## When is it cause, not correlation?

Prepare for the LSAT or discuss it with others in this forum.
glucose101

Posts: 409
Joined: Sun Mar 29, 2009 12:23 am

### When is it cause, not correlation?

"Correlation does not imply causation." I can't tell you how many times I've heard that. On the LSAT, it seems like causation is never the "correct" answer, and that correlation is always mistaken for causation. How would causation be "correct" on the LSAT?

AlabamaIceman

Posts: 94
Joined: Mon Feb 14, 2011 3:32 pm

### Re: When is it cause, not correlation?

I'm not sure what you're asking. Most of the time when that phrase appears, it's in reference to a "weaken" question, and you are pointing out that the faulty argument in the prompt shows correlation but not causation. If it shows causation, then the argument being put forth is most likely correct.

Conversely, if the question asks you to prove something or strengthen an argument, you would be looking for causation rather than correlation.

NYC Law

Posts: 1563
Joined: Thu May 26, 2011 3:33 pm

### Re: When is it cause, not correlation?

There's almost never causation, researchers have a very difficult time finding actual direct causation. There are some research and statistical methods that can get you close to finding a causation, but it still usually can't prove it.

Just assume there's never a causation, there can always be an extraneous variable.

glucose101

Posts: 409
Joined: Sun Mar 29, 2009 12:23 am

### Re: When is it cause, not correlation?

NYC Law wrote:There's almost never causation, researchers have a very difficult time finding actual direct causation. There are some research and statistical methods that can get you close to finding a causation, but it still usually can't prove it.

Will the LSAT ever have such scenarios? I realize the LSAT will probably only use correlation/causation on flaw/weaken questions, but I just want to make sure there aren't instances that the LSAT would defy this.

suspicious android

Posts: 919
Joined: Tue Feb 09, 2010 4:54 pm

### Re: When is it cause, not correlation?

Are you asking when a causal conclusion could be validly drawn on the basis of a mere correlation? Never. A causal conclusion based on a mere correlation may be true, but never validly drawn. So I'm not sure what you mean by "correct" here, but there are definitely some instances where a causal argument based on a correlation has not hinged on that correlation issue, but on some other flaw instead.

TMC116

Posts: 284
Joined: Mon Jun 06, 2011 6:08 pm

### Re: When is it cause, not correlation?

suspicious android wrote:Are you asking when a causal conclusion could be validly drawn on the basis of a mere correlation? Never. A causal conclusion based on a mere correlation may be true, but never validly drawn. So I'm not sure what you mean by "correct" here, but there are definitely some instances where a causal argument based on a correlation has not hinged on that correlation issue, but on some other flaw instead.

Exactly.

If you see a causal argument, assume its wrong/flawed. If it is the basis for a conclusion, then that's where the flaw is (so act accordingly when asked to weaken, strengthen, or identify the flaw)

EarlCat

Posts: 606
Joined: Mon Mar 12, 2007 4:04 pm

### Re: When is it cause, not correlation?

glucose101 wrote:"Correlation does not imply causation." I can't tell you how many times I've heard that. On the LSAT, it seems like causation is never the "correct" answer, and that correlation is always mistaken for causation. How would causation be "correct" on the LSAT?

You can only conclude causation if the premise tells you there's causation.

glucose101

Posts: 409
Joined: Sun Mar 29, 2009 12:23 am

### Re: When is it cause, not correlation?

EarlCat wrote:
glucose101 wrote:"Correlation does not imply causation." I can't tell you how many times I've heard that. On the LSAT, it seems like causation is never the "correct" answer, and that correlation is always mistaken for causation. How would causation be "correct" on the LSAT?

You can only conclude causation if the premise tells you there's causation.

Ok! Thanks! This is what I assumed, but just wanted to make sure.

MCockerill08

Posts: 6
Joined: Tue Jun 07, 2011 9:44 am

### Re: When is it cause, not correlation?

Causation can rarely if ever be "proven" on an LSAT LR question (though in sufficient assumption questions, the correct answer could "assume" the casual link that would justify an argument's conclusion can be assumed), but arguments alleging causation can be strengthened by eliminating alternative causes or buttressing the credibility of the link (e.g.: if expanding and diversifying the sample size of our original study yields the same correlation, it is more likely that some casual link exists) .

Audio Technica Guy

Posts: 317
Joined: Tue Jun 29, 2010 5:21 pm

### Re: When is it cause, not correlation?

yeah, there has never been a single question on the LSAT were causation was successfully proven. LSAC would likely shy away from this, because there isn't a logically agreed upon definition of the word causation. It would be too ripe for counterargument.

So, on the LSAT, the only possible way you can have causation is if a premise straight up tells you it is, without a doubt, causation. Off the top of my head, I can't actually recall any question that did that either.

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am

### Re: When is it cause, not correlation?

The above isn't necessarily true. Since we are told to accept premises at face value, if we are given causal premises, like 'smoking always causes cancer' and 'cancer always causes death', we could be asked what 'must be true' given those facts, and in the example above, it would be absolutely true that smoking always causes death.

And naturally, I'm smoking a cigarette as I write this..

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am

### Re: When is it cause, not correlation?

Audio Technica Guy wrote:yeah, there has never been a single question on the LSAT were causation was successfully proven. LSAC would likely shy away from this, because there isn't a logically agreed upon definition of the word causation. It would be too ripe for counterargument.

So, on the LSAT, the only possible way you can have causation is if a premise straight up tells you it is, without a doubt, causation. Off the top of my head, I can't actually recall any question that did that either.
whoops.. Didn't read this last bit.

But from my recollection there have been a couple (though rare) valid causal arguments.

From what I remember the answers are usually phrased something like, 'doing blah increases risk of something blah'

suspicious android

Posts: 919
Joined: Tue Feb 09, 2010 4:54 pm

### Re: When is it cause, not correlation?

skip james wrote:But from my recollection there have been a couple (though rare) valid causal arguments.

From what I remember the answers are usually phrased something like, 'doing blah increases risk of something blah'

So what are you saying? That a causal statement can be validly concluded? That's already the consensus of the thread, it seems. Just need a premise that indicates causality. Absent that, it's impossible.

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am

### Re: When is it cause, not correlation?

suspicious android wrote:
skip james wrote:But from my recollection there have been a couple (though rare) valid causal arguments.

From what I remember the answers are usually phrased something like, 'doing blah increases risk of something blah'

So what are you saying? That a causal statement can be validly concluded? That's already the consensus of the thread, it seems. Just need a premise that indicates causality. Absent that, it's impossible.

I take it you don't buy into the second analogy of Kant's Critique of Pure Reason then.

suspicious android

Posts: 919
Joined: Tue Feb 09, 2010 4:54 pm

### Re: When is it cause, not correlation?

skip james wrote:I take it you don't buy into the second analogy of Kant's Critique of Pure Reason then.

I'll admit I've forgotten almost everything about Kant I ever knew, but wasn't that position you're referring to the idea that we must assume causality as we intuitively understand it to be real? That's not to say that it's really real, just that it's how we're stuck interpreting the world. Regardless, I always found Hume to be on much firmer ground on this sort of thing. That might be because I'm not smart enough to really understand Kant, but well.. I'm okay with that.

westinghouse60

Posts: 392
Joined: Mon Aug 02, 2010 2:27 am

### Re: When is it cause, not correlation?

suspicious android wrote:
skip james wrote:I take it you don't buy into the second analogy of Kant's Critique of Pure Reason then.

I'll admit I've forgotten almost everything about Kant I ever knew, but wasn't that position you're referring to the idea that we must assume causality as we intuitively understand it to be real? That's not to say that it's really real, just that it's how we're stuck interpreting the world. Regardless, I always found Hume to be on much firmer ground on this sort of thing. That might be because I'm not smart enough to really understand Kant, but well.. I'm okay with that.

Didn't Hume say we can never know that one event causes another, only that one event seems to always follow the other? I'm probably oversimplifying that, or maybe I'm way off entirely.

suspicious android

Posts: 919
Joined: Tue Feb 09, 2010 4:54 pm

### Re: When is it cause, not correlation?

westinghouse60 wrote:Didn't Hume say we can never know that one event causes another, only that one event seems to always follow the other? I'm probably oversimplifying that, or maybe I'm way off entirely.

Yeah, that's basically right. It went like this:

Hume writes a bunch of stuff, says that induction and causation and just reason in general are all bullshit, we can only know what we can see.

Kant says "Man, you just blew my mind." He then resolves to prove Hume completely wrong because he is the most boring man in the history of the world and doesn't like Hume ruining all the nice ideas everyone has about God, the soul, rationality, etc.

Ten years later, he puts out a book that says it turns out all that stuff is real after all, and we can sorta prove it if we just think about it really, really hard. It doesn't make much sense, but he sounds so smart that everyone just kinda goes along with it for awhile, because they don't want to look dumb and it's what they wanted to believe anyway.

That was my take anyway, I didn't do too great in that class.

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am

### Re: When is it cause, not correlation?

westinghouse60 wrote:Didn't Hume say we can never know that one event causes another, only that one event seems to always follow the other? I'm probably oversimplifying that, or maybe I'm way off entirely.

Yup. Hume's the guy who made correlation => causation a flaw. Major lsat props to him.

suspicious android wrote:Yeah, that's basically right. It went like this:

Hume writes a bunch of stuff, says that induction and causation and just reason in general are all bullshit, we can only know what we can see.

Kant says "Man, you just blew my mind." He then resolves to prove Hume completely wrong because he is the most boring man in the history of the world and doesn't like Hume ruining all the nice ideas everyone has about God, the soul, rationality, etc.

I don't think this is fair. I'm not sure if Kant is right about the whole ordeal, but the dude did put a decade into thinking about it, and I think to be charitable, we should at least think about it for more than 10 minutes before making a dismissive forum post about it.

I struggled with Kant a lot myself, but only because I thought there might be something there that I wasn't seeing. It's easy to just shrug it off as philosophic nonsense, but it's quite difficult to encounter some complex, very well-thought out arguments and to really try and understand where it's all coming from.

I'm not sure if I buy into Kant myself, but I'm definitely not as dismissive as you about it, or at the very least, I try not to be..

Who knows.. I could be wasting my time.. I've definitely done that before.

jlee282

Posts: 17
Joined: Tue Mar 16, 2010 11:28 am

### Re: When is it cause, not correlation?

If I remember correctly, there were few problems on the LR that asked to look for the necessary assumption of a causal relationship. There is a great chapter on PowerScore Bible about causal relationships. Perhaps you can look that up.

suspicious android

Posts: 919
Joined: Tue Feb 09, 2010 4:54 pm

### Re: When is it cause, not correlation?

skip james wrote:I don't think this is fair. I'm not sure if Kant is right about the whole ordeal, but the dude did put a decade into thinking about it, and I think to be charitable, we should at least think about it for more than 10 minutes before making a dismissive forum post about it.

What makes you think I only thought about it for 10 minutes? I wasted three good months of my life trying to understand Kant before I dismissed him.

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am

### Re: When is it cause, not correlation?

Sorta embarrassing.. but I can top that. By a lot.

And three months really isn't too long, especially for Kant. Three months is tantamount to saying 'I took a Kant class once'. That's barely enough time to a grasp of what's going on in the Aesthetics of the Critique.

But I'm biased. I like Kant, regardless of whether or not he ultimately succeeds. He's got heart, plus he's probably a nice guy, having invented the categorical imperative and all.

westinghouse60

Posts: 392
Joined: Mon Aug 02, 2010 2:27 am

### Re: When is it cause, not correlation?

skip james wrote:Sorta embarrassing.. but I can top that. By a lot.

And three months really isn't too long, especially for Kant. Three months is tantamount to saying 'I took a Kant class once'. That's barely enough time to a grasp of what's going on in the Aesthetics of the Critique.

But I'm biased. I like Kant, regardless of whether or not he ultimately succeeds. He's got heart, plus he's probably a nice guy, having invented the categorical imperative and all.

Only Kant I've read is Groundwork of the Metaphysics of Morals. Although I'm about to start Critique of Pure Reason now that the LSAT is done and I have tons of free time lol.

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am

### Re: When is it cause, not correlation?

you're gonna need a lot of coffee.

suspicious android

Posts: 919
Joined: Tue Feb 09, 2010 4:54 pm

### Re: When is it cause, not correlation?

skip james wrote:But I'm biased. I like Kant, regardless of whether or not he ultimately succeeds. He's got heart, plus he's probably a nice guy, having invented the categorical imperative and all.

I read that he was a pretty nice guy, and apparently a beloved professor, which is pretty atypical for big deal philosophers. I'm sure there's lots to be gained from a thorough, years long study of his works, but . . . I haven't even seen season 3 of Breaking Bad yet, I gotta prioritize.

skip james

Posts: 262
Joined: Sat Sep 19, 2009 2:53 am