectual internet blowhard tries to take away your carefully thought-out arguments with that "correlation does not imply causation" yarn, just send them here because YOU WIN   (slate.com) divider line 226
    More: Cool  
•       •       •

27660 clicks; posted to Main » on 02 Oct 2012 at 5:31 PM (1 year ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



226 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

Archived thread

First | « | 1 | 2 | 3 | 4 | 5 | » | Last | Show all
 
2012-10-03 02:41:22 AM
Correlation, Causation, and Communism all start with 'C'.
 
2012-10-03 02:42:03 AM

kieran57: "By convention, we call an effect "significant" if the chances of its deriving from a twist of fate-as opposed to some more genuine relationship-are less than 5 percent."

P VALUES DO NOT WORK THAT WAY.


Or...

1 in 20 significant events happen by pure chance.

growlersoftware.com
 
2012-10-03 02:58:03 AM
In psychology (which for those who've never studied it is actually very heavily statistics based) this crap gets incredibly messy. (A happened to B which may have been a factor in C when we use this statistical model and these criteria for blah blah blah)

My brain has been forever scarred reading research papers which all seem to end with "It may be that A was the catalyst for B, however E, F and G which occurred and appear to have been influenced by B (etc)"

Or co-morbidity. Easy to explain at a really high level, hard to scientifically explain.
 
2012-10-03 03:00:56 AM

rcf1105: As I tell my students, correlation may not show causation, but it does show association, which can be equally important. And besides, use your common sense. Studying is highly correlated with getting good grades, and I'm pretty sure that getting good grades doesn't cause one to go out and study more.


Well, except if you don't excel in early in school, when they're teaching the number of syllables in words, spelling, and multiplication tables, which you hate so you dislike this "school stuff," so you don't try hard. Which then carries over to the algebra and calculus that you practically find to be second nature, but you don't try at that either...
 
2012-10-03 03:02:25 AM

Bhruic: ThrobblefootSpectre: I've noticed the vast majority of people who use the "correlation does not imply causation" soundbite use it in completely inappropriate context. They use it as a complete non-sequiter in a context where no causation was claimed, apparently in the hope someone thinks it means something.

Not really true. Most people have come to use it because of the prevalence of jumping to conclusions - especially in the media. "Scientistists find that some people with X also have Y" becomes "X causes Y!". The whole correlation != causation came as a push-back to that. Or so it seems to me.


Yep. Especially when they don't consider Y causes X.

Like I said, bad drivers drink so they feel more at ease at the wheel.
 
2012-10-03 03:05:01 AM

BolloxReader: Dude, have you seen all the beer ads during football games? There's your link. A ton of money sunk into consumer psychology to get you to associate "game time" with "drinking time." Or most any sport for that matter.

There IS an underlying link, and watching lots of football will induce more people to drink during football season. It may not happen for YOU or for any one person in particular because you can't predict specific results for a given person. But you CAN say that out of 1 million viewers, x% will drink more beer during the football season because of the prompts in the advertising. It's not the game of football that does it, it's the social environment that Madison Avenue has spent a lot of money cultivating around the game.


Or maybe advertisers know the correlation between watching football and drinking beer.

It's called "target marketing."
 
2012-10-03 03:12:05 AM

doglover: Carefully framed logical arguments can easily arrive at conclusions that simply don't mirror reality thanks to faulty assumptions at the start.


If they have faulty assumptions, they aren't carefully framed, moron.
 
2012-10-03 03:15:43 AM

LiberalEastCoastElitist: I've pretty much never seen anyone who uses this phrase or applies logical fallacy names to arguments to have an intelligent, original thought. I can very much picture them in their freshman writing or statistics class mentally rubbing their hands together in glee at all the internet arguments they will now vanquish.


This is what stupid people say to avoid the realization they're stupid.
 
2012-10-03 03:23:07 AM

spamdog: Animatronik: and is still important to 21st century progressives, who cherry-pick correlations and declare them to be facts with no supporting logic to back them up

I suppose the irony of that flew right over your head.

You know what pisses me off about arguing online? People who cry ad hominem. I see this frequently used as an "I win" clause - you insult me, that means you have no argument and I'm right.

What is usually happening in this situation is that the person is a certified moron, so instead of wasting their breath trying to convince the person of something, they call them a cocksucker instead.


Many computers ago, I had a fark thread about the "monty hall problem" bookmarked. It started off with a discussion. Those that understood the undebatable true mathematical answer tried to explain it to those that didn't understand. After some time, the people with the right answer started rightfully calling those with the wrong answer "stupid," at which point those that didn't understand roughly said "LOL, you're name calling, which means you have no argument left." Implying that they were in fact right.

They were right, but not how they thought. There was no argument left that could show them how they were wrong (that they could understand), but that doesn't mean they were right. It just means they were stupid.
 
2012-10-03 03:26:56 AM

Fano: trappedspirit: Oh FFS, everyone with half a brain has the list of logical fallacies in their clipboard and just pastes "moving the goal posts" until they hit "back-pedaling" and slip in some "tautology" dripping from your "strawman". And when they start sounding really cool they throw in some "appeal to authority" with a side order of "anecdotal evidence" until we get to their "ad hominem" money shot. That's not cliched. That's how real brainiac powerhouses roll! Huzza!

Wow, you really went down the slippery slope there.


+2
 
2012-10-03 03:34:14 AM
ITT: Internet geniuses.

/douchebags
 
2012-10-03 03:38:03 AM

Firethorn: It's logic along the lines of how in science today you get 'theories' instead of 'laws'. Einstein's theory of relativity is more advanced than the 'law' of gravity, but it's still 'only' a theory. When they find something that's 'more true', it'll be replaced/modified.


That's not how scientific theories and laws work. Theories do not "graduate" into being laws. Laws merely describe physical relationships that always behave the same way under certain conditions (ex. Fg=(GmM)/(r^2) ) without explaining how it works. Theories are rigorous explanations of why those relationships work the way they do.
 
2012-10-03 04:02:44 AM

This About That: Correlation does imply causation. Correlation does not confirm causation.


The example given in "How to Lie with Statistics" (a book you need to read) is the close correlation between the wholesale price of rum in Havana and the salaries of presbyterian ministers in New England.

Does that imply causation to you?
 
2012-10-03 06:34:23 AM
I'll just leave this here...

Freakonomics

It's strange that nobody has brought this one up yet.
 
2012-10-03 06:46:10 AM

KrispyKritter: major bingo there. people parroting this and that. life is full of them. those that really have something going on don't have time to fark around and piss away time on teh internet.


Start listening to people, instead of what they are saying, try figuring out where they heard that. 99% of everything people say is parroted from somewhere. Downside, once you see it, other peoples conversations will be 99% more annoying.
 
2012-10-03 07:51:51 AM

Coolfusis: Oh look, it's another one of the "that thing doesn't exactly mean that!" articles. I'll step aside and let the pedants masturbate furiously over this one.


I'm not a pedant, so can I just leisurely masturbate to this one?
 
2012-10-03 08:36:40 AM

KrispyKritter: major bingo there. people parroting this and that. life is full of them. those that really have something going on don't have time to fark around and piss away time on teh internet.


So, what have you done that is so great for humanity?
 
2012-10-03 09:34:44 AM

kxs401: I guess I'm just ignorant, because I certainly realize that correlation doesn't PROVE causation, but why doesn't it imply it? If A and B are correlated, possible explanations are that A causes B or B causes A. When we notice that smoking is correlated with lung cancer, why wouldn't we go looking to find causality?


Because "possible" and "implies" are not the same thing.


Smokers stand outside to smoke. Standing outside correlates to lung cancer. Standing outside must cause lung cancer.
 
2012-10-03 09:36:18 AM
perfect article aimed at about half of all Farkers who post regularly.

Should have thrown in some of the other favorites as well, like:

" strawman argument"

"false equivalency"

"ad hominem attack"

"poe's law"
 
2012-10-03 09:45:47 AM

orbister: Does that imply causation to you?


If one were considering the correlation without the axe to be ground, it certainly would. The next step would be to investigate the possible cause-and-effect relationship. The fact that you appear to presume, without investigation, that any possible such relationship is absurd, along with the fact that you are quoting a book called "How to Lie With Statistics" leads me to believe you have taken no action to either prove or disprove the premise you are presenting as "proof" of... something.
 
2012-10-03 09:57:54 AM

Bullseyed: kxs401: I guess I'm just ignorant, because I certainly realize that correlation doesn't PROVE causation, but why doesn't it imply it? If A and B are correlated, possible explanations are that A causes B or B causes A. When we notice that smoking is correlated with lung cancer, why wouldn't we go looking to find causality?

Because "possible" and "implies" are not the same thing.


Smokers stand outside to smoke. Standing outside correlates to lung cancer. Standing outside must cause lung cancer.


There are cars belching deadly exhaust into the atmosphere near smokers standing outside. Therefore smokers cause global warming.
 
2012-10-03 10:13:46 AM
Wow, this thread is almost as much of a clusterf*ck as watching people discuss the phrase "exception that proves the rule."

ProTip for any of you operating under the assumption that "imply" in the phrase is synonymous with "suggest": You're wrong. Shut your pie-hole, learn from the few posters in here that have pointed out (more than once no less) what the phrase actually means, and get on with your lives.

Everybody starts out ignorant; the true measure of character is seeing who chooses to remain that way.
 
2012-10-03 02:22:17 PM

moran: I think your last sentence is a bit too strong. The theory that eating cyanide causes death was well-established by experiment (or could have been, anyway) long before it was known how eating cyanide causes death.


Perhaps you're right. Still, one can at least do a study where you 'assign' people to eat cyanide and track the resulting deaths. Well, you can if you're heartless, and there have been plenty of heartless people through history willing to kill by poisoning. It's still a study.that checks for causation by deliberately administering cyanide, rather than waiting for some segment of the population to decide to ingest the (presumed unknown) substance of their own volition.

ChuDogg: I think it's more fallicous to automatically assume that "more texting = more accidents". I can think of a few reasons right off the top of my head that more texting can correlate with less accidents (less people driving for one). If I saw a study or something that correlated texting with decreased accidents I would not dismiss it out of hand saying "correlation does not equal causation"


The latest texting studies I've seen are starting to say that the people who are texting wile driving are the people who drive distracted anyways, making is so that 'driving while texting' accidents are more displacing 'driving while drinking/eating/applying makeup/combing hair/talking on cellphone' accidents than increasing the rate.
 
2012-10-03 05:52:39 PM

likesass: Correlation does not always equal causation, but, without correlation, you can eliminate causation.


Correlation is a measure of the *linear* dependence between two variables. If a causal relationship produces data that are not linearly dependent, then in fact you can have causation without seeing a correlation absolutely large enough to raise an eyebrow.

For instance, placing a probe further out in the Solar system causes it to receive less energy from the sun. But the relationship between distance and energy will be as the inverse-square of the distance. Thus if we summarize distance/energy data from the probe only via correlation, a linear construct, we might not suspect a causal relationship between the two values. In fact, the correlation you will see with this kind of relationship will depend very much on the distances at which you measured.

Nice to finally have a reason to post on here... I think.
 
2012-10-03 06:06:10 PM

RRod: likesass: Correlation does not always equal causation, but, without correlation, you can eliminate causation.

Correlation is a measure of the *linear* dependence between two variables.


Uhhh. No. Pearson correlation methods are sensitive only to linear correlation, but that is only one way of measuring it among many others.
 
2012-10-04 10:08:09 AM
Well when we see correlation between A and B we have several possibilities.

1) A caused B (or helped cause)
2) B caused A (ditto)
3) Something else caused both A and B (ditto)
4) If both A and B are both increasing (or decreasing) over time, it could that they are doing so by coincidence (and a fairly likely one at that as anything that trends in one direction will at least crudely correlate with something that also trends in one direction)
5) Sampling error.

Of course you can use other evidence to eliminate these. For smoking and lung cancer, way too many studies exist for 5 to be viable, 2 and 3 are a bit absurd, and 4 could theoretically be cited for smoking decrease correlating with lung cancer decrease. But 4 could not explain studies which take a group of smokers and a group of people similar in other regards expect they don't smoke, follow them and see that the smokers really are far more likely to get cancer.
 
Displayed 26 of 226 comments

First | « | 1 | 2 | 3 | 4 | 5 | » | Last | Show all

View Voting Results: Smartest and Funniest


This thread is archived, and closed to new comments.

Continue Farking
Submit a Link »






Report