Mixing
If we have two non-zero-correlated random variables, then why do we say that “correlation does not imply...
$begingroup$
If we have two non-zero correlated random variables then they are dependent. Why then do we have the saying "Correlation does not imply Causation". A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link.
statistics logic causal-diagrams
edited Jan 20 at 18:59
nbro
2,41563174
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
$endgroup$
add a comment |
$begingroup$
If we have two non-zero correlated random variables then they are dependent. Why then do we have the saying "Correlation does not imply Causation". A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link.
statistics logic causal-diagrams
edited Jan 20 at 18:59
nbro
2,41563174
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
$endgroup$
$begingroup$
X=number of sunglasses sold per day is correlated with Y=number of icecreams sold per day. There is a relation (summer), but X is not cause of Y, nor the other way.
$endgroup$
– leonbloy
Sep 1 '17 at 17:10
add a comment |
$begingroup$
If we have two non-zero correlated random variables then they are dependent. Why then do we have the saying "Correlation does not imply Causation". A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link.
statistics logic causal-diagrams
edited Jan 20 at 18:59
nbro
2,41563174
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
$endgroup$
If we have two non-zero correlated random variables then they are dependent. Why then do we have the saying "Correlation does not imply Causation". A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link.
statistics logic causal-diagrams
statistics logic causal-diagrams
edited Jan 20 at 18:59
nbro
2,41563174
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
edited Jan 20 at 18:59
nbro
2,41563174
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
edited Jan 20 at 18:59
nbro
2,41563174
edited Jan 20 at 18:59
nbro
2,41563174
edited Jan 20 at 18:59
nbro
2,41563174
2,41563174
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
asked Jan 29 '16 at 11:20
usainlightningusainlightning
299518
299518
$begingroup$
X=number of sunglasses sold per day is correlated with Y=number of icecreams sold per day. There is a relation (summer), but X is not cause of Y, nor the other way.
$endgroup$
– leonbloy
Sep 1 '17 at 17:10
add a comment |
$begingroup$
X=number of sunglasses sold per day is correlated with Y=number of icecreams sold per day. There is a relation (summer), but X is not cause of Y, nor the other way.
$endgroup$
– leonbloy
Sep 1 '17 at 17:10
$begingroup$
X=number of sunglasses sold per day is correlated with Y=number of icecreams sold per day. There is a relation (summer), but X is not cause of Y, nor the other way.
$endgroup$
– leonbloy
Sep 1 '17 at 17:10
$begingroup$
X=number of sunglasses sold per day is correlated with Y=number of icecreams sold per day. There is a relation (summer), but X is not cause of Y, nor the other way.
$endgroup$
– leonbloy
Sep 1 '17 at 17:10
add a comment |
3 Answers
3
active
oldest
votes
$begingroup$
Things can be correlated by having a third common cause while not having any causal influence on each other. For example, a fire siren going off and flashing warning lights may be near 100% correlated but neither causes the other. Instead, there is a third causal influence that is the common cause of both, namely pulling the fire alarm.
I recommend learning about the work primarily developed by Judea Pearl on causal inference. An assumption in causal inference, called Reichenbach's common-cause assumption and sloganized as "no correlation without causation", states that two variables are only dependent if one is the cause of the other or there is a third variable causing both. So you definitely aren't completely off in your intuition.
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
$endgroup$
add a comment |
$begingroup$
There are three different concepts in play, two from probability theory and one from philosophy.
The two mathematical concepts are (in)dependence and correlation. Independence is a general property of events and random variables. Correlation is a property typically associated with random variables that are square integrable. If two square integrable variables are independent they they have zero correlation but not vice versa.
Causation is a philosophical concept that implies comparing the observable world with alternative worlds (that could have happened but did not). It is so dangerous that only lawyers and judges use it. The closest approximation that scientists use is causality, and it is mainly a negative axiom: an event cannot be the cause of another event unless it precedes the other event in time. There is a slightly adapted version in relativity theory due to the fact that time is not an absolute observable.
The statement 'correlation does not imply causation' refers to the fact that correlation can be inferred statistically, whereas causal connections imply much more a priori choices of what worlds are possible.
As an example, it is relatively easy to establish a positive correlation, and therefore also a dependence, between smoking and lung cancer. To prove that smoking causes lung cancer, however, you need to set up an experiment that effectively manages several parallel worlds, one in which a person smokes and another where they don't smoke with all other things being equal whatever that means.
As a good surrogate, scientists will approximate cause-effect relationships by multivariate correlation analysis. The best thing we can then come up with is statements like 'smoking is positively correlated to lung cancer even after factoring out annual income and level of education'. That is how research is reported in professional journals. Unfortunately the word cause will often creep back into the press release.
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
$endgroup$
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
add a comment |
$begingroup$
If we have two non-zero correlated random variables then they are dependent.
Yes, this is true. Non-zero correlation implies dependence.
Why then do we have the saying "Correlation does not imply Causation"
Because this is also true.
A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link
There is not necessarily any causal link. If two things change together, it doesn't mean that one is the cause of the other. For example, in the Summer, ice cream sales increase and, apparently, the crime also increases, but one is not the cause of the other. There is one or more hidden variables which makes these variables to change together. In general, we often think that if two things change together, then they have a causal relationship, but this is actually rarely the case.
answered Jan 20 at 18:58
nbronbro
2,41563174
$endgroup$
add a comment |
Your Answer
StackExchange.ifUsing("editor", function () {
return StackExchange.using("mathjaxEditing", function () {
StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
});
});
}, "mathjax-editing");
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "69"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
noCode: true, onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f1631870%2fif-we-have-two-non-zero-correlated-random-variables-then-why-do-we-say-that-co%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
3 Answers
3
active
oldest
votes
3 Answers
3
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
Things can be correlated by having a third common cause while not having any causal influence on each other. For example, a fire siren going off and flashing warning lights may be near 100% correlated but neither causes the other. Instead, there is a third causal influence that is the common cause of both, namely pulling the fire alarm.
I recommend learning about the work primarily developed by Judea Pearl on causal inference. An assumption in causal inference, called Reichenbach's common-cause assumption and sloganized as "no correlation without causation", states that two variables are only dependent if one is the cause of the other or there is a third variable causing both. So you definitely aren't completely off in your intuition.
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
$endgroup$
add a comment |
$begingroup$
Things can be correlated by having a third common cause while not having any causal influence on each other. For example, a fire siren going off and flashing warning lights may be near 100% correlated but neither causes the other. Instead, there is a third causal influence that is the common cause of both, namely pulling the fire alarm.
I recommend learning about the work primarily developed by Judea Pearl on causal inference. An assumption in causal inference, called Reichenbach's common-cause assumption and sloganized as "no correlation without causation", states that two variables are only dependent if one is the cause of the other or there is a third variable causing both. So you definitely aren't completely off in your intuition.
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
$endgroup$
add a comment |
$begingroup$
Things can be correlated by having a third common cause while not having any causal influence on each other. For example, a fire siren going off and flashing warning lights may be near 100% correlated but neither causes the other. Instead, there is a third causal influence that is the common cause of both, namely pulling the fire alarm.
I recommend learning about the work primarily developed by Judea Pearl on causal inference. An assumption in causal inference, called Reichenbach's common-cause assumption and sloganized as "no correlation without causation", states that two variables are only dependent if one is the cause of the other or there is a third variable causing both. So you definitely aren't completely off in your intuition.
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
$endgroup$
Things can be correlated by having a third common cause while not having any causal influence on each other. For example, a fire siren going off and flashing warning lights may be near 100% correlated but neither causes the other. Instead, there is a third causal influence that is the common cause of both, namely pulling the fire alarm.
I recommend learning about the work primarily developed by Judea Pearl on causal inference. An assumption in causal inference, called Reichenbach's common-cause assumption and sloganized as "no correlation without causation", states that two variables are only dependent if one is the cause of the other or there is a third variable causing both. So you definitely aren't completely off in your intuition.
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
answered Jan 29 '16 at 11:35
Derek ElkinsDerek Elkins
17.1k11437
17.1k11437
add a comment |
add a comment |
$begingroup$
There are three different concepts in play, two from probability theory and one from philosophy.
The two mathematical concepts are (in)dependence and correlation. Independence is a general property of events and random variables. Correlation is a property typically associated with random variables that are square integrable. If two square integrable variables are independent they they have zero correlation but not vice versa.
Causation is a philosophical concept that implies comparing the observable world with alternative worlds (that could have happened but did not). It is so dangerous that only lawyers and judges use it. The closest approximation that scientists use is causality, and it is mainly a negative axiom: an event cannot be the cause of another event unless it precedes the other event in time. There is a slightly adapted version in relativity theory due to the fact that time is not an absolute observable.
The statement 'correlation does not imply causation' refers to the fact that correlation can be inferred statistically, whereas causal connections imply much more a priori choices of what worlds are possible.
As an example, it is relatively easy to establish a positive correlation, and therefore also a dependence, between smoking and lung cancer. To prove that smoking causes lung cancer, however, you need to set up an experiment that effectively manages several parallel worlds, one in which a person smokes and another where they don't smoke with all other things being equal whatever that means.
As a good surrogate, scientists will approximate cause-effect relationships by multivariate correlation analysis. The best thing we can then come up with is statements like 'smoking is positively correlated to lung cancer even after factoring out annual income and level of education'. That is how research is reported in professional journals. Unfortunately the word cause will often creep back into the press release.
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
$endgroup$
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
add a comment |
$begingroup$
There are three different concepts in play, two from probability theory and one from philosophy.
The two mathematical concepts are (in)dependence and correlation. Independence is a general property of events and random variables. Correlation is a property typically associated with random variables that are square integrable. If two square integrable variables are independent they they have zero correlation but not vice versa.
Causation is a philosophical concept that implies comparing the observable world with alternative worlds (that could have happened but did not). It is so dangerous that only lawyers and judges use it. The closest approximation that scientists use is causality, and it is mainly a negative axiom: an event cannot be the cause of another event unless it precedes the other event in time. There is a slightly adapted version in relativity theory due to the fact that time is not an absolute observable.
The statement 'correlation does not imply causation' refers to the fact that correlation can be inferred statistically, whereas causal connections imply much more a priori choices of what worlds are possible.
As an example, it is relatively easy to establish a positive correlation, and therefore also a dependence, between smoking and lung cancer. To prove that smoking causes lung cancer, however, you need to set up an experiment that effectively manages several parallel worlds, one in which a person smokes and another where they don't smoke with all other things being equal whatever that means.
As a good surrogate, scientists will approximate cause-effect relationships by multivariate correlation analysis. The best thing we can then come up with is statements like 'smoking is positively correlated to lung cancer even after factoring out annual income and level of education'. That is how research is reported in professional journals. Unfortunately the word cause will often creep back into the press release.
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
$endgroup$
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
add a comment |
$begingroup$
There are three different concepts in play, two from probability theory and one from philosophy.
The two mathematical concepts are (in)dependence and correlation. Independence is a general property of events and random variables. Correlation is a property typically associated with random variables that are square integrable. If two square integrable variables are independent they they have zero correlation but not vice versa.
Causation is a philosophical concept that implies comparing the observable world with alternative worlds (that could have happened but did not). It is so dangerous that only lawyers and judges use it. The closest approximation that scientists use is causality, and it is mainly a negative axiom: an event cannot be the cause of another event unless it precedes the other event in time. There is a slightly adapted version in relativity theory due to the fact that time is not an absolute observable.
The statement 'correlation does not imply causation' refers to the fact that correlation can be inferred statistically, whereas causal connections imply much more a priori choices of what worlds are possible.
As an example, it is relatively easy to establish a positive correlation, and therefore also a dependence, between smoking and lung cancer. To prove that smoking causes lung cancer, however, you need to set up an experiment that effectively manages several parallel worlds, one in which a person smokes and another where they don't smoke with all other things being equal whatever that means.
As a good surrogate, scientists will approximate cause-effect relationships by multivariate correlation analysis. The best thing we can then come up with is statements like 'smoking is positively correlated to lung cancer even after factoring out annual income and level of education'. That is how research is reported in professional journals. Unfortunately the word cause will often creep back into the press release.
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
$endgroup$
There are three different concepts in play, two from probability theory and one from philosophy.
The two mathematical concepts are (in)dependence and correlation. Independence is a general property of events and random variables. Correlation is a property typically associated with random variables that are square integrable. If two square integrable variables are independent they they have zero correlation but not vice versa.
Causation is a philosophical concept that implies comparing the observable world with alternative worlds (that could have happened but did not). It is so dangerous that only lawyers and judges use it. The closest approximation that scientists use is causality, and it is mainly a negative axiom: an event cannot be the cause of another event unless it precedes the other event in time. There is a slightly adapted version in relativity theory due to the fact that time is not an absolute observable.
The statement 'correlation does not imply causation' refers to the fact that correlation can be inferred statistically, whereas causal connections imply much more a priori choices of what worlds are possible.
As an example, it is relatively easy to establish a positive correlation, and therefore also a dependence, between smoking and lung cancer. To prove that smoking causes lung cancer, however, you need to set up an experiment that effectively manages several parallel worlds, one in which a person smokes and another where they don't smoke with all other things being equal whatever that means.
As a good surrogate, scientists will approximate cause-effect relationships by multivariate correlation analysis. The best thing we can then come up with is statements like 'smoking is positively correlated to lung cancer even after factoring out annual income and level of education'. That is how research is reported in professional journals. Unfortunately the word cause will often creep back into the press release.
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
edited Jan 29 '16 at 11:54
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
answered Jan 29 '16 at 11:36
JustpassingbyJustpassingby
9,094726
9,094726
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
add a comment |
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
$begingroup$
Causation is not just a philosophical topic. It is also a statistical or machine learning topic. See causal inference.
$endgroup$
– nbro
Jan 20 at 18:52
add a comment |
$begingroup$
If we have two non-zero correlated random variables then they are dependent.
Yes, this is true. Non-zero correlation implies dependence.
Why then do we have the saying "Correlation does not imply Causation"
Because this is also true.
A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link
There is not necessarily any causal link. If two things change together, it doesn't mean that one is the cause of the other. For example, in the Summer, ice cream sales increase and, apparently, the crime also increases, but one is not the cause of the other. There is one or more hidden variables which makes these variables to change together. In general, we often think that if two things change together, then they have a causal relationship, but this is actually rarely the case.
answered Jan 20 at 18:58
nbronbro
2,41563174
$endgroup$
add a comment |
$begingroup$
If we have two non-zero correlated random variables then they are dependent.
Yes, this is true. Non-zero correlation implies dependence.
Why then do we have the saying "Correlation does not imply Causation"
Because this is also true.
A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link
There is not necessarily any causal link. If two things change together, it doesn't mean that one is the cause of the other. For example, in the Summer, ice cream sales increase and, apparently, the crime also increases, but one is not the cause of the other. There is one or more hidden variables which makes these variables to change together. In general, we often think that if two things change together, then they have a causal relationship, but this is actually rarely the case.
answered Jan 20 at 18:58
nbronbro
2,41563174
$endgroup$
add a comment |
$begingroup$
If we have two non-zero correlated random variables then they are dependent.
Yes, this is true. Non-zero correlation implies dependence.
Why then do we have the saying "Correlation does not imply Causation"
Because this is also true.
A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link
There is not necessarily any causal link. If two things change together, it doesn't mean that one is the cause of the other. For example, in the Summer, ice cream sales increase and, apparently, the crime also increases, but one is not the cause of the other. There is one or more hidden variables which makes these variables to change together. In general, we often think that if two things change together, then they have a causal relationship, but this is actually rarely the case.
answered Jan 20 at 18:58
nbronbro
2,41563174
$endgroup$
If we have two non-zero correlated random variables then they are dependent.
Yes, this is true. Non-zero correlation implies dependence.
Why then do we have the saying "Correlation does not imply Causation"
Because this is also true.
A change in one variable may not cause exactly the same change in another but there is at least some 'causal' link
There is not necessarily any causal link. If two things change together, it doesn't mean that one is the cause of the other. For example, in the Summer, ice cream sales increase and, apparently, the crime also increases, but one is not the cause of the other. There is one or more hidden variables which makes these variables to change together. In general, we often think that if two things change together, then they have a causal relationship, but this is actually rarely the case.
answered Jan 20 at 18:58
nbronbro
2,41563174
answered Jan 20 at 18:58
nbronbro
2,41563174
answered Jan 20 at 18:58
nbronbro
2,41563174
answered Jan 20 at 18:58
nbronbro
2,41563174
2,41563174
add a comment |
add a comment |
Thanks for contributing an answer to Mathematics Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f1631870%2fif-we-have-two-non-zero-correlated-random-variables-then-why-do-we-say-that-co%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
$begingroup$
X=number of sunglasses sold per day is correlated with Y=number of icecreams sold per day. There is a relation (summer), but X is not cause of Y, nor the other way.
$endgroup$
– leonbloy
Sep 1 '17 at 17:10