|You are working with the text-only light edition of "H.Lohninger: Teach/Me Data Analysis, Springer-Verlag, Berlin-New York-Tokyo, 1999. ISBN 3-540-14743-8". Click here for further information.|
|Table of Contents Univariate Data Sampling Distributions Chi-Square Distribution|
|See also: sampling distributions, t-distribution, F-distribution|
In order to make inferences about the population variance on the basis of the sample variance, we have to consider a special distribution, called chi-square () distribution: if a random variable Y is normally distributed with mean µ and variance s2, then the quantity
shows a distribution with n-1 degrees of freedom for a random sample of size n. Several examples of distributions for different degrees of freedom are shown in the figure below.
As you can see, the distribution is skewed and is always positive. The mean of the distribution is equal to the number of degrees of freedom n-1, the variance is twice the degrees of freedom. The distribution is tabulated in statistical tables, or can be calculated online by means of the distribution calculator. The distribution is used to test differences between population and sample variances, and between theoretical and observed distributions.
An important property of the distribution is its additivity: if two independent variables follow a distribution (exihibiting the degrees of freedom f1 and f2), then the sum of the two variables is also -distributed with a degree of freedom of f1+f2.
Last Update: 2005-Jšn-25