knabe
Well-known member
if something is correlated, is it causal?
correlated vs. causal
- Thread starter knabe
- Start date
Help Support Steer Planet:
JbarL
Well-known member
i would say as in its most diluted of relationships......yes....
ELBEE
Well-known member
Nope, your too late, it's already done. The Fox correlated the chicken causal he was hungry. You might'a saved the chicken if'n you'da uncausaled the fox whil'ist he was still correlatin'.
AAOK
Well-known member
knabe, I found an interesting perspective ???
Pet Peeve: Correlation is not Causality
“Good grief! What is he up to now?” - Fine, I deserve that. Now, give me back control of my frontal lobe. Confusing correlation and causality (unlike alliteration) is indeed one of my pet peeves. Pet peeves: I do hope that term translates well. You can’t buy peeves at the pet store, and I know for a fact that all the wild ones can’t be house-broken. Pet peeves, my dear friends in Malta, are “those concepts, utterance, and related manifestations of life that bug the crap out of us.” As Skatman Crothers once said, “Let me elucidate here…”
Correlation means that two things (let’s call them A and B) are somehow related. Causality, of course, means that A and B have some causal relationship. For fun and giggles, let’s say that A cause B. All is good, yes? Well, the problem that I see is that too many people are confusing correlation with causality. More elucidation follows:
Let’s say that I hit you - in the snout - and that you get a bloody nose. Most courts would see that as a clear case of correlation and causality. There is a high correlation between a punch in the nose and the issuance of blood from same. A jury would say: “He caused your damage.” Actually, they would probably let me off, seeing as how I can blame caffeine, growing up in The South, or The Dukes of Hazzard TV show for my poor social graces. I would create my own obscure causal chain and basically absolve myself of any blame. But, that’s beside the point. The real nuggets is that there is a clear, direct and irrefutable link between A and B. A happens first, and B always happens next, as a direct result of A. That is causality. If the results are not direct, or if there are intervening factors, we might not be looking at good ol’ causality. However, I can attest that a bloody punch in the beak is a statistically significant case of causality, and is punishable as such in most, if not all civilized jurisdictions.
Correlation is less directed. Correlation just says “these two things have a relationship.” Ideally, the relationship is clear: As A goes up, so does B. As A happens, so does B. Correlation might indicate causality, or it might not. This is where the problem comes in. Follow this one…
Being the wildly naive student of life that I am, I believe there is a strong correlation between in-ground swimming pools and household income. I pretty much believe that if you live in a trailer park, you are unlikely to have an in-ground pool. It’s not impossible, but I don’t see it as an everyday occurrence. On the other hand, I bet most of you rich folks have a pool or two in your backyard, and probably one in the master closet. I see a correlation - a really simple correlation. Now, the less rigorous among us might see a causal relationship (and yes, there are two we could talk about here, but I’m only interested in one). “Hey! In-ground pools cause high-income! Let’s set up a government funding program to install in-ground pools in low-income neighborhoods! Let’s raise the average income level.”
Well, I know you wouldn’t do this, but you’re special, and most of you already have in-ground pools. But, think of the average metrics that you have built around your business processes. Think of correlations that you might see there. Do you think someone, somewhere in the process food chain might be tempted to confuse correlation with causality? You don’t need an in-ground pool to fall on your head to answer this one. Too many times, we attempt to transform a predictive relationship into a causal relationship. My old stats prof - rest his now-departed soul - used to drill us with the three goals for measurements: describe, predict and control. Correlation can be predictive, yes. If you have a lot of expensive items to sell (say caviar), look for the in-ground pools and start knocking on doors. But correlation does not beget control. You can’t go and remove someone’s in-ground pool and watch their income level drop. It doesn’t work that way, even though I think it would be a hoot if it did. If you want to prove causality, you have to do some experimentation. No one likes to do experimentation - it’s too costly, too time-consuming, and besides, everyone knows it causes warts.
One last fish-in-the-barrel: “What about high-income causing in-ground pools? Isn’t that a causal relationship?” Well, if I raise your income to $1 million per year, will that cause an in-ground pool to appear in your backyard? Not directly - I hope - but if it does, call me at home. Most likely, it will cause you to get an inflated ego, dump your current house, move across town, put your kids in private school, go to a lot of fancy parties, swim in other peoples’ pools, feel left out, and order a new in-ground pool. That might be causality, but it doesn’t feel the same as a punch in the nose, does it? Besides, if we go down that path, then I have to conclude that birth causes death, marriage causes babies, and Ferrari ownership causes mustaches.
On that last point, if any university out there is willing to fund a wart-free experiment, I am happy to drive a Ferrari California for the rest of the year and report back to you… provided I’m not out installing in-ground pools as part of some poorly-conceived social program, or too engrossed in the Dukes, or too busy shaving under my nose.
P.S. For the record, I don’t have an in-ground pool. But, if you ever want to go wading, come on over and bring your own towel. Conclude what you want about my income, but don’t talk to me about causality unless you know what’s what.
Pet Peeve: Correlation is not Causality
“Good grief! What is he up to now?” - Fine, I deserve that. Now, give me back control of my frontal lobe. Confusing correlation and causality (unlike alliteration) is indeed one of my pet peeves. Pet peeves: I do hope that term translates well. You can’t buy peeves at the pet store, and I know for a fact that all the wild ones can’t be house-broken. Pet peeves, my dear friends in Malta, are “those concepts, utterance, and related manifestations of life that bug the crap out of us.” As Skatman Crothers once said, “Let me elucidate here…”
Correlation means that two things (let’s call them A and B) are somehow related. Causality, of course, means that A and B have some causal relationship. For fun and giggles, let’s say that A cause B. All is good, yes? Well, the problem that I see is that too many people are confusing correlation with causality. More elucidation follows:
Let’s say that I hit you - in the snout - and that you get a bloody nose. Most courts would see that as a clear case of correlation and causality. There is a high correlation between a punch in the nose and the issuance of blood from same. A jury would say: “He caused your damage.” Actually, they would probably let me off, seeing as how I can blame caffeine, growing up in The South, or The Dukes of Hazzard TV show for my poor social graces. I would create my own obscure causal chain and basically absolve myself of any blame. But, that’s beside the point. The real nuggets is that there is a clear, direct and irrefutable link between A and B. A happens first, and B always happens next, as a direct result of A. That is causality. If the results are not direct, or if there are intervening factors, we might not be looking at good ol’ causality. However, I can attest that a bloody punch in the beak is a statistically significant case of causality, and is punishable as such in most, if not all civilized jurisdictions.
Correlation is less directed. Correlation just says “these two things have a relationship.” Ideally, the relationship is clear: As A goes up, so does B. As A happens, so does B. Correlation might indicate causality, or it might not. This is where the problem comes in. Follow this one…
Being the wildly naive student of life that I am, I believe there is a strong correlation between in-ground swimming pools and household income. I pretty much believe that if you live in a trailer park, you are unlikely to have an in-ground pool. It’s not impossible, but I don’t see it as an everyday occurrence. On the other hand, I bet most of you rich folks have a pool or two in your backyard, and probably one in the master closet. I see a correlation - a really simple correlation. Now, the less rigorous among us might see a causal relationship (and yes, there are two we could talk about here, but I’m only interested in one). “Hey! In-ground pools cause high-income! Let’s set up a government funding program to install in-ground pools in low-income neighborhoods! Let’s raise the average income level.”
Well, I know you wouldn’t do this, but you’re special, and most of you already have in-ground pools. But, think of the average metrics that you have built around your business processes. Think of correlations that you might see there. Do you think someone, somewhere in the process food chain might be tempted to confuse correlation with causality? You don’t need an in-ground pool to fall on your head to answer this one. Too many times, we attempt to transform a predictive relationship into a causal relationship. My old stats prof - rest his now-departed soul - used to drill us with the three goals for measurements: describe, predict and control. Correlation can be predictive, yes. If you have a lot of expensive items to sell (say caviar), look for the in-ground pools and start knocking on doors. But correlation does not beget control. You can’t go and remove someone’s in-ground pool and watch their income level drop. It doesn’t work that way, even though I think it would be a hoot if it did. If you want to prove causality, you have to do some experimentation. No one likes to do experimentation - it’s too costly, too time-consuming, and besides, everyone knows it causes warts.
One last fish-in-the-barrel: “What about high-income causing in-ground pools? Isn’t that a causal relationship?” Well, if I raise your income to $1 million per year, will that cause an in-ground pool to appear in your backyard? Not directly - I hope - but if it does, call me at home. Most likely, it will cause you to get an inflated ego, dump your current house, move across town, put your kids in private school, go to a lot of fancy parties, swim in other peoples’ pools, feel left out, and order a new in-ground pool. That might be causality, but it doesn’t feel the same as a punch in the nose, does it? Besides, if we go down that path, then I have to conclude that birth causes death, marriage causes babies, and Ferrari ownership causes mustaches.
On that last point, if any university out there is willing to fund a wart-free experiment, I am happy to drive a Ferrari California for the rest of the year and report back to you… provided I’m not out installing in-ground pools as part of some poorly-conceived social program, or too engrossed in the Dukes, or too busy shaving under my nose.
P.S. For the record, I don’t have an in-ground pool. But, if you ever want to go wading, come on over and bring your own towel. Conclude what you want about my income, but don’t talk to me about causality unless you know what’s what.
Show Dad
Well-known member
Oh crap! Logic on a Saturday morning.
Knabe has used a correlation (mutual relationship) of the discussion function of Steer Planet and it's members to show the causality (the relationship between cause and effect) of the member's replies. (argue)
<alien>
Knabe has used a correlation (mutual relationship) of the discussion function of Steer Planet and it's members to show the causality (the relationship between cause and effect) of the member's replies. (argue)
<alien>
ELBEE
Well-known member
A causal B correlated C, but A and B have allready transpired so C cannot be changed. Therefore in the future when C has a negative impact on me (poor poor pitiful me), I must intervene prior to the completion of cause A and B. Just cause'!
Dah! What are you guy's thinkin'? The governmet can help us with this!
Dah! What are you guy's thinkin'? The governmet can help us with this!
