# Cramér-von-Mises criterion

Cramér-von-Mises criterion

In statistics the Cramér-von-Mises criterion is a form of minimum distance estimation used for judging the goodness of fit of a probability distribution $F^*$ compared to a given distribution $F$ is given by

:$n W^2 = n int_\left\{-infty\right\}^\left\{infty\right\} \left[F\left(x\right)-F^*\left(x\right)\right] ^2,mathrm\left\{d\right\}F\left(x\right)$

In one-sample applications $F$ is the theoretical distribution and $F^*$ is the empirically observed distribution. Alternatively the two distributions can both be empirically estimated ones; this is called the two-sample case.

The criterion is named after Harald Cramér and Richard Edler von Mises who first proposed it in 1928-1930. The generalization to two samples is due to Anderson (1962).

The Cramér-von-Mises test is an alternative to the Kolmogorov-Smirnov test. It is thought that the CvM test is more powerful than the KS test, but this has not been shown theoretically.

Cramér-von-Mises test (one sample)

Let $x_1,x_2,cdots,x_n$ be the observed values, in increasing order. Then it is possible to show that

:$T = n W^2 = frac\left\{1\right\}\left\{12n\right\} + sum_\left\{i=1\right\}^n left \left[ frac\left\{2i-1\right\}\left\{2n\right\}-F\left(x_i\right) ight\right] ^2.$

If this value is larger than the tabulated value we can reject the hypothesis that the data come from the distribution $F\left(.\right)$.

Cramér-von-Mises test (two samples)

Let $x_1,x_2,cdots,x_n$ and $y_1,y_2,cdots,y_m$ be the observed values in the first and second sample respectively, in increasing order. Let $r_1,r_2,cdots,r_n$ be the ranks of the x's in the combined sample, and let $s_1,s_2,cdots,s_m$ be the ranks of the y's in the combined sample. It can be shown that

:$T = n W^2 = frac\left\{U\right\}\left\{n m \left(n+m\right)\right\}-frac\left\{4 m n - 1\right\}\left\{6\left(m+n\right)\right\}$

where U is defined as

:$U = n sum_\left\{i=1\right\}^n \left(r_i-i\right)^2 + m sum_\left\{j=1\right\}^m \left(s_j-j\right)^2$

If the value of T is larger than the tabulated value we can reject the hypothesis that the two samples come from the same distribution. (Some books give critical values for U, which is more convenient, as it avoids the need to compute T via the expression above. The conclusion will be the same).

References

*cite journal
last= Anderson
first= TW
year= 1962
title= On the Distribution of the Two-Sample Cramer-von Mises Criterion
journal= The Annals of Mathematical Statistics
volume= 33
issue= 3
pages= 1148–1159
publisher= Institute of Mathematical Statistics
issn= 0003-4851
doi= 10.1214/aoms/1177704477
url= http://projecteuclid.org/DPubS/Repository/1.0/Disseminate?view=body&id=pdf_1&handle=euclid.aoms/1177704477
format= PDF
accessdate= 2008-09-23

Xiao, Gordon, Yakovlev: 'A C++ program for the Cramér-von-Mises two sample test', Journal of Statistical Software, 17 #8, January 2007 [http://www.jstatsoft.org/v17/i08/paper]

Wikimedia Foundation. 2010.

### Look at other dictionaries:

• Cramér–von Mises criterion — In statistics the Cramér–von Mises criterion is a criterion used for judging the goodness of fit of a cumulative distribution function F * compared to a given empirical distribution function Fn, or for comparing two empirical distributions. It is …   Wikipedia

• Cramér-von-Mises-Test — Der Cramér von Mises Test ist ein statistischer Test, mit dem untersucht werden kann, ob die Häufigkeitsverteilung der Daten einer Stichprobe von einer vorgegebenen hypothetischen Wahrscheinlichkeitsverteilung abweicht (Ein Stichproben Fall),… …   Deutsch Wikipedia

• Richard von Mises — Infobox Scientist name = Richard von Mises box width = image width =150px caption = Richard von Mises birth date = 19 April 1883 birth place = Lemberg death date = 14 July 1953 death place = Boston residence = citizenship = nationality =… …   Wikipedia

• Harald Cramér — Born 25 September 1893(1893 09 25) Stockholm, Sweden …   Wikipedia

• Minimum distance estimation — (MDE) is a statistical method for fitting a mathematical model to data, usually the empirical distribution. Contents 1 Definition 2 Statistics used in estimation 2.1 Chi square criterion …   Wikipedia

• List of statistics topics — Please add any Wikipedia articles related to statistics that are not already on this list.The Related changes link in the margin of this page (below search) leads to a list of the most recent changes to the articles listed below. To see the most… …   Wikipedia

• List of mathematics articles (C) — NOTOC C C closed subgroup C minimal theory C normal subgroup C number C semiring C space C symmetry C* algebra C0 semigroup CA group Cabal (set theory) Cabibbo Kobayashi Maskawa matrix Cabinet projection Cable knot Cabri Geometry Cabtaxi number… …   Wikipedia

• Normality test — In statistics, normality tests are used to determine whether a data set is well modeled by a normal distribution or not, or to compute how likely an underlying random variable is to be normally distributed. More precisely, they are a form of… …   Wikipedia

• Test de normalité — En statistiques, les tests de normalité permettent de vérifier si des données réelles suivent une loi normale ou non. Les tests de normalité sont des cas particuliers des tests d adéquation (ou tests d ajustement, tests permettant de comparer des …   Wikipédia en Français

• Tests de normalité — Test de normalité En statistiques, les tests de normalité permettent de vérifier que des données réelles suivent une loi normale ou non. Les tests de normalité sont des cas particuliers des tests d adéquation (ou tests d ajustement, tests… …   Wikipédia en Français