Are there any examples of a variable being normally distributed that is *not* due to the Central Limit Theorem?Central limit theorem and the law of large numbersCentral limit theorem when the mean is not constantWhy does the central limit theorem work with a single sample?The central limit theorem, What it meansUnderstanding the Central Limit Theorem (CLT)How does the Central Limit Theorem show that the Binomial Distribution is approximately Normal for a large value of n?For which parameters does the Central Limit Theorem work?What distributions don't follow the central limit theorem?How can the central limit theorem hold for distributions which have limits on the random variable?Are there any examples of where the central limit theorem does not hold?
What's the point of deactivating Num Lock on login screens?
In Japanese, what’s the difference between “Tonari ni” (となりに) and “Tsugi” (つぎ)? When would you use one over the other?
Has the BBC provided arguments for saying Brexit being cancelled is unlikely?
What would happen to a modern skyscraper if it rains micro blackholes?
What is the offset in a seaplane's hull?
What do you call a Matrix-like slowdown and camera movement effect?
Problem of parity - Can we draw a closed path made up of 20 line segments...
What are these boxed doors outside store fronts in New York?
Can an x86 CPU running in real mode be considered to be basically an 8086 CPU?
How can I prevent hyper evolved versions of regular creatures from wiping out their cousins?
Modeling an IPv4 Address
To string or not to string
How is it possible to have an ability score that is less than 3?
Can I ask the recruiters in my resume to put the reason why I am rejected?
Prove that NP is closed under karp reduction?
Risk of getting Chronic Wasting Disease (CWD) in the United States?
What is the word for reserving something for yourself before others do?
How to test if a transaction is standard without spending real money?
How does one intimidate enemies without having the capacity for violence?
What does CI-V stand for?
Today is the Center
What typically incentivizes a professor to change jobs to a lower ranking university?
Can a Warlock become Neutral Good?
A newer friend of my brother's gave him a load of baseball cards that are supposedly extremely valuable. Is this a scam?
Are there any examples of a variable being normally distributed that is *not* due to the Central Limit Theorem?
Central limit theorem and the law of large numbersCentral limit theorem when the mean is not constantWhy does the central limit theorem work with a single sample?The central limit theorem, What it meansUnderstanding the Central Limit Theorem (CLT)How does the Central Limit Theorem show that the Binomial Distribution is approximately Normal for a large value of n?For which parameters does the Central Limit Theorem work?What distributions don't follow the central limit theorem?How can the central limit theorem hold for distributions which have limits on the random variable?Are there any examples of where the central limit theorem does not hold?
.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty margin-bottom:0;
$begingroup$
The normal distribution seems unintuitive until you learn the CLT, which explains why it is so prevalent in real life. But does it ever arise as the "natural" distribution for some quantity?
normal-distribution central-limit-theorem
$endgroup$
add a comment |
$begingroup$
The normal distribution seems unintuitive until you learn the CLT, which explains why it is so prevalent in real life. But does it ever arise as the "natural" distribution for some quantity?
normal-distribution central-limit-theorem
$endgroup$
4
$begingroup$
The physical theory of diffusion, to the extent it is applicable to any system, predicts Normal distributions of quantities (like temperature or concentration) that originate at a point. Indeed, a great many systems are diffusive (options prices, particle transport in homogeneous media, etc.), suggesting that examples are abundant assuming one is not so naive as to suppose that a Normal distribution must hold exactly out to unrealistically large or small values--that would be a misunderstanding of all physical theory.
$endgroup$
– whuber♦
Apr 3 at 22:27
2
$begingroup$
The normal distribution seems unintuitive until you learn that it maximizes the entropy under the constraint of fixed variance.
$endgroup$
– leonbloy
Apr 4 at 3:18
add a comment |
$begingroup$
The normal distribution seems unintuitive until you learn the CLT, which explains why it is so prevalent in real life. But does it ever arise as the "natural" distribution for some quantity?
normal-distribution central-limit-theorem
$endgroup$
The normal distribution seems unintuitive until you learn the CLT, which explains why it is so prevalent in real life. But does it ever arise as the "natural" distribution for some quantity?
normal-distribution central-limit-theorem
normal-distribution central-limit-theorem
asked Apr 3 at 21:18
gardenheadgardenhead
1764
1764
4
$begingroup$
The physical theory of diffusion, to the extent it is applicable to any system, predicts Normal distributions of quantities (like temperature or concentration) that originate at a point. Indeed, a great many systems are diffusive (options prices, particle transport in homogeneous media, etc.), suggesting that examples are abundant assuming one is not so naive as to suppose that a Normal distribution must hold exactly out to unrealistically large or small values--that would be a misunderstanding of all physical theory.
$endgroup$
– whuber♦
Apr 3 at 22:27
2
$begingroup$
The normal distribution seems unintuitive until you learn that it maximizes the entropy under the constraint of fixed variance.
$endgroup$
– leonbloy
Apr 4 at 3:18
add a comment |
4
$begingroup$
The physical theory of diffusion, to the extent it is applicable to any system, predicts Normal distributions of quantities (like temperature or concentration) that originate at a point. Indeed, a great many systems are diffusive (options prices, particle transport in homogeneous media, etc.), suggesting that examples are abundant assuming one is not so naive as to suppose that a Normal distribution must hold exactly out to unrealistically large or small values--that would be a misunderstanding of all physical theory.
$endgroup$
– whuber♦
Apr 3 at 22:27
2
$begingroup$
The normal distribution seems unintuitive until you learn that it maximizes the entropy under the constraint of fixed variance.
$endgroup$
– leonbloy
Apr 4 at 3:18
4
4
$begingroup$
The physical theory of diffusion, to the extent it is applicable to any system, predicts Normal distributions of quantities (like temperature or concentration) that originate at a point. Indeed, a great many systems are diffusive (options prices, particle transport in homogeneous media, etc.), suggesting that examples are abundant assuming one is not so naive as to suppose that a Normal distribution must hold exactly out to unrealistically large or small values--that would be a misunderstanding of all physical theory.
$endgroup$
– whuber♦
Apr 3 at 22:27
$begingroup$
The physical theory of diffusion, to the extent it is applicable to any system, predicts Normal distributions of quantities (like temperature or concentration) that originate at a point. Indeed, a great many systems are diffusive (options prices, particle transport in homogeneous media, etc.), suggesting that examples are abundant assuming one is not so naive as to suppose that a Normal distribution must hold exactly out to unrealistically large or small values--that would be a misunderstanding of all physical theory.
$endgroup$
– whuber♦
Apr 3 at 22:27
2
2
$begingroup$
The normal distribution seems unintuitive until you learn that it maximizes the entropy under the constraint of fixed variance.
$endgroup$
– leonbloy
Apr 4 at 3:18
$begingroup$
The normal distribution seems unintuitive until you learn that it maximizes the entropy under the constraint of fixed variance.
$endgroup$
– leonbloy
Apr 4 at 3:18
add a comment |
2 Answers
2
active
oldest
votes
$begingroup$
To an extent I think this this may be a philosophical issue as much as a statistical one.
A lot of naturally occurring phenomena are approximately normally distributed. One can argue
whether the underlying cause of that may be something like the CLT:
Heights of people may be considered as the the sum of many smaller causes (perhaps independent, unlikely identically distributed): lengths of various bones, or results of various gene expressions, or results of many dietary
influences, or some combination of all of the above.Test scores may be considered as the sums of scores on many individual test questions (possibly identically distributed, unlikely entirely independent).
Distance a particle travels in one dimension as a result of Brownian motion in a fluid: Motion may be considered abstractly as a random walk resulting from IID random hits by molecules.
One example where the CLT is not necessarily involved is the dispersion of shots around a bull's eye: The distance from the bull's eye can be modeled as a Rayleigh
distribution (proportional to square root of chi-sq with 2 DF) and the counterclockwise angle from the the positive horizontal axis can be modeled as uniform on $(0, 2pi).$ Then after changing from polar to rectangular coordinates, distances in horizontal (x) and
vertical (y) directions turn out to be uncorrelated bivariate normal. [This is the essence of the Box-Muller transformation, which you can google.] However, the normal x and y coordinates might be considered as the sum of many small inaccuracies in targeting, which might justify a CLT-related mechanism in the background.
In a historical sense, the widespread use of normal (Gaussian) distributions instead of double-exponential (Laplace) distributions to model astronomical observations may be partly due to the CLT. In the early days of modeling errors of such observations, there was a debate between Gauss and Laplace, each arguing for his own favorite distribution. For various reasons, the normal model has won out. One can argue that one reason for the eventual success of the normal distribution was mathematical convenience based on normal limits of the CLT. This seems to be true even when it is unclear which family of distributions provides the better fit. (Even now, there are still astronomers who feel that the "one best observation" made by
a meticulous, respected astronomer is bound to be a better value than than the average of many observations made by presumably less-gifted observers. In effect, they would prefer no intervention at all by statisticians.)
$endgroup$
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
add a comment |
$begingroup$
Lots of naturally occurring variables are normally distributed. Heights of humans? Size of animal colonies?
New contributor
$endgroup$
5
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
1
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
2
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
2
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
2
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
|
show 7 more comments
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: "65"
;
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: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
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
,
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%2fstats.stackexchange.com%2fquestions%2f401055%2fare-there-any-examples-of-a-variable-being-normally-distributed-that-is-not-du%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
2 Answers
2
active
oldest
votes
2 Answers
2
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
To an extent I think this this may be a philosophical issue as much as a statistical one.
A lot of naturally occurring phenomena are approximately normally distributed. One can argue
whether the underlying cause of that may be something like the CLT:
Heights of people may be considered as the the sum of many smaller causes (perhaps independent, unlikely identically distributed): lengths of various bones, or results of various gene expressions, or results of many dietary
influences, or some combination of all of the above.Test scores may be considered as the sums of scores on many individual test questions (possibly identically distributed, unlikely entirely independent).
Distance a particle travels in one dimension as a result of Brownian motion in a fluid: Motion may be considered abstractly as a random walk resulting from IID random hits by molecules.
One example where the CLT is not necessarily involved is the dispersion of shots around a bull's eye: The distance from the bull's eye can be modeled as a Rayleigh
distribution (proportional to square root of chi-sq with 2 DF) and the counterclockwise angle from the the positive horizontal axis can be modeled as uniform on $(0, 2pi).$ Then after changing from polar to rectangular coordinates, distances in horizontal (x) and
vertical (y) directions turn out to be uncorrelated bivariate normal. [This is the essence of the Box-Muller transformation, which you can google.] However, the normal x and y coordinates might be considered as the sum of many small inaccuracies in targeting, which might justify a CLT-related mechanism in the background.
In a historical sense, the widespread use of normal (Gaussian) distributions instead of double-exponential (Laplace) distributions to model astronomical observations may be partly due to the CLT. In the early days of modeling errors of such observations, there was a debate between Gauss and Laplace, each arguing for his own favorite distribution. For various reasons, the normal model has won out. One can argue that one reason for the eventual success of the normal distribution was mathematical convenience based on normal limits of the CLT. This seems to be true even when it is unclear which family of distributions provides the better fit. (Even now, there are still astronomers who feel that the "one best observation" made by
a meticulous, respected astronomer is bound to be a better value than than the average of many observations made by presumably less-gifted observers. In effect, they would prefer no intervention at all by statisticians.)
$endgroup$
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
add a comment |
$begingroup$
To an extent I think this this may be a philosophical issue as much as a statistical one.
A lot of naturally occurring phenomena are approximately normally distributed. One can argue
whether the underlying cause of that may be something like the CLT:
Heights of people may be considered as the the sum of many smaller causes (perhaps independent, unlikely identically distributed): lengths of various bones, or results of various gene expressions, or results of many dietary
influences, or some combination of all of the above.Test scores may be considered as the sums of scores on many individual test questions (possibly identically distributed, unlikely entirely independent).
Distance a particle travels in one dimension as a result of Brownian motion in a fluid: Motion may be considered abstractly as a random walk resulting from IID random hits by molecules.
One example where the CLT is not necessarily involved is the dispersion of shots around a bull's eye: The distance from the bull's eye can be modeled as a Rayleigh
distribution (proportional to square root of chi-sq with 2 DF) and the counterclockwise angle from the the positive horizontal axis can be modeled as uniform on $(0, 2pi).$ Then after changing from polar to rectangular coordinates, distances in horizontal (x) and
vertical (y) directions turn out to be uncorrelated bivariate normal. [This is the essence of the Box-Muller transformation, which you can google.] However, the normal x and y coordinates might be considered as the sum of many small inaccuracies in targeting, which might justify a CLT-related mechanism in the background.
In a historical sense, the widespread use of normal (Gaussian) distributions instead of double-exponential (Laplace) distributions to model astronomical observations may be partly due to the CLT. In the early days of modeling errors of such observations, there was a debate between Gauss and Laplace, each arguing for his own favorite distribution. For various reasons, the normal model has won out. One can argue that one reason for the eventual success of the normal distribution was mathematical convenience based on normal limits of the CLT. This seems to be true even when it is unclear which family of distributions provides the better fit. (Even now, there are still astronomers who feel that the "one best observation" made by
a meticulous, respected astronomer is bound to be a better value than than the average of many observations made by presumably less-gifted observers. In effect, they would prefer no intervention at all by statisticians.)
$endgroup$
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
add a comment |
$begingroup$
To an extent I think this this may be a philosophical issue as much as a statistical one.
A lot of naturally occurring phenomena are approximately normally distributed. One can argue
whether the underlying cause of that may be something like the CLT:
Heights of people may be considered as the the sum of many smaller causes (perhaps independent, unlikely identically distributed): lengths of various bones, or results of various gene expressions, or results of many dietary
influences, or some combination of all of the above.Test scores may be considered as the sums of scores on many individual test questions (possibly identically distributed, unlikely entirely independent).
Distance a particle travels in one dimension as a result of Brownian motion in a fluid: Motion may be considered abstractly as a random walk resulting from IID random hits by molecules.
One example where the CLT is not necessarily involved is the dispersion of shots around a bull's eye: The distance from the bull's eye can be modeled as a Rayleigh
distribution (proportional to square root of chi-sq with 2 DF) and the counterclockwise angle from the the positive horizontal axis can be modeled as uniform on $(0, 2pi).$ Then after changing from polar to rectangular coordinates, distances in horizontal (x) and
vertical (y) directions turn out to be uncorrelated bivariate normal. [This is the essence of the Box-Muller transformation, which you can google.] However, the normal x and y coordinates might be considered as the sum of many small inaccuracies in targeting, which might justify a CLT-related mechanism in the background.
In a historical sense, the widespread use of normal (Gaussian) distributions instead of double-exponential (Laplace) distributions to model astronomical observations may be partly due to the CLT. In the early days of modeling errors of such observations, there was a debate between Gauss and Laplace, each arguing for his own favorite distribution. For various reasons, the normal model has won out. One can argue that one reason for the eventual success of the normal distribution was mathematical convenience based on normal limits of the CLT. This seems to be true even when it is unclear which family of distributions provides the better fit. (Even now, there are still astronomers who feel that the "one best observation" made by
a meticulous, respected astronomer is bound to be a better value than than the average of many observations made by presumably less-gifted observers. In effect, they would prefer no intervention at all by statisticians.)
$endgroup$
To an extent I think this this may be a philosophical issue as much as a statistical one.
A lot of naturally occurring phenomena are approximately normally distributed. One can argue
whether the underlying cause of that may be something like the CLT:
Heights of people may be considered as the the sum of many smaller causes (perhaps independent, unlikely identically distributed): lengths of various bones, or results of various gene expressions, or results of many dietary
influences, or some combination of all of the above.Test scores may be considered as the sums of scores on many individual test questions (possibly identically distributed, unlikely entirely independent).
Distance a particle travels in one dimension as a result of Brownian motion in a fluid: Motion may be considered abstractly as a random walk resulting from IID random hits by molecules.
One example where the CLT is not necessarily involved is the dispersion of shots around a bull's eye: The distance from the bull's eye can be modeled as a Rayleigh
distribution (proportional to square root of chi-sq with 2 DF) and the counterclockwise angle from the the positive horizontal axis can be modeled as uniform on $(0, 2pi).$ Then after changing from polar to rectangular coordinates, distances in horizontal (x) and
vertical (y) directions turn out to be uncorrelated bivariate normal. [This is the essence of the Box-Muller transformation, which you can google.] However, the normal x and y coordinates might be considered as the sum of many small inaccuracies in targeting, which might justify a CLT-related mechanism in the background.
In a historical sense, the widespread use of normal (Gaussian) distributions instead of double-exponential (Laplace) distributions to model astronomical observations may be partly due to the CLT. In the early days of modeling errors of such observations, there was a debate between Gauss and Laplace, each arguing for his own favorite distribution. For various reasons, the normal model has won out. One can argue that one reason for the eventual success of the normal distribution was mathematical convenience based on normal limits of the CLT. This seems to be true even when it is unclear which family of distributions provides the better fit. (Even now, there are still astronomers who feel that the "one best observation" made by
a meticulous, respected astronomer is bound to be a better value than than the average of many observations made by presumably less-gifted observers. In effect, they would prefer no intervention at all by statisticians.)
edited Apr 4 at 0:56
answered Apr 3 at 21:53
BruceETBruceET
6,3431721
6,3431721
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
add a comment |
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
$begingroup$
Yep. Still fixing typos. Thanks for noticing this one. Same error in 'test scores' also fixed.
$endgroup$
– BruceET
Apr 3 at 22:20
add a comment |
$begingroup$
Lots of naturally occurring variables are normally distributed. Heights of humans? Size of animal colonies?
New contributor
$endgroup$
5
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
1
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
2
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
2
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
2
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
|
show 7 more comments
$begingroup$
Lots of naturally occurring variables are normally distributed. Heights of humans? Size of animal colonies?
New contributor
$endgroup$
5
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
1
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
2
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
2
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
2
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
|
show 7 more comments
$begingroup$
Lots of naturally occurring variables are normally distributed. Heights of humans? Size of animal colonies?
New contributor
$endgroup$
Lots of naturally occurring variables are normally distributed. Heights of humans? Size of animal colonies?
New contributor
New contributor
answered Apr 3 at 21:20
HappyHappy
112
112
New contributor
New contributor
5
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
1
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
2
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
2
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
2
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
|
show 7 more comments
5
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
1
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
2
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
2
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
2
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
5
5
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
$begingroup$
But are those examples really normally distributed, or is that just a useful approximation? True normally distributed random variables take negative values with positive probability.
$endgroup$
– Artem Mavrin
Apr 3 at 21:27
1
1
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
$begingroup$
@Happy Actually neither example given here is normally distributed because the support of the normal distribution is -infinity to +infinity and the examples given can never be zero or less. In each case the normal distribution might be a useful approximation, but not if you were interested in the tails of the distribution.
$endgroup$
– JeremyC
Apr 3 at 21:28
2
2
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
$begingroup$
Human height is the result of the sum of (approximately) independent genes, so they actually are due to the CLT.
$endgroup$
– gardenhead
Apr 3 at 22:09
2
2
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
$begingroup$
@ArtemMavrin: getting a negative height would be something like 8+ standard deviations. If one objects to a normal approximation not being valid because it places zero probability mass beyond 8 sd's, you might as well also complain that a truly Normally distributed value is irrational with probability 1, yet all our measurements are rational numbers.
$endgroup$
– Cliff AB
Apr 3 at 22:15
2
2
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
$begingroup$
@gardenhead have a read of the statement of the central limit theorem and note where it doesn't match what you said (number of genes is finite while CLT is about the limiting distribution of a standardized sum as $ntoinfty$. (By the way, there's another theorem that says that a finite sum of iid non-normal variates cannot be normal; this theorem doesn't contradict the actual CLT!)
$endgroup$
– Glen_b♦
Apr 4 at 0:36
|
show 7 more comments
Thanks for contributing an answer to Cross Validated!
- 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%2fstats.stackexchange.com%2fquestions%2f401055%2fare-there-any-examples-of-a-variable-being-normally-distributed-that-is-not-du%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
4
$begingroup$
The physical theory of diffusion, to the extent it is applicable to any system, predicts Normal distributions of quantities (like temperature or concentration) that originate at a point. Indeed, a great many systems are diffusive (options prices, particle transport in homogeneous media, etc.), suggesting that examples are abundant assuming one is not so naive as to suppose that a Normal distribution must hold exactly out to unrealistically large or small values--that would be a misunderstanding of all physical theory.
$endgroup$
– whuber♦
Apr 3 at 22:27
2
$begingroup$
The normal distribution seems unintuitive until you learn that it maximizes the entropy under the constraint of fixed variance.
$endgroup$
– leonbloy
Apr 4 at 3:18