# Scaled inverse chi-squared distribution

Template:Probability distribution
The **scaled inverse chi-squared distribution** is the distribution for *x* = 1/*s*^{2}, where *s*^{2} is a sample mean of the squares of ν independent normal random variables that have mean 0 and inverse variance 1/σ^{2} = τ^{2}. The distribution is therefore parametrised by the two quantities ν and τ^{2}, referred to as the *number of chi-squared degrees of freedom* and the *scaling parameter*, respectively.

This family of scaled inverse chi-squared distributions is closely related to two other distribution families, those of the inverse-chi-squared distribution and the inverse gamma distribution. Compared to the inverse-chi-squared distribution, the scaled distribution has an extra parameter *τ*^{2}, which scales the distribution horizontally and vertically, representing the inverse-variance of the original underlying process. Also, the scale inverse chi-squared distribution is presented as the distribution for the inverse of the *mean* of ν squared deviates, rather than the inverse of their *sum*. The two distributions thus have the relation that if

Compared to the inverse gamma distribution, the scaled inverse chi-squared distribution describes the same data distribution, but using a different parametrization, which may be more convenient in some circumstances. Specifically, if

Either form may be used to represent the maximum entropy distribution for a fixed first inverse moment and first logarithmic moment .

The scaled inverse chi-squared distribution also has a particular use in Bayesian statistics, somewhat unrelated to its use as a predictive distribution for *x* = 1/*s*^{2}. Specifically, the scaled inverse chi-squared distribution can be used as a conjugate prior for the variance parameter of a normal distribution. In this context the scaling parameter is denoted by σ_{0}^{2} rather than by τ^{2}, and has a different interpretation. The application has been more usually presented using the inverse gamma distribution formulation instead; however, some authors, following in particular Gelman *et al.* (1995/2004) argue that the inverse chi-squared parametrisation is more intuitive.

## Characterization

The probability density function of the scaled inverse chi-squared distribution extends over the domain and is

where is the degrees of freedom parameter and is the scale parameter. The cumulative distribution function is

where is the incomplete Gamma function, is the Gamma function and is a regularized Gamma function. The characteristic function is

where is the modified Bessel function of the second kind.

## Parameter estimation

The maximum likelihood estimate of is

The maximum likelihood estimate of can be found using Newton's method on:

where is the digamma function. An initial estimate can be found by taking the formula for mean and solving it for Let be the sample mean. Then an initial estimate for is given by:

## Bayesian estimation of the variance of a Normal distribution

The scaled inverse chi-squared distribution has a second important application, in the Bayesian estimation of the variance of a Normal distribution.

According to Bayes theorem, the posterior probability distribution for quantities of interest is proportional to the product of a prior distribution for the quantities and a likelihood function:

where *D* represents the data and *I* represents any initial information about σ^{2} that we may already have.

The simplest scenario arises if the mean μ is already known; or, alternatively, if it is the conditional distribution of σ^{2} that is sought, for a particular assumed value of μ.

Then the likelihood term *L*(σ^{2}|*D*) = *p*(*D*|σ^{2}) has the familiar form

Combining this with the rescaling-invariant prior p(σ^{2}|*I*) = 1/σ^{2}, which can be argued (e.g. following Jeffreys) to be the least informative possible prior for σ^{2} in this problem, gives a combined posterior probability

This form can be recognised as that of a scaled inverse chi-squared distribution, with parameters ν = *n* and τ^{2} = *s*^{2} = (1/*n*) Σ (x_{i}-μ)^{2}

Gelman *et al* remark that the re-appearance of this distribution, previously seen in a sampling context, may seem remarkable; but given the choice of prior the "result is not surprising".^{[1]}

In particular, the choice of a rescaling-invariant prior for σ^{2} has the result that the probability for the ratio of σ^{2} / *s*^{2} has the same form (independent of the conditioning variable) when conditioned on *s*^{2} as when conditioned on σ^{2}:

In the sampling-theory case, conditioned on σ^{2}, the probability distribution for (1/s^{2}) is a scaled inverse chi-squared distribution; and so the probability distribution for σ^{2} conditioned on *s*^{2}, given a scale-agnostic prior, is also a scaled inverse chi-squared distribution.

### Use as an informative prior

If more is known about the possible values of σ^{2}, a distribution from the scaled inverse chi-squared family, such as Scale-inv-χ^{2}(*n*_{0}, *s*_{0}^{2}) can be a convenient form to represent a less uninformative prior for σ^{2}, as if from the result of *n*_{0} previous observations (though *n*_{0} need not necessarily be a whole number):

Such a prior would lead to the posterior distribution

which is itself a scaled inverse chi-squared distribution. The scaled inverse chi-squared distributions are thus a convenient conjugate prior family for σ^{2} estimation.

### Estimation of variance when mean is unknown

If the mean is not known, the most uninformative prior that can be taken for it is arguably the translation-invariant prior *p*(μ|*I*) ∝ const., which gives the following joint posterior distribution for μ and σ^{2},

The marginal posterior distribution for σ^{2} is obtained from the joint posterior distribution by integrating out over μ,

This is again a scaled inverse chi-squared distribution, with parameters and .

## Related distributions

- If then
- If (Inverse-chi-squared distribution) then
- If then (Inverse-chi-squared distribution)
- If then (Inverse-gamma distribution)
- Scaled inverse chi square distribution is a special case of type 5 Pearson distribution

## References

- Gelman A.
*et al*(1995),*Bayesian Data Analysis*, pp 474–475; also pp 47, 480

- ↑ Gelman
*et al*(1995),*Bayesian Data Analysis*(1st ed), p.68

- REDIRECT Template:Probability distributions