variance
出典: meddic
 n.
 関
 disagreement、discordance、discrete、dispersal、disperse、dispersion、fluctuate、fluctuation、inconsistency、inconsistent、intersperse、oscillate、variation、vary
WordNet ［license wordnet］
「the second moment around the mean; the expected value of the square of the deviations of a random variable from its mean value」PrepTutorEJDIC ［license prepejdic］
「変化,多様性;相違,不一致」WordNet ［license wordnet］
「an official dispensation to act contrary to a rule or regulation (typically a building regulation); "a zoning variance"」
Wikipedia preview
出典(authority):フリー百科事典『ウィキペディア（Wikipedia）』「2017/07/12 03:44:58」(JST)
wiki en
[Wiki en表示]In probability theory and statistics, variance is the expectation of the squared deviation of a random variable from its mean, and it informally measures how far a set of (random) numbers are spread out from their mean. The variance has a central role in statistics. It is used in descriptive statistics, statistical inference, hypothesis testing, goodness of fit, and Monte Carlo sampling, amongst many others. This makes it a central quantity in numerous fields such as physics, biology, chemistry, cryptography, economics, and finance. The variance is the square of the standard deviation, the second central moment of a distribution, and the covariance of the random variable with itself, and it is often represented by $\sigma ^{2}$, $s^{2}$, or $\operatorname {Var} (X)$.
Contents
 1 Definition
 1.1 Continuous random variable
 1.2 Discrete random variable
 2 Examples
 2.1 Normal distribution
 2.2 Exponential distribution
 2.3 Poisson distribution
 2.4 Binomial distribution
 2.5 Fair die
 3 Properties
 3.1 Basic properties
 3.2 Sum of uncorrelated variables (Bienaymé formula)
 3.3 Sum of correlated variables
 3.4 Matrix notation for the variance of a linear combination
 3.5 Weighted sum of variables
 3.6 Product of independent variables
 3.7 Product of statistically dependent variables
 3.8 Decomposition
 3.9 Formulae for the variance
 3.10 Calculation from the CDF
 3.11 Characteristic property
 3.12 Units of measurement
 4 Approximating the variance of a function
 5 Population variance and sample variance
 5.1 Population variance
 5.2 Sample variance
 5.3 Distribution of the sample variance
 5.4 Samuelson's inequality
 5.5 Relations with the harmonic and arithmetic means
 6 Tests of equality of variances
 7 History
 8 Moment of inertia
 9 Semivariance
 10 Generalizations
 11 See also
 12 Notes
Definition
The variance of a random variable $X$ is the expected value of the squared deviation from the mean of $X$, $\mu =\operatorname {E} [X]$:
 $$
Var ( X ) = E [ ( X − μ ) 2 ] . {\displaystyle \operatorname {Var} (X)=\operatorname {E} \left[(X\mu )^{2}\right].}
This definition encompasses random variables that are generated by processes that are discrete, continuous, neither, or mixed. The variance can also be thought of as the covariance of a random variable with itself:
 $$
Var ( X ) = Cov ( X , X ) . {\displaystyle \operatorname {Var} (X)=\operatorname {Cov} (X,X).}
The variance is also equivalent to the second cumulant of a probability distribution that generates $X$. The variance is typically designated as $\operatorname {Var} (X)$, $\sigma _{X}^{2}$, or simply $\sigma ^{2}$ (pronounced "sigma squared"). The expression for the variance can be expanded:
 $$
Var ( X ) = E [ ( X − E [ X ] ) 2 ] = E [ X 2 − 2 X E [ X ] + E [ X ] 2 ] = E [ X 2 ] − 2 E [ X ] E [ X ] + E [ X ] 2 = E [ X 2 ] − E [ X ] 2 {\displaystyle {\begin{aligned}\operatorname {Var} (X)&=\operatorname {E} \left[(X\operatorname {E} [X])^{2}\right]\\&=\operatorname {E} \left[X^{2}2X\operatorname {E} [X]+\operatorname {E} [X]^{2}\right]\\&=\operatorname {E} \left[X^{2}\right]2\operatorname {E} [X]\operatorname {E} [X]+\operatorname {E} [X]^{2}\\&=\operatorname {E} \left[X^{2}\right]\operatorname {E} [X]^{2}\end{aligned}}}
A mnemonic for the above expression is "mean of square minus square of mean". On computational floating point arithmetic, this equation should not be used, because it suffers from catastrophic cancellation if the two components of the equation are similar in magnitude. There exist numerically stable alternatives.
Continuous random variable
If the random variable $X$ represents samples generated by a continuous distribution with probability density function $f(x)$, then the population variance is given by
 $$
Var ( X ) = σ 2 = ∫ ( x − μ ) 2 f ( x ) d x = ∫ x 2 f ( x ) d x − 2 μ ∫ x f ( x ) d x + ∫ μ 2 f ( x ) d x = ∫ x 2 f ( x ) d x − μ 2 {\displaystyle \operatorname {Var} (X)=\sigma ^{2}=\int (x\mu )^{2}\,f(x)\,dx\,=\int x^{2}f(x)\,dx2\mu \int xf(x)\,dx+\int \mu ^{2}f(x)\,dx=\int x^{2}\,f(x)\,dx\,\mu ^{2}}
where $\mu$ is the expected value
 $$
μ = ∫ x f ( x ) d x {\displaystyle \mu =\int x\,f(x)\,dx\,}
and where the integrals are definite integrals taken for $x$ ranging over the range of $X$.
If a continuous distribution does not have an expected value, as is the case for the Cauchy distribution, it does not have a variance either. Many other distributions for which the expected value does exist also do not have a finite variance because the integral in the variance definition diverges. An example is a Pareto distribution whose index $k$ satisfies $1<k\leq 2$.
Discrete random variable
If the generator of random variable $X$ is discrete with probability mass function $x_{1}\mapsto p_{1},x_{2}\mapsto p_{2},\ldots ,x_{n}\mapsto p_{n}$ then
 $$
Var ( X ) = ∑ i = 1 n p i ⋅ ( x i − μ ) 2 , {\displaystyle \operatorname {Var} (X)=\sum _{i=1}^{n}p_{i}\cdot (x_{i}\mu )^{2},}
or equivalently
 $$
Var ( X ) = ∑ i = 1 n p i x i 2 − μ 2 , {\displaystyle \operatorname {Var} (X)=\sum _{i=1}^{n}p_{i}x_{i}^{2}\mu ^{2},}
where $\mu$ is the average value, i.e.
 $$
μ = ∑ i = 1 n p i ⋅ x i . {\displaystyle \mu =\sum _{i=1}^{n}p_{i}\cdot x_{i}.}
(When such a discrete weighted variance is specified by weights whose sum is not 1, then one divides by the sum of the weights.)
The variance of a set of $n$ equally likely values can be written as
 $$
Var ( X ) = 1 n ∑ i = 1 n ( x i − μ ) 2 , {\displaystyle \operatorname {Var} (X)={\frac {1}{n}}\sum _{i=1}^{n}(x_{i}\mu )^{2},}
where $\mu$ is the expected value, i.e.,
 $$
μ = 1 n ∑ i = 1 n x i . {\displaystyle \mu ={\frac {1}{n}}\sum _{i=1}^{n}x_{i}.}
The variance of a set of $n$ equally likely values can be equivalently expressed, without directly referring to the mean, in terms of squared deviations of all points from each other:^{[1]}
 $$
Var ( X ) = 1 n 2 ∑ i = 1 n ∑ j = 1 n 1 2 ( x i − x j ) 2 = 1 n 2 ∑ i ∑ j > i ( x i − x j ) 2 . {\displaystyle \operatorname {Var} (X)={\frac {1}{n^{2}}}\sum _{i=1}^{n}\sum _{j=1}^{n}{\frac {1}{2}}(x_{i}x_{j})^{2}={\frac {1}{n^{2}}}\sum _{i}\sum _{j>i}(x_{i}x_{j})^{2}.}
Examples
Normal distribution
The normal distribution with parameters $\mu$ and $\sigma$ is a continuous distribution whose probability density function is given by
 $$
f ( x ) = 1 2 π σ 2 e − ( x − μ ) 2 2 σ 2 . {\displaystyle f(x)={\frac {1}{\sqrt {2\pi \sigma ^{2}}}}e^{{\frac {(x\mu )^{2}}{2\sigma ^{2}}}}.}
In this distribution, $\operatorname {E} [X]=\mu$ and the variance $\operatorname {Var} (X)$ is related with $\sigma$ via
 $$
Var ( X ) = ∫ − ∞ ∞ x 2 2 π σ 2 e − ( x − μ ) 2 2 σ 2 d x − μ 2 = σ 2 . {\displaystyle \operatorname {Var} (X)=\int _{\infty }^{\infty }{\frac {x^{2}}{\sqrt {2\pi \sigma ^{2}}}}e^{{\frac {(x\mu )^{2}}{2\sigma ^{2}}}}\,dx\mu ^{2}=\sigma ^{2}.}
The role of the normal distribution in the central limit theorem is in part responsible for the prevalence of the variance in probability and statistics.
Exponential distribution
The exponential distribution with parameter $\lambda$ is a continuous distribution whose support is the semiinfinite interval $[0,\infty )$. Its probability density function is given by
 $$
f ( x ) = λ e − λ x {\displaystyle f(x)=\lambda e^{\lambda x}}
and it has expected value $\mu =\lambda ^{1}$. The variance is equal to
 $$
Var ( X ) = ∫ 0 ∞ x 2 λ e − λ x d x − μ 2 = λ − 2 . {\displaystyle \operatorname {Var} (X)=\int _{0}^{\infty }x^{2}\lambda e^{\lambda x}\,dx\mu ^{2}=\lambda ^{2}.}
So for an exponentially distributed random variable, $\sigma ^{2}=\mu ^{2}.$
Poisson distribution
The Poisson distribution with parameter $\lambda$ is a discrete distribution for $k=0,1,2,\ldots$. Its probability mass function is given by
 $$
p ( k ) = λ k k ! e − λ , {\displaystyle p(k)={\frac {\lambda ^{k}}{k!}}e^{\lambda },}
and it has expected value $\mu =\lambda$. The variance is equal to
 $$
Var ( X ) = ( ∑ k = 0 ∞ k 2 λ k k ! e − λ ) − μ 2 = λ , {\displaystyle \operatorname {Var} (X)=\left(\sum _{k=0}^{\infty }k^{2}{\frac {\lambda ^{k}}{k!}}e^{\lambda }\right)\mu ^{2}=\lambda ,}
So for a Poissondistributed random variable, $\sigma ^{2}=\mu$.
Binomial distribution
The binomial distribution with parameters $n$ and $p$ is a discrete distribution for $k=0,1,2,\ldots ,n$. Its probability mass function is given by
 $$
p ( k ) = ( n k ) p k ( 1 − p ) n − k , {\displaystyle p(k)={n \choose k}p^{k}(1p)^{nk},}
and it has expected value $\mu =np$. The variance is equal to
 $$
Var ( X ) = ( ∑ k = 0 n k 2 ( n k ) p k ( 1 − p ) n − k ) − μ 2 = n p ( 1 − p ) . {\displaystyle \operatorname {Var} (X)=\left(\sum _{k=0}^{n}k^{2}{n \choose k}p^{k}(1p)^{nk}\right)\mu ^{2}=np(1p).}
As a simple example, the binomial distribution with $p=1/2$ describes the probability of getting $k$ heads in $n$ tosses. Thus the expected value of the number of heads is ${\frac {n}{2}}$, and the variance is ${\frac {n}{4}}$.
Fair die
A sixsided fair die can be modelled with a discrete random variable with outcomes 1 through 6, each with equal probability ${\textstyle {\frac {1}{6}}}$. The expected value is ${\textstyle {\frac {1+2+3+4+5+6}{6}}=3.5}$. Therefore, the variance can be computed to be
 $$
∑ i = 1 6 1 6 ( i − 3.5 ) 2 = 1 6 ∑ i = 1 6 ( i − 3.5 ) 2 = 1 6 ( ( − 2.5 ) 2 + ( − 1.5 ) 2 + ( − 0.5 ) 2 + 0.5 2 + 1.5 2 + 2.5 2 ) = 1 6 ⋅ 17.50 = 35 12 ≈ 2.92. {\displaystyle {\begin{aligned}\sum _{i=1}^{6}{\tfrac {1}{6}}(i3.5)^{2}={\tfrac {1}{6}}\sum _{i=1}^{6}(i3.5)^{2}&={\tfrac {1}{6}}\left((2.5)^{2}{+}(1.5)^{2}{+}(0.5)^{2}{+}0.5^{2}{+}1.5^{2}{+}2.5^{2}\right)\\&={\tfrac {1}{6}}\cdot 17.50={\tfrac {35}{12}}\approx 2.92.\end{aligned}}}
The general formula for the variance of the outcome X of a die of n sides is
 $$
σ 2 = E ( X 2 ) − ( E ( X ) ) 2 = 1 n ∑ i = 1 n i 2 − ( 1 n ∑ i = 1 n i ) 2 = 1 6 ( n + 1 ) ( 2 n + 1 ) − 1 4 ( n + 1 ) 2 = n 2 − 1 12 . {\displaystyle {\begin{aligned}\sigma ^{2}=E(X^{2})(E(X))^{2}&={\frac {1}{n}}\sum _{i=1}^{n}i^{2}\left({\frac {1}{n}}\sum _{i=1}^{n}i\right)^{2}\\&={\tfrac {1}{6}}(n+1)(2n+1){\tfrac {1}{4}}(n+1)^{2}\\&={\frac {n^{2}1}{12}}.\end{aligned}}}
Properties
Basic properties
Variance is nonnegative because the squares are positive or zero:
 $$
Var ( X ) ≥ 0. {\displaystyle \operatorname {Var} (X)\geq 0.}
The variance of a constant random variable is zero, and if the variance of a variable in a data set is 0, then all the entries have the same value:
 $$
P ( X = a ) = 1 ⇔ Var ( X ) = 0. {\displaystyle P(X=a)=1\Leftrightarrow \operatorname {Var} (X)=0.}
Variance is invariant with respect to changes in a location parameter. That is, if a constant is added to all values of the variable, the variance is unchanged:
 $$
Var ( X + a ) = Var ( X ) . {\displaystyle \operatorname {Var} (X+a)=\operatorname {Var} (X).}
If all values are scaled by a constant, the variance is scaled by the square of that constant:
 $$
Var ( a X ) = a 2 Var ( X ) . {\displaystyle \operatorname {Var} (aX)=a^{2}\operatorname {Var} (X).}
The variance of a sum of two random variables is given by
 $$
Var ( a X + b Y ) = a 2 Var ( X ) + b 2 Var ( Y ) + 2 a b Cov ( X , Y ) , {\displaystyle \operatorname {Var} (aX+bY)=a^{2}\operatorname {Var} (X)+b^{2}\operatorname {Var} (Y)+2ab\,\operatorname {Cov} (X,Y),}
 $$
Var ( a X − b Y ) = a 2 Var ( X ) + b 2 Var ( Y ) − 2 a b Cov ( X , Y ) , {\displaystyle \operatorname {Var} (aXbY)=a^{2}\operatorname {Var} (X)+b^{2}\operatorname {Var} (Y)2ab\,\operatorname {Cov} (X,Y),}
where Cov(⋅, ⋅) is the covariance. In general we have for the sum of $N$ random variables $\{X_{1},\dots ,X_{N}\}$:
 $$
Var ( ∑ i = 1 N X i ) = ∑ i , j = 1 N Cov ( X i , X j ) = ∑ i = 1 N Var ( X i ) + ∑ i ≠ j Cov ( X i , X j ) . {\displaystyle \operatorname {Var} \left(\sum _{i=1}^{N}X_{i}\right)=\sum _{i,j=1}^{N}\operatorname {Cov} (X_{i},X_{j})=\sum _{i=1}^{N}\operatorname {Var} (X_{i})+\sum _{i\neq j}\operatorname {Cov} (X_{i},X_{j}).}
These results lead to the variance of a linear combination as:
 $$
Var ( ∑ i = 1 N a i X i ) = ∑ i , j = 1 N a i a j Cov ( X i , X j ) = ∑ i = 1 N a i 2 Var ( X i ) + ∑ i ≠ j a i a j Cov ( X i , X j ) = ∑ i = 1 N a i 2 Var ( X i ) + 2 ∑ 1 ≤ i < j ≤ N a i a j Cov ( X i , X j ) . {\displaystyle {\begin{aligned}\operatorname {Var} \left(\sum _{i=1}^{N}a_{i}X_{i}\right)&=\sum _{i,j=1}^{N}a_{i}a_{j}\operatorname {Cov} (X_{i},X_{j})\\&=\sum _{i=1}^{N}a_{i}^{2}\operatorname {Var} (X_{i})+\sum _{i\not =j}a_{i}a_{j}\operatorname {Cov} (X_{i},X_{j})\\&=\sum _{i=1}^{N}a_{i}^{2}\operatorname {Var} (X_{i})+2\sum _{1\leq i<j\leq N}a_{i}a_{j}\operatorname {Cov} (X_{i},X_{j}).\end{aligned}}}
If the random variables $X_{1},\dots ,X_{N}$ are such that
 $$
Cov ( X i , X j ) = 0 , ∀ ( i ≠ j ) , {\displaystyle \operatorname {Cov} (X_{i},X_{j})=0\ ,\ \forall \ (i\neq j),}
they are said to be uncorrelated. It follows immediately from the expression given earlier that if the random variables $X_{1},\dots ,X_{N}$ are uncorrelated, then the variance of their sum is equal to the sum of their variances, or, expressed symbolically:
 $$
Var ( ∑ i = 1 N X i ) = ∑ i = 1 N Var ( X i ) . {\displaystyle \operatorname {Var} \left(\sum _{i=1}^{N}X_{i}\right)=\sum _{i=1}^{N}\operatorname {Var} (X_{i}).}
Since independent random variables are always uncorrelated, the equation above holds in particular when the random variables $X_{1},\dots ,X_{n}$ are independent. Thus independence is sufficient but not necessary for the variance of the sum to equal the sum of the variances.
Sum of uncorrelated variables (Bienaymé formula)
One reason for the use of the variance in preference to other measures of dispersion is that the variance of the sum (or the difference) of uncorrelated random variables is the sum of their variances:
 $$
Var ( ∑ i = 1 n X i ) = ∑ i = 1 n Var ( X i ) . {\displaystyle \operatorname {Var} \left(\sum _{i=1}^{n}X_{i}\right)=\sum _{i=1}^{n}\operatorname {Var} (X_{i}).}
This statement is called the Bienaymé formula^{[2]} and was discovered in 1853.^{[3]}^{[4]} It is often made with the stronger condition that the variables are independent, but being uncorrelated suffices. So if all the variables have the same variance σ^{2}, then, since division by n is a linear transformation, this formula immediately implies that the variance of their mean is
 $$
Var ( X ¯ ) = Var ( 1 n ∑ i = 1 n X i ) = 1 n 2 ∑ i = 1 n Var ( X i ) = σ 2 n . {\displaystyle \operatorname {Var} \left({\overline {X}}\right)=\operatorname {Var} \left({\frac {1}{n}}\sum _{i=1}^{n}X_{i}\right)={\frac {1}{n^{2}}}\sum _{i=1}^{n}\operatorname {Var} \left(X_{i}\right)={\frac {\sigma ^{2}}{n}}.}
That is, the variance of the mean decreases when n increases. This formula for the variance of the mean is used in the definition of the standard error of the sample mean, which is used in the central limit theorem.
To prove the initial statement, it suffices to show that
 $$
Var ( X + Y ) = Var ( X ) + Var ( Y ) . {\displaystyle \operatorname {Var} (X+Y)=\operatorname {Var} (X)+\operatorname {Var} (Y).}
The general result then follows by induction. Starting with the definition,
 $$
Var ( X + Y ) = E [ ( X + Y ) 2 ] − ( E [ X + Y ] ) 2 = E [ X 2 + 2 X Y + Y 2 ] − ( E [ X ] + E [ Y ] ) 2 . {\displaystyle {\begin{aligned}\operatorname {Var} (X+Y)&=E[(X+Y)^{2}](E[X+Y])^{2}\\&=E[X^{2}+2XY+Y^{2}](E[X]+E[Y])^{2}.\\\end{aligned}}}
Using the linearity of the expectation operator and the assumption of independence (or uncorrelatedness) of X and Y, this further simplifies as follows:
 $$
Var ( X + Y ) = E [ X 2 ] + 2 E [ X Y ] + E [ Y 2 ] − ( E [ X ] 2 + 2 E [ X ] E [ Y ] + E [ Y ] 2 ) = E [ X 2 ] + E [ Y 2 ] − E [ X ] 2 − E [ Y ] 2 = Var ( X ) + Var ( Y ) . {\displaystyle {\begin{aligned}\operatorname {Var} (X+Y)&=E[X^{2}]+2E[XY]+E[Y^{2}](E[X]^{2}+2E[X]E[Y]+E[Y]^{2})\\&=E[X^{2}]+E[Y^{2}]E[X]^{2}E[Y]^{2}\\&=\operatorname {Var} (X)+\operatorname {Var} (Y).\end{aligned}}}
Sum of correlated variables
In general, if the variables are correlated, then the variance of their sum is the sum of their covariances:
 $$
Var ( ∑ i = 1 n X i ) = ∑ i = 1 n ∑ j = 1 n Cov ( X i , X j ) = ∑ i = 1 n Var ( X i ) + 2 ∑ 1 ≤ i < j ≤ n Cov ( X i , X j ) . {\displaystyle \operatorname {Var} \left(\sum _{i=1}^{n}X_{i}\right)=\sum _{i=1}^{n}\sum _{j=1}^{n}\operatorname {Cov} (X_{i},X_{j})=\sum _{i=1}^{n}\operatorname {Var} (X_{i})+2\sum _{1\leq i<j\leq n}\operatorname {Cov} (X_{i},X_{j}).}
(Note: The second equality comes from the fact that Cov(X_{i},X_{i}) = Var(X_{i}).)
Here Cov(⋅, ⋅) is the covariance, which is zero for independent random variables (if it exists). The formula states that the variance of a sum is equal to the sum of all elements in the covariance matrix of the components. The next expression states equivalently that the variance of the sum is the sum of the diagonal of covariance matrix plus two times the sum of its upper triangular elements (or its lower triangular elements); this emphasizes that the covariance matrix is symmetric. This formula is used in the theory of Cronbach's alpha in classical test theory.
So if the variables have equal variance σ^{2} and the average correlation of distinct variables is ρ, then the variance of their mean is
 $$
Var ( X ¯ ) = σ 2 n + n − 1 n ρ σ 2 . {\displaystyle \operatorname {Var} ({\overline {X}})={\frac {\sigma ^{2}}{n}}+{\frac {n1}{n}}\rho \sigma ^{2}.}
This implies that the variance of the mean increases with the average of the correlations. In other words, additional correlated observations are not as effective as additional independent observations at reducing the uncertainty of the mean. Moreover, if the variables have unit variance, for example if they are standardized, then this simplifies to
 $$
Var ( X ¯ ) = 1 n + n − 1 n ρ . {\displaystyle \operatorname {Var} ({\overline {X}})={\frac {1}{n}}+{\frac {n1}{n}}\rho .}
This formula is used in the Spearman–Brown prediction formula of classical test theory. This converges to ρ if n goes to infinity, provided that the average correlation remains constant or converges too. So for the variance of the mean of standardized variables with equal correlations or converging average correlation we have
 $$
lim n → ∞ Var ( X ¯ ) = ρ . {\displaystyle \lim _{n\to \infty }\operatorname {Var} ({\overline {X}})=\rho .}
Therefore, the variance of the mean of a large number of standardized variables is approximately equal to their average correlation. This makes clear that the sample mean of correlated variables does not generally converge to the population mean, even though the law of large numbers states that the sample mean will converge for independent variables.
Matrix notation for the variance of a linear combination
Define $X$ as a column vector of $n$ random variables $X_{1},\ldots ,X_{n}$, and $c$ as a column vector of $n$ scalars $c_{1},\ldots ,c_{n}$. Therefore, $c^{T}X$ is a linear combination of these random variables, where $c^{T}$ denotes the transpose of $c$. Also let $\Sigma$ be the covariance matrix of $X$. The variance of $c^{T}X$ is then given by:^{[5]}
 $$
Var ( c T X ) = c T Σ c . {\displaystyle \operatorname {Var} (c^{T}X)=c^{T}\Sigma c.}
Weighted sum of variables
The scaling property and the Bienaymé formula, along with the property of the covariance Cov(aX, bY) = ab Cov(X, Y) jointly imply that
 $$
Var ( a X ± b Y ) = a 2 Var ( X ) + b 2 Var ( Y ) ± 2 a b Cov ( X , Y ) . {\displaystyle \operatorname {Var} (aX\pm bY)=a^{2}\operatorname {Var} (X)+b^{2}\operatorname {Var} (Y)\pm 2ab\,\operatorname {Cov} (X,Y).}
This implies that in a weighted sum of variables, the variable with the largest weight will have a disproportionally large weight in the variance of the total. For example, if X and Y are uncorrelated and the weight of X is two times the weight of Y, then the weight of the variance of X will be four times the weight of the variance of Y.
The expression above can be extended to a weighted sum of multiple variables:
 $$
Var ( ∑ i n a i X i ) = ∑ i = 1 n a i 2 Var ( X i ) + 2 ∑ 1 ≤ i ∑ < j ≤ n a i a j Cov ( X i , X j ) {\displaystyle \operatorname {Var} \left(\sum _{i}^{n}a_{i}X_{i}\right)=\sum _{i=1}^{n}a_{i}^{2}\operatorname {Var} (X_{i})+2\sum _{1\leq i}\sum _{<j\leq n}a_{i}a_{j}\operatorname {Cov} (X_{i},X_{j})}
Product of independent variables
If two variables X and Y are independent, the variance of their product is given by^{[6]}
 $$
Var ( X Y ) = [ E ( X ) ] 2 Var ( Y ) + [ E ( Y ) ] 2 Var ( X ) + Var ( X ) Var ( Y ) . {\displaystyle {\begin{aligned}\operatorname {Var} (XY)&=[E(X)]^{2}\operatorname {Var} (Y)+[E(Y)]^{2}\operatorname {Var} (X)+\operatorname {Var} (X)\operatorname {Var} (Y).\end{aligned}}}
Equivalently, using the basic properties of expectation, it is given by
 $$
Var ( X Y ) = E ( X 2 ) E ( Y 2 ) − [ E ( X ) ] 2 [ E ( Y ) ] 2 . {\displaystyle \operatorname {Var} (XY)=E(X^{2})E(Y^{2})[E(X)]^{2}[E(Y)]^{2}.}
Product of statistically dependent variables
In general, if two variables are statistically dependent, the variance of their product is given by:
 $$
Var ( X Y ) = E [ X 2 Y 2 ] − [ E ( X Y ) ] 2 = Cov ( X 2 , Y 2 ) + E ( X 2 ) E ( Y 2 ) − [ E ( X Y ) ] 2 = Cov ( X 2 , Y 2 ) + ( Var ( X ) + [ E ( X ) ] 2 ) ( Var ( Y ) + [ E ( Y ) ] 2 ) − [ Cov ( X , Y ) + E ( X ) E ( Y ) ] 2 {\displaystyle {\begin{aligned}\operatorname {Var} (XY)&=E[X^{2}Y^{2}][E(XY)]^{2}\\&=\operatorname {Cov} (X^{2},Y^{2})+E(X^{2})E(Y^{2})[E(XY)]^{2}\\&=\operatorname {Cov} (X^{2},Y^{2})+(\operatorname {Var} (X)+[E(X)]^{2})(\operatorname {Var} (Y)+[E(Y)]^{2})[\operatorname {Cov} (X,Y)+E(X)E(Y)]^{2}\end{aligned}}}
Decomposition
The general formula for variance decomposition or the law of total variance is: If $X$ and $Y$ are two random variables, and the variance of $X$ exists, then
 $$
Var [ X ] = E Y ( Var [ X ∣ Y ] ) + Var Y ( E [ X ∣ Y ] ) . {\displaystyle \operatorname {Var} [X]=\operatorname {E} _{Y}(\operatorname {Var} [X\mid Y])+\operatorname {Var} _{Y}(\operatorname {E} [X\mid Y]).\,}
where $\operatorname {E} (XY)$ is the conditional expectation of $X$ given $Y$, and $\operatorname {Var} (X\mid Y)$ is the conditional variance of $X$ given $Y$. (A more intuitive explanation is that given a particular value of $Y$, then $X$ follows a distribution with mean $\operatorname {E} (X\mid Y)$ and variance $\operatorname {Var} (X\mid Y)$). As $\operatorname {E} (X\mid Y)$ is a function of the variable $Y$, the outer expectation or variance is taken with respect to Y. The above formula tells how to find $\operatorname {Var} (X)$ based on the distributions of these two quantities when $Y$ is allowed to vary.
In particular, if $Y$ is a discrete random variable assuming $y_{1},y_{2},\ldots ,y_{n}$ with corresponding probability masses $p_{1},p_{2},\ldots ,p_{n}$, then in the formula for total variance, the first term on the righthand side becomes
 $$
E Y ( Var [ X ∣ Y ] ) = ∑ i = 1 n p i σ i 2 , {\displaystyle \operatorname {E} _{Y}(\operatorname {Var} [X\mid Y])=\sum _{i=1}^{n}p_{i}\sigma _{i}^{2},}
where $\sigma _{i}^{2}=\operatorname {Var} [X\mid y_{i}]$. Similarly, the second term on the righthand side becomes
 $$
Var Y ( E [ X ∣ Y ] ) = ∑ i = 1 n p i μ i 2 − ( ∑ i = 1 n p i μ i ) 2 = ∑ i = 1 n p i μ i 2 − μ 2 , {\displaystyle \operatorname {Var} _{Y}(\operatorname {E} [X\mid Y])=\sum _{i=1}^{n}p_{i}\mu _{i}^{2}\left(\sum _{i=1}^{n}p_{i}\mu _{i}\right)^{2}=\sum _{i=1}^{n}p_{i}\mu _{i}^{2}\mu ^{2},}
where $\mu _{i}=\operatorname {E} [X\mid y_{i}]$ and $\mu =\sum _{i=1}^{n}p_{i}\mu _{i}$. Thus the total variance is given by
 $$
Var [ X ] = ∑ i = 1 n p i σ i 2 + ( ∑ i = 1 n p i μ i 2 − μ 2 ) . {\displaystyle \operatorname {Var} [X]=\sum _{i=1}^{n}p_{i}\sigma _{i}^{2}+\left(\sum _{i=1}^{n}p_{i}\mu _{i}^{2}\mu ^{2}\right).}
A similar formula is applied in analysis of variance, where the corresponding formula is
 $$
M S total = M S between + M S within ; {\displaystyle {\mathit {MS}}_{\text{total}}={\mathit {MS}}_{\text{between}}+{\mathit {MS}}_{\text{within}};}
here ${\mathit {MS}}$ refers to the Mean of the Squares. In linear regression analysis the corresponding formula is
 $$
M S total = M S regression + M S residual . {\displaystyle {\mathit {MS}}_{\text{total}}={\mathit {MS}}_{\text{regression}}+{\mathit {MS}}_{\text{residual}}.}
This can also be derived from the additivity of variances, since the total (observed) score is the sum of the predicted score and the error score, where the latter two are uncorrelated.
Similar decompositions are possible for the sum of squared deviations (sum of squares, ${\mathit {SS}}$):
 $$
S S total = S S between + S S within , {\displaystyle {\mathit {SS}}_{\text{total}}={\mathit {SS}}_{\text{between}}+{\mathit {SS}}_{\text{within}},}  $$
S S total = S S regression + S S residual . {\displaystyle {\mathit {SS}}_{\text{total}}={\mathit {SS}}_{\text{regression}}+{\mathit {SS}}_{\text{residual}}.}
Formulae for the variance
A formula often used for deriving the variance of a theoretical distribution is as follows:
 $$
Var ( X ) = E ( X 2 ) − ( E ( X ) ) 2 . {\displaystyle \operatorname {Var} (X)=\operatorname {E} (X^{2})(\operatorname {E} (X))^{2}.}
This will be useful when it is possible to derive formulae for the expected value and for the expected value of the square.
This formula is also sometimes used in connection with the sample variance. While useful for hand calculations, it is not advised for computer calculations as it suffers from catastrophic cancellation if the two components of the equation are similar in magnitude and floating point arithmetic is used. This is discussed in the article Algorithms for calculating variance.
Calculation from the CDF
The population variance for a nonnegative random variable can be expressed in terms of the cumulative distribution function F using
 $$
2 ∫ 0 ∞ u ( 1 − F ( u ) ) d u − ( ∫ 0 ∞ 1 − F ( u ) d u ) 2 . {\displaystyle 2\int _{0}^{\infty }u(1F(u))\,du{\Big (}\int _{0}^{\infty }1F(u)\,du{\Big )}^{2}.}
This expression can be used to calculate the variance in situations where the CDF, but not the density, can be conveniently expressed.
Characteristic property
The second moment of a random variable attains the minimum value when taken around the first moment (i.e., mean) of the random variable, i.e. $\mathrm {argmin} _{m}\,\mathrm {E} ((Xm)^{2})=\mathrm {E} (X)\,$. Conversely, if a continuous function $\varphi$ satisfies $\mathrm {argmin} _{m}\,\mathrm {E} (\varphi (Xm))=\mathrm {E} (X)\,$ for all random variables X, then it is necessarily of the form $\varphi (x)=ax^{2}+b$, where a > 0. This also holds in the multidimensional case.^{[7]}
Units of measurement
Unlike expected absolute deviation, the variance of a variable has units that are the square of the units of the variable itself. For example, a variable measured in meters will have a variance measured in meters squared. For this reason, describing data sets via their standard deviation or root mean square deviation is often preferred over using the variance. In the dice example the standard deviation is √2.9 ≈ 1.7, slightly larger than the expected absolute deviation of 1.5.
The standard deviation and the expected absolute deviation can both be used as an indicator of the "spread" of a distribution. The standard deviation is more amenable to algebraic manipulation than the expected absolute deviation, and, together with variance and its generalization covariance, is used frequently in theoretical statistics; however the expected absolute deviation tends to be more robust as it is less sensitive to outliers arising from measurement anomalies or an unduly heavytailed distribution.
Approximating the variance of a function
The delta method uses secondorder Taylor expansions to approximate the variance of a function of one or more random variables: see Taylor expansions for the moments of functions of random variables. For example, the approximate variance of a function of one variable is given by

 $$
Var [ f ( X ) ] ≈ ( f ′ ( E [ X ] ) ) 2 Var [ X ] {\displaystyle \operatorname {Var} \left[f(X)\right]\approx \left(f'(\operatorname {E} \left[X\right])\right)^{2}\operatorname {Var} \left[X\right]}
 $$
provided that f is twice differentiable and that the mean and variance of X are finite.
Population variance and sample variance
Realworld observations such as the measurements of yesterday's rain throughout the day typically cannot be complete sets of all possible observations that could be made. As such, the variance calculated from the finite set will in general not match the variance that would have been calculated from the full population of possible observations. This means that one estimates the mean and variance that would have been calculated from an omniscient set of observations by using an estimator equation. The estimator is a function of the sample of n observations drawn without observational bias from the whole population of potential observations. In this example that sample would be the set of actual measurements of yesterday's rainfall from available rain gauges within the geography of interest.
The simplest estimators for population mean and population variance are simply the mean and variance of the sample, the sample mean and (uncorrected) sample variance – these are consistent estimators (they converge to the correct value as the number of samples increases), but can be improved. Estimating the population variance by taking the sample's variance is close to optimal in general, but can be improved in two ways. Most simply, the sample variance is computed as an average of squared deviations about the (sample) mean, by dividing by n. However, using values other than n improves the estimator in various ways. Four common values for the denominator are n, n − 1, n + 1, and n − 1.5: n is the simplest (population variance of the sample), n − 1 eliminates bias, n + 1 minimizes mean squared error for the normal distribution, and n − 1.5 mostly eliminates bias in unbiased estimation of standard deviation for the normal distribution.
Firstly, if the omniscient mean is unknown (and is computed as the sample mean), then the sample variance is a biased estimator: it underestimates the variance by a factor of (n − 1) / n; correcting by this factor (dividing by n − 1 instead of n) is called Bessel's correction. The resulting estimator is unbiased, and is called the (corrected) sample variance or unbiased sample variance. For example, when n = 1 the variance of a single observation about the sample mean (itself) is obviously zero regardless of the population variance. If the mean is determined in some other way than from the same samples used to estimate the variance then this bias does not arise and the variance can safely be estimated as that of the samples about the (independently known) mean.
Secondly, the sample variance does not generally minimize mean squared error between sample variance and population variance. Correcting for bias often makes this worse: one can always choose a scale factor that performs better than the corrected sample variance, though the optimal scale factor depends on the excess kurtosis of the population (see mean squared error: variance), and introduces bias. This always consists of scaling down the unbiased estimator (dividing by a number larger than n − 1), and is a simple example of a shrinkage estimator: one "shrinks" the unbiased estimator towards zero. For the normal distribution, dividing by n + 1 (instead of n − 1 or n) minimizes mean squared error. The resulting estimator is biased, however, and is known as the biased sample variation.
Population variance
In general, the population variance of a finite population of size N with values x_{i} is given by
 $$
σ 2 = 1 N ∑ i = 1 N ( x i − μ ) 2 = 1 N ∑ i = 1 N ( x i 2 − 2 μ x i + μ 2 ) = ( 1 N ∑ i = 1 N x i 2 ) − 2 μ ( 1 N ∑ i = 1 N x i ) + μ 2 = ( 1 N ∑ i = 1 N x i 2 ) − μ 2 {\displaystyle \sigma ^{2}={\frac {1}{N}}\sum _{i=1}^{N}\left(x_{i}\mu \right)^{2}={\frac {1}{N}}\sum _{i=1}^{N}\left(x_{i}^{2}2\mu x_{i}+\mu ^{2}\right)=\left({\frac {1}{N}}\sum _{i=1}^{N}x_{i}^{2}\right)2\mu \left({\frac {1}{N}}\sum _{i=1}^{N}x_{i}\right)+\mu ^{2}=\left({\frac {1}{N}}\sum _{i=1}^{N}x_{i}^{2}\right)\mu ^{2}}
where the population mean is
 $$
μ = 1 N ∑ i = 1 N x i {\displaystyle \mu ={\frac {1}{N}}\sum _{i=1}^{N}x_{i}} .
The population variance can also be computed using
 $$
σ 2 = 1 N 2 ∑ i < j ( x i − x j ) 2 = 1 2 N 2 ∑ i , j = 1 N ( x i − x j ) 2 {\displaystyle \sigma ^{2}={\frac {1}{N^{2}}}\sum _{i<j}\left(x_{i}x_{j}\right)^{2}={\frac {1}{2N^{2}}}\sum _{i,j=1}^{N}\left(x_{i}x_{j}\right)^{2}}
This is true because
 $$
1 2 N 2 ∑ i , j = 1 N ( x i − x j ) 2 = 1 2 N 2 ∑ i , j = 1 N ( x i 2 − 2 x i x j + x j 2 ) = 1 2 N ∑ j = 1 N ( 1 N ∑ i = 1 N x i 2 ) − ( 1 N ∑ i = 1 N x i ) ( 1 N ∑ j = 1 N x j ) + 1 2 N ∑ i = 1 N ( 1 N ∑ j = 1 N x j 2 ) = 1 2 ( σ 2 + μ 2 ) − μ 2 + 1 2 ( σ 2 + μ 2 ) = σ 2 {\displaystyle {\begin{aligned}{\frac {1}{2N^{2}}}\sum _{i,j=1}^{N}\left(x_{i}x_{j}\right)^{2}&={\frac {1}{2N^{2}}}\sum _{i,j=1}^{N}\left(x_{i}^{2}2x_{i}x_{j}+x_{j}^{2}\right)\\&={\frac {1}{2N}}\sum _{j=1}^{N}\left({\frac {1}{N}}\sum _{i=1}^{N}x_{i}^{2}\right)\left({\frac {1}{N}}\sum _{i=1}^{N}x_{i}\right)\left({\frac {1}{N}}\sum _{j=1}^{N}x_{j}\right)+{\frac {1}{2N}}\sum _{i=1}^{N}\left({\frac {1}{N}}\sum _{j=1}^{N}x_{j}^{2}\right)\\&={\frac {1}{2}}\left(\sigma ^{2}+\mu ^{2}\right)\mu ^{2}+{\frac {1}{2}}\left(\sigma ^{2}+\mu ^{2}\right)=\sigma ^{2}\end{aligned}}}
The population variance matches the variance of the generating probability distribution. In this sense, the concept of population can be extended to continuous random variables with infinite populations.
Sample variance
In many practical situations, the true variance of a population is not known a priori and must be computed somehow. When dealing with extremely large populations, it is not possible to count every object in the population, so the computation must be performed on a sample of the population.^{[8]} Sample variance can also be applied to the estimation of the variance of a continuous distribution from a sample of that distribution.
We take a sample with replacement of n values y_{1}, ..., y_{n} from the population, where n < N, and estimate the variance on the basis of this sample.^{[9]} Directly taking the variance of the sample data gives the average of the squared deviations:
 $$
σ y 2 = 1 n ∑ i = 1 n ( y i − y ¯ ) 2 = ( 1 n ∑ i = 1 n y i 2 ) − y ¯ 2 = 1 n 2 ∑ i < j ( y i − y j ) 2 . {\displaystyle \sigma _{y}^{2}={\frac {1}{n}}\sum _{i=1}^{n}\left(y_{i}{\overline {y}}\right)^{2}=\left({\frac {1}{n}}\sum _{i=1}^{n}y_{i}^{2}\right){\overline {y}}^{2}={\frac {1}{n^{2}}}\sum _{i<j}\left(y_{i}y_{j}\right)^{2}.}
Here, ${\overline {y}}$ denotes the sample mean:
 $$
y ¯ = 1 n ∑ i = 1 n y i . {\displaystyle {\overline {y}}={\frac {1}{n}}\sum _{i=1}^{n}y_{i}.}
Since the y_{i} are selected randomly, both $\scriptstyle {\overline {y}}$ and $\scriptstyle \sigma _{y}^{2}$ are random variables. Their expected values can be evaluated by averaging over the ensemble of all possible samples {y_{i}} of size n from the population. For $\scriptstyle \sigma _{y}^{2}$ this gives:
 $$
E [ σ y 2 ] = E [ 1 n ∑ i = 1 n ( y i − 1 n ∑ j = 1 n y j ) 2 ] = 1 n ∑ i = 1 n E [ y i 2 − 2 n y i ∑ j = 1 n y j + 1 n 2 ∑ j = 1 n y j ∑ k = 1 n y k ] = 1 n ∑ i = 1 n [ n − 2 n E [ y i 2 ] − 2 n ∑ j ≠ i E [ y i y j ] + 1 n 2 ∑ j = 1 n ∑ k ≠ j n E [ y j y k ] + 1 n 2 ∑ j = 1 n E [ y j 2 ] ] = 1 n ∑ i = 1 n [ n − 2 n ( σ 2 + μ 2 ) − 2 n ( n − 1 ) μ 2 + 1 n 2 n ( n − 1 ) μ 2 + 1 n ( σ 2 + μ 2 ) ] = n − 1 n σ 2 . {\displaystyle {\begin{aligned}E[\sigma _{y}^{2}]&=E\left[{\frac {1}{n}}\sum _{i=1}^{n}\left(y_{i}{\frac {1}{n}}\sum _{j=1}^{n}y_{j}\right)^{2}\right]\\&={\frac {1}{n}}\sum _{i=1}^{n}E\left[y_{i}^{2}{\frac {2}{n}}y_{i}\sum _{j=1}^{n}y_{j}+{\frac {1}{n^{2}}}\sum _{j=1}^{n}y_{j}\sum _{k=1}^{n}y_{k}\right]\\&={\frac {1}{n}}\sum _{i=1}^{n}\left[{\frac {n2}{n}}E[y_{i}^{2}]{\frac {2}{n}}\sum _{j\neq i}E[y_{i}y_{j}]+{\frac {1}{n^{2}}}\sum _{j=1}^{n}\sum _{k\neq j}^{n}E[y_{j}y_{k}]+{\frac {1}{n^{2}}}\sum _{j=1}^{n}E[y_{j}^{2}]\right]\\&={\frac {1}{n}}\sum _{i=1}^{n}\left[{\frac {n2}{n}}(\sigma ^{2}+\mu ^{2}){\frac {2}{n}}(n1)\mu ^{2}+{\frac {1}{n^{2}}}n(n1)\mu ^{2}+{\frac {1}{n}}(\sigma ^{2}+\mu ^{2})\right]\\&={\frac {n1}{n}}\sigma ^{2}.\end{aligned}}}
Hence $\scriptstyle \sigma _{y}^{2}$ gives an estimate of the population variance that is biased by a factor of ${\frac {n1}{n}}$. For this reason, $\scriptstyle \sigma _{y}^{2}$ is referred to as the biased sample variance. Correcting for this bias yields the unbiased sample variance:
 $$
s 2 = n n − 1 σ y 2 = n n − 1 ( 1 n ∑ i = 1 n ( y i − y ¯ ) 2 ) = 1 n − 1 ∑ i = 1 n ( y i − y ¯ ) 2 {\displaystyle s^{2}={\frac {n}{n1}}\sigma _{y}^{2}={\frac {n}{n1}}\left({\frac {1}{n}}\sum _{i=1}^{n}\left(y_{i}{\overline {y}}\right)^{2}\right)={\frac {1}{n1}}\sum _{i=1}^{n}\left(y_{i}{\overline {y}}\right)^{2}}
Either estimator may be simply referred to as the sample variance when the version can be determined by context. The same proof is also applicable for samples taken from a continuous probability distribution.
The use of the term n − 1 is called Bessel's correction, and it is also used in sample covariance and the sample standard deviation (the square root of variance). The square root is a concave function and thus introduces negative bias (by Jensen's inequality), which depends on the distribution, and thus the corrected sample standard deviation (using Bessel's correction) is biased. The unbiased estimation of standard deviation is a technically involved problem, though for the normal distribution using the term n − 1.5 yields an almost unbiased estimator.
The unbiased sample variance is a Ustatistic for the function ƒ(y_{1}, y_{2}) = (y_{1} − y_{2})^{2}/2, meaning that it is obtained by averaging a 2sample statistic over 2element subsets of the population.
Distribution of the sample variance
Being a function of random variables, the sample variance is itself a random variable, and it is natural to study its distribution. In the case that y_{i} are independent observations from a normal distribution, Cochran's theorem shows that s^{2} follows a scaled chisquared distribution:^{[10]}
 $$
( n − 1 ) s 2 σ 2 ∼ χ n − 1 2 . {\displaystyle (n1){\frac {s^{2}}{\sigma ^{2}}}\sim \chi _{n1}^{2}.}
As a direct consequence, it follows that
 $$
E ( s 2 ) = E ( σ 2 n − 1 χ n − 1 2 ) = σ 2 , {\displaystyle \operatorname {E} (s^{2})=\operatorname {E} \left({\frac {\sigma ^{2}}{n1}}\chi _{n1}^{2}\right)=\sigma ^{2},}
and^{[11]}
 $$
Var [ s 2 ] = Var ( σ 2 n − 1 χ n − 1 2 ) = σ 4 ( n − 1 ) 2 Var ( χ n − 1 2 ) = 2 σ 4 n − 1 . {\displaystyle \operatorname {Var} [s^{2}]=\operatorname {Var} \left({\frac {\sigma ^{2}}{n1}}\chi _{n1}^{2}\right)={\frac {\sigma ^{4}}{(n1)^{2}}}\operatorname {Var} \left(\chi _{n1}^{2}\right)={\frac {2\sigma ^{4}}{n1}}.}
If the y_{i} are independent and identically distributed, but not necessarily normally distributed, then^{[12]}^{[13]}
 $$
E [ s 2 ] = σ 2 , Var [ s 2 ] = σ 4 n ( ( κ − 1 ) + 2 n − 1 ) = 1 n ( μ 4 − n − 3 n − 1 σ 4 ) , {\displaystyle \operatorname {E} [s^{2}]=\sigma ^{2},\quad \operatorname {Var} [s^{2}]={\frac {\sigma ^{4}}{n}}\left((\kappa 1)+{\frac {2}{n1}}\right)={\frac {1}{n}}\left(\mu _{4}{\frac {n3}{n1}}\sigma ^{4}\right),}
where κ is the kurtosis of the distribution and μ_{4} is the fourth central moment.
If the conditions of the law of large numbers hold for the squared observations, s^{2} is a consistent estimator of σ^{2}. One can see indeed that the variance of the estimator tends asymptotically to zero. An asymptotically equivalent formula was given in Kenney and Keeping (1951:164), Rose and Smith (2002:264), and Weisstein (n.d.).^{[14]}^{[15]}^{[16]}
Samuelson's inequality
Samuelson's inequality is a result that states bounds on the values that individual observations in a sample can take, given that the sample mean and (biased) variance have been calculated.^{[17]} Values must lie within the limits ${\bar {y}}\pm \sigma _{y}(n1)^{1/2}.$
Relations with the harmonic and arithmetic means
It has been shown^{[18]} that for a sample {y_{i}} of real numbers,
 $$
σ y 2 ≤ 2 y max ( A − H ) , {\displaystyle \sigma _{y}^{2}\leq 2y_{\max }(AH),}
where y_{max} is the maximum of the sample, A is the arithmetic mean, H is the harmonic mean of the sample and $\sigma _{y}^{2}$ is the (biased) variance of the sample.
This bound has been improved, and it is known that variance is bounded by
 $$
σ y 2 ≤ y max ( A − H ) ( y max − A ) y max − H , {\displaystyle \sigma _{y}^{2}\leq {\frac {y_{\max }(AH)(y_{\max }A)}{y_{\max }H}},}
 $$
σ y 2 ≥ y min ( A − H ) ( A − y min ) H − y min , {\displaystyle \sigma _{y}^{2}\geq {\frac {y_{\min }(AH)(Ay_{\min })}{Hy_{\min }}},}
where y_{min} is the minimum of the sample.^{[19]}
Tests of equality of variances
Testing for the equality of two or more variances is difficult. The F test and chi square tests are both adversely affected by nonnormality and are not recommended for this purpose.
Several non parametric tests have been proposed: these include the Barton–David–Ansari–Freund–Siegel–Tukey test, the Capon test, Mood test, the Klotz test and the Sukhatme test. The Sukhatme test applies to two variances and requires that both medians be known and equal to zero. The Mood, Klotz, Capon and Barton–David–Ansari–Freund–Siegel–Tukey tests also apply to two variances. They allow the median to be unknown but do require that the two medians are equal.
The Lehmann test is a parametric test of two variances. Of this test there are several variants known. Other tests of the equality of variances include the Box test, the Box–Anderson test and the Moses test.
Resampling methods, which include the bootstrap and the jackknife, may be used to test the equality of variances.
History
The term variance was first introduced by Ronald Fisher in his 1918 paper The Correlation Between Relatives on the Supposition of Mendelian Inheritance:^{[20]}
The great body of available statistics show us that the deviations of a human measurement from its mean follow very closely the Normal Law of Errors, and, therefore, that the variability may be uniformly measured by the standard deviation corresponding to the square root of the mean square error. When there are two independent causes of variability capable of producing in an otherwise uniform population distributions with standard deviations $\sigma _{1}$ and $\sigma _{2}$, it is found that the distribution, when both causes act together, has a standard deviation ${\sqrt {\sigma _{1}^{2}+\sigma _{2}^{2}}}$. It is therefore desirable in analysing the causes of variability to deal with the square of the standard deviation as the measure of variability. We shall term this quantity the Variance...
Moment of inertia
The variance of a probability distribution is analogous to the moment of inertia in classical mechanics of a corresponding mass distribution along a line, with respect to rotation about its center of mass.^{[citation needed]} It is because of this analogy that such things as the variance are called moments of probability distributions.^{[citation needed]} The covariance matrix is related to the moment of inertia tensor for multivariate distributions. The moment of inertia of a cloud of n points with a covariance matrix of $\Sigma$ is given by^{[citation needed]}
 $$
I = n ( 1 3 × 3 tr ( Σ ) − Σ ) . {\displaystyle I=n(\mathbf {1} _{3\times 3}\operatorname {tr} (\Sigma )\Sigma ).}
This difference between moment of inertia in physics and in statistics is clear for points that are gathered along a line. Suppose many points are close to the x axis and distributed along it. The covariance matrix might look like
 $$
Σ = [ 10 0 0 0 0.1 0 0 0 0.1 ] . {\displaystyle \Sigma ={\begin{bmatrix}10&0&0\ &0.1&0\ &0&0.1\end{bmatrix}}.}
That is, there is the most variance in the x direction. Physicists would consider this to have a low moment about the x axis so the momentofinertia tensor is
 $$
I = n [ 0.2 0 0 0 10.1 0 0 0 10.1 ] . {\displaystyle I=n{\begin{bmatrix}0.2&0&0\ &10.1&0\ &0&10.1\end{bmatrix}}.}
Semivariance
The semivariance is calculated in the same manner as the variance but only those observations that fall below the mean are included in the calculation. It is sometimes described as a measure of downside risk in an investments context. For skewed distributions, the semivariance can provide additional information that a variance does not.^{[citation needed]}
Generalizations
It has been suggested that this section be merged with Covariance#Complex random vectors to Complex variance and covariance. (Discuss) Proposed since June 2015. 
If $x$ is a scalar complexvalued random variable, with values in $\mathbb {C} ,$ then its variance is $\operatorname {E} \left[(x\mu )(x\mu )^{*}\right],$ where $x^{*}$ is the complex conjugate of $x.$ This variance is a real scalar.
If $X$ is a vectorvalued random variable, with values in $\mathbb {R} ^{n},$ and thought of as a column vector, then a natural generalization of variance is $\operatorname {E} \left[(X\mu )(X\mu )^{\operatorname {T} }\right],$ where $\mu =\operatorname {E} (X)$ and $X^{\operatorname {T} }$ is the transpose of $X,$ and so is a row vector. The result is a positive semidefinite square matrix, commonly referred to as the variancecovariance matrix (or simply as the covariance matrix).
Another natural generalization of variance for such vectorvalued random variables $X,$ which results in a scalar value rather than in a matrix, is obtained by interpreting the deviation between the random variable and its mean as the Euclidean distance. This results in $\operatorname {E} \left[(X\mu )^{\operatorname {T} }(X\mu )\right]=\operatorname {tr} (C),$ which is the trace of the covariance matrix.
If $X$ is a vector and complexvalued random variable, with values in $\mathbb {C} ^{n},$ then the covariance matrix is $\operatorname {E} \left[(X\mu )(X\mu )^{\dagger }\right],$ where $X^{\dagger }$ is the conjugate transpose of $X.$^{[citation needed]} This matrix is also positive semidefinite and square.
See also
This "see also" section may contain an excessive number of suggestions. Please ensure that only the most relevant links are given, that they are not red links, and that any links are not already in this article. (May 2017) (Learn how and when to remove this template message) 
 Statistics portal
Look up variance in Wiktionary, the free dictionary. 
 Algorithms for calculating variance
 Average absolute deviation
 Bhatia–Davis inequality
 Commonmethod variance
 Correlation
 Covariance
 Chebyshev's inequality
 Distance variance
 Estimation of covariance matrices
 Explained variance
 Homoscedasticity
 Mean absolute error
 Mean absolute difference
 Mean preserving spread
 Pooled variance (also known as combined, composite, or overall variance)
 Popoviciu's inequality on variances
 Qualitative variation
 Quasivariance, used in linear regression when the explanatory variable is categorical
 Reduced chisquared
 Sample mean and covariance
 Semivariance
 Skewness
 Taylor's law
 Weighted sample variance
Notes
 ^ Yuli Zhang,Huaiyu Wu,Lei Cheng (June 2012). Some new deformation formulas about variance and covariance. Proceedings of 4th International Conference on Modelling, Identification and Control(ICMIC2012). pp. 987–992. CS1 maint: Uses authors parameter (link)
 ^ Loeve, M. (1977) "Probability Theory", Graduate Texts in Mathematics, Volume 45, 4th edition, SpringerVerlag, p. 12.
 ^ Bienaymé, I.J. (1853) "Considérations à l'appui de la découverte de Laplace sur la loi de probabilité dans la méthode des moindres carrés", Comptes rendus de l'Académie des sciences Paris, 37, p. 309–317; digital copy available [1]
 ^ Bienaymé, I.J. (1867) "Considérations à l'appui de la découverte de Laplace sur la loi de probabilité dans la méthode des moindres carrés", Journal de Mathématiques Pures et Appliquées, Série 2, Tome 12, p. 158–167; digital copy available [2][3]
 ^ Johnson, Richard; Wichern, Dean (2001). Applied Multivariate Statistical Analysis. Prentice Hall. p. 76. ISBN 0131877151
 ^ Goodman, Leo A., "On the exact variance of products," Journal of the American Statistical Association, December 1960, 708–713. DOI: 10.2307/2281592
 ^ Kagan, A.; Shepp, L. A. (1998). "Why the variance?". Statistics & Probability Letters. 38 (4): 329–333. doi:10.1016/S01677152(98)000418.
 ^ Navidi, William (2006) Statistics for Engineers and Scientists, McGrawHill, pg 14.
 ^ Montgomery, D. C. and Runger, G. C. (1994) Applied statistics and probability for engineers, page 201. John Wiley & Sons New York
 ^ Knight K. (2000), Mathematical Statistics, Chapman and Hall, New York. (proposition 2.11)
 ^ Casella and Berger (2002) Statistical Inference, Example 7.3.3, p. 331^{[full citation needed]}
 ^ Cho, Eungchun; Cho, Moon Jung; Eltinge, John (2005) The Variance of Sample Variance From a Finite Population. International Journal of Pure and Applied Mathematics 21 (3): 387394. http://www.ijpam.eu/contents/2005213/10/10.pdf
 ^ Cho, Eungchun; Cho, Moon Jung (2009) Variance of Sample Variance With Replacement. International Journal of Pure and Applied Mathematics 52 (1): 4347. http://www.ijpam.eu/contents/2009521/5/5.pdf
 ^ Kenney, John F.; Keeping, E.S. (1951) Mathematics of Statistics. Part Two. 2nd ed. D. Van Nostrand Company, Inc. Princeton: New Jersey. http://krishikosh.egranth.ac.in/bitstream/1/2025521/1/G2257.pdf
 ^ Rose, Colin; Smith, Murray D. (2002) Mathematical Statistics with Mathematica. SpringerVerlag, New York. http://www.mathstatica.com/book/Mathematical_Statistics_with_Mathematica.pdf
 ^ Weisstein, Eric W. (n.d.) Sample Variance Distribution. MathWorldA Wolfram Web Resource. http://mathworld.wolfram.com/SampleVarianceDistribution.html
 ^ Samuelson, Paul (1968). "How Deviant Can You Be?". Journal of the American Statistical Association. 63 (324): 1522–1525. JSTOR 2285901. doi:10.1080/01621459.1968.10480944.
 ^ Mercer, A. McD. (2000). "Bounds for A–G, A–H, G–H, and a family of inequalities of Ky Fan’s type, using a general method". J. Math. Anal. Appl. 243 (1): 163–173. doi:10.1006/jmaa.1999.6688.
 ^ Sharma, R. (2008). "Some more inequalities for arithmetic mean, harmonic mean and variance". J. Math. Inequalities. 2 (1): 109–114. doi:10.7153/jmi0211.
 ^ Ronald Fisher (1918) The correlation between relatives on the supposition of Mendelian Inheritance
Theory of probability distributions








Statistics


















Authority control 


UpToDate Contents
全文を閲覧するには購読必要です。 To read the full text you will need to subscribe.
 1. QT間隔のばらつき：臨床応用 qt dispersion clinical applications
 2. QT間隔のばらつき：測定および解釈 qt dispersion measurement and interpretation
 3. 手首の解剖学および基本的な生体力学 anatomy and basic biomechanics of the wrist
 4. 末梢動脈疾患（PAD）の疫学、危険因子、および自然経過 epidemiology risk factors and natural history of peripheral artery disease
 5. 成人における間欠性爆発性障害：疫学、臨床的特徴、評価、および診断 intermittent explosive disorder in adults epidemiology clinical features assessment and diagnosis
英文文献
 Main differences between volatiles of sparkling and base wines accessed through comprehensive two dimensional gas chromatography with timeofflight mass spectrometric detection and chemometric tools.
 Welke JE1, Zanus M2, Lazzarotto M3, Pulgati FH4, Zini CA5.
 Food chemistry.Food Chem.2014 Dec 1;164:42737. doi: 10.1016/j.foodchem.2014.05.025. Epub 2014 May 17.
 The main changes in the volatile profile of base wines and their corresponding sparkling wines produced by traditional method were evaluated and investigated for the first time using headspace solidphase microextraction combined with comprehensive twodimensional gas chromatography with timeoffli
 PMID 24996354
 Genetic diversity of amylose content and RVA pasting parameters in 20 rice accessions grown in Hainan, China.
 Tong C1, Chen Y1, Tang F1, Xu F1, Huang Y1, Chen H2, Bao J3.
 Food chemistry.Food Chem.2014 Oct 15;161:23945. doi: 10.1016/j.foodchem.2014.04.011. Epub 2014 Apr 12.
 Starch physicochemical properties determine the eating and cooking quality of rice. The genetic diversity in the apparent amylose content (AAC) and pasting viscosity parameters of 20 geographically diverse rice accessions were investigated. It was found that AAC and pasting viscosities differed wide
 PMID 24837946
 Health care students' reactions towards HIV patients: examining prejudice, emotions, attribution of blame and willingness to interact with HIV/AIDS patients.
 Philip J1, Chadee D, Yearwood RP.
 AIDS care.AIDS Care.2014 Oct;26(10):123641. doi: 10.1080/09540121.2014.896449. Epub 2014 Mar 20.
 One of the most pervasive stigmatising conditions in society today is HIV/AIDS. In Trinidad and Tobago, stigma and discrimination are still pervasive especially against persons living with HIV/AIDS (PLHIV) and atrisk groups. HIV stigmatisation takes place at all levels including health care institu
 PMID 24646404
 A test of the resource security and the body mass index reference point hypotheses of body dissatisfaction amongst adolescents in eight countries.
 Mellor D1, FullerTyszkiewicz M, McCabe MP, Ricciardelli LA, Skouteris H, Mussap AJ.
 Ethnicity & health.Ethn Health.2014 Oct;19(5):54864. doi: 10.1080/13557858.2013.857761. Epub 2013 Nov 21.
 Objective. This study aimed to identify culturallevel variables that may influence the extent to which adolescents from different cultural groups are dissatisfied with their bodies. Design. A sample of 1730 male and 2000 female adolescents from Australia, Fiji, Malaysia, Tonga, Tongans in New Zeala
 PMID 24261816
和文文献
 対人過敏・自己優先尺度の作成――「新型うつ」の心理学的特徴の測定――
 心理学研究 87(6), 622632, 2018
 NAID 130005399530
 学習方略の使用に対する時期と使用法の有効性の認知
 心理学研究, 2018
 NAID 130005290737
 初年次教育としての体育科教育がもたらす 心理・社会的効果の研究 －準実験的研究手法にもとづく授業効果の検証－
 国際基督教大学学報. IA 教育研究 = Educational Studies 59, 1729, 20170331
 NAID 120006029789
 部活動における主将のリーダーシップが部員の集団効力感に及ぼす影響
 学習開発学研究 (10), 117124, 20170316
 NAID 120005998236
関連リンク
関連画像
■★リンクテーブル★
先読み  「fluctuation」 
リンク元  「variation」「disperse」「inconsistent」「vary」「intersperse」 
拡張検索  「invariance」「covariance」「analysis of covariance」 
「fluctuation」
 n.
 波動、不動。動揺、揺らぎ。高下、変動。不安定、気迷い。(pl.)盛衰、興亡
WordNet ［license wordnet］
「the quality of being unsteady and subject to changes; "he kept a record of price fluctuations"」 同
 wavering
WordNet ［license wordnet］
「a wave motion; "the fluctuations of the sea"」PrepTutorEJDIC ［license prepejdic］
「波動;動揺,変動」
「variation」
 関
 alter、alteration、change、difference、disparity、distinction、fluctuate、fluctuation、heteromorphic、heteromorphy、mutant、mutate、mutation、oscillate、shift、transmutation、turn、turn to、variance、variant、variational、vary
WordNet ［license wordnet］
「(astronomy) any perturbation of the mean motion or orbit of a planet or satellite (especially a perturbation of the earth''s moon)」WordNet ［license wordnet］
「the process of varying or being varied」WordNet ［license wordnet］
「a repetition of a musical theme in which it is modified or embellished」PrepTutorEJDIC ［license prepejdic］
「{U}『変化(変動)すること』 / {C}(個々の)変化;変化の量(程度) / {C}変形,変わり種;(生物の)変種 / {C}(…の)変奏曲《+on+名》」WordNet ［license wordnet］
「an activity that varies from a norm or standard; "any variation in his routine was immediately reported"」WordNet ［license wordnet］
「an artifact that deviates from a norm or standard; "he patented a variation on the sandal"」WordNet ［license wordnet］
「the act of changing or altering something slightly but noticeably from the norm or standard; "who is responsible for these variations in taxation?"」
「disperse」
 vt.
 (人々)を司法に散らす、分散させる。(ビラなど)をまき散らす
 (苦痛など)を追い払う。(疑いなど)をはらす
 (知識など)を広める、普及させる。(人)を各所に配置する。
 (物理)(光)を分光する。(化)(微粒子)を分散させる
 vi.
 adj.
 関
 discrete、dispersal、dispersion、intersperse、variance
WordNet ［license wordnet］
「separate (light) into spectral rays; "the prosm disperses light"」PrepTutorEJDIC ［license prepejdic］
「…‘を'『四方に散らす』 / …‘を'追い払う;〈不安など〉‘を'消し去る / 〈微粒子〉‘を'分散させる;〈光〉‘を'分光する / …‘を'各所に配置する / 『散らばる』,分散(離散)する / 〈微粒子が〉分散する,〈光が〉分光する」WordNet ［license wordnet］
「to cause to separate and go in different directions; "She waved her hand and scattered the crowds"」WordNet ［license wordnet］
「move away from each other; "The crowds dispersed"; "The children scattered in all directions when the teacher approached";」
「inconsistent」
 adj.
 [叙述](～の点で)一貫性のない、無定見な(in)。(～と)一致しない、調和しない、(信条などと)矛盾する(with)。ムラのある
 (数学)不能な、整合しない(incompatible)
 (論理)(1組の命題が)明らかな矛盾を導き出すことのできる
 関
 (adv.)inconsistently
 disagree、disagreement、discordance、incommensurate、inconsistency、indefinite、reciprocal、repulsion、variable、variably、variance
WordNet ［license wordnet］
「not capable of being made consistent or harmonious; "inconsistent accounts"」PrepTutorEJDIC ［license prepejdic］
「食い違っている,矛盾する」WordNet ［license wordnet］
「displaying a lack of consistency; "inconsistent statements cannot both be true at the same time"; "inconsistent with the roadmap"」
「vary」
 v.
 関
 change、differ、different、disparate、dissimilar、distinct、fluctuate、fluctuation、oscillate、shift、switch、turn to、variance、variant、variation
WordNet ［license wordnet］
「be subject to change in accordance with a variable; "Prices vary"; "His moods vary depending on the weather"」PrepTutorEJDIC ［license prepejdic］
「(…から…に)『変わる』,変化する《+from+名+to+名》 / (…の点で)『互いに異なる』,いろいろである《+in(on)+名》 / …‘を'『変える』,変更する / …‘に'変化をつける,‘を'変化に富ませる」WordNet ［license wordnet］
「make something more diverse and varied; "Vary the menu"」
「intersperse」
 分散させる、散在させる
WordNet ［license wordnet］
「place at intervals in or among; "intersperse exclamation marks in the text"」PrepTutorEJDIC ［license prepejdic］
「(…の間に)…‘を'散らばせる,まき散らす《+『名』+『among』(『between』)+『名』》 / 《通例受動態で》(他のものを散りばめて)…‘に'変化を与える;(他のものを)…‘に'散りばめる《+『名』+『with』+『名』》」
「invariance」
WordNet ［license wordnet］
「the nature of a quantity or property or function that remains unchanged when a given transformation is applied to it; "the invariance of the configuration under translation"」
「covariance」
WordNet ［license wordnet］
「(statistics) the mean value of the product of the deviations of two variates from their respective means」
「analysis of covariance」