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 Statistical Tests Comparing Variances One Sample Chi-Square-Test | |
See also: two sample F-Test, Chi-Square Distribution, survey on statistical tests |
Certain problems require not only that the mean conforms to some restrictions, but also that the variance is within certain limits, i.e. not larger than a given value. So we have to compare the estimated sample variance with the hypothetical variance s^{2}. When the samples are normally distributed, the ratio .(n-1) / s^{2} follows a -distribution (pronounced: chi-square).
The upper tails of the distribution have been tabulated (or you may
use the distribution calculator). (a)
depicts the area of a% in the upper tail of
the distribution,
i.e. Prob(
> (a
))
= a . The shape of the -distribution
depends on the degrees of freedom n-1.
Last Update: 2005-Jul-16