Chi-Square Distribution


Also found in: Dictionary, Thesaurus, Medical, Legal, Financial, Wikipedia.

chi-square distribution

[′kī ¦skwer dis·trə′byü·shən]
(statistics)
The distribution of the sum of the squares of a set of variables, each of which has a normal distribution and is expressed in standardized units.

Chi-Square Distribution

 

The probability distribution of the sum

of the squares of the normally distributed random variables X1, . . ., Xf, with zero mathematical expectation and unit variance is known as a chi-square distribution with f degrees of freedom. The distribution function for a chi-square random variable is

The first three moments (the mathematical expectation, the variance, and the third central moment) of χ2 are f, 2f, and 8f, respectively. The sum of two independent random variables Chi-Square Distribution and Chi-Square Distribution with f1 and f2 degrees of freedom has a chi-square distribution with f1 + f2 degrees of freedom.

Examples of chi-square distributions are the distributions of the squares of random variables that obey the Rayleigh and Maxwellian distributions. The Poisson distribution can be expressed in terms of a chi-square distribution with an even number of degrees of freedom:

If the number f of terms of the sum χ2 increases without bound, then, according to the central limit theorem, the distribution function of the standardized ratio Chi-Square Distribution converges to the standard normal distribution:

where

A consequence of this fact is another limit relation, which is convenient for calculating Ff(x) when f has large values:

In mathematical statistics, the chi-square distribution is used to construct interval estimates and statistical tests. Let Yi, . . ., Yn be random variables representing independent measurements of an unknown constant a. Suppose the measurement errors Yia are independent and are distributed identically normally. We have

E(Yia) = 0 E(Yia)2 = σ2

The statistical estimate of the unknown variance σ2 is then expressed by the equation

s2 = S2/(n – 1)

where

The ratio S22 obeys a chi-square distribution with f = n – 1 degrees of freedom. Let x1 and x2 be positive numbers that are solutions of the equations Ff(x1) = α/2 and Ff(x2) = 1 – α/2, where α is a specified number in the interval (0,1/2). In this case

P{x1 < S22 < x2} = P{S2/x2 < σ2 < S2/x1} = 1 – α

The interval (S2/x1, S2/x2 is called the confidence interval for σ2 with confidence coefficient 1 – α.

This method of constructing an interval estimate for σ2 is often used to test the hypothesis that Chi-Square Distribution, where Chi-Square Distribution is a given number. Thus, if Chi-Square Distribution belongs to the confidence interval indicated, then one concludes that the measurements do not contradict the hypothesis Chi-Square Distribution. If, however, Chi-Square Distribution or Chi-Square Distribution, then it must be assumed that Chi-Square Distribution or Chi-Square Distribution, respectively. This test corresponds to a significance level equal to α.

REFERENCE

Cramer, H. Matematicheskie metody statistiki, 2nd ed. Moscow, 1975. (Translated from English.)

L. N. BOLSHEV

References in periodicals archive ?
The actual computation of the exact assurance probability [GAMMA]([upsilon], [[??].sup.2.sub.g]) in Equation 8 requires the evaluation of noncentral t cumulative distribution function and the one-dimensional integration with respect to a chi-square distribution. Specifically, for each possible value of the sample variance [[??].sup.2.sub.p] with a scaled chi-square distribution, the least sample size [[??].sub.g] is computed so that the corresponding power [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], where [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] and g is calculated as in Equation 11.
2.2 If [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] MVN(0,1), then if A is a symmetric idempotent matrix of rank p, then ZAZ has a chi-square distribution with p degrees of freedom.
If the forecast error has a multivariate normal distribution with mean 0 and variance [OMEGA], then the analog of (1) is Box Multivariate Normal and Chi-Square Distributions A univariate normal distribution is characterized by two numbers, the mean and the variance.
The median of the chi-square distribution for a given number of degrees of freedom is the score that corresponds to a P value of 0.500.
The shape of the chi-square distribution, like the t, depends upon [nu], degrees of freedom associated with [S.sup.2].
Under the null hypothesis of no ARCH effects, the LM statistic follows a Chi-square distribution with degree of freedom (df) equal to the number of lags in the auxiliary regression.
We undertook to do this by generating large "Monte Carlo" chi-square distributions (see Data Sources and Methods, Section 2) representing thousands of first-difference turn tests structured as in Figure 1.
The chi-square statistic calculated in each case was evaluated with the chi-square distribution for (L - 1)(S - 1) degrees of freedom.
In any case where 10% [is less than] Q([X.sup.2] [where] v) [is less than] 90%, [X.sup.2] falls sufficiently close to the center of a Chi-square distribution with v degrees of freedom that it is very probable from such a distribution, validating the hypothesis being tested.
In our example, the p-values based on the nominal chi-square distribution were often similar to those based on the Monte Carlo simulations.
The statistical power of the chi-square test is evaluated from the noncentral chi-square distribution (Broffitt and Randles, 1977; Guenther, 1977; Weir and Cockerham, 1978).