[R] D'agostino test

Diethelm Wuertz wuertz at itp.phys.ethz.ch
Wed Sep 1 19:34:20 CEST 2004

Ludwig Baringhaus wrote:

>>Several versions of the D'Agostino Test are implemented in
>>"fBasics" from Rmetrics beside many other tests for normality

>Unlike the Shapiro-Wilk or the Anderson-Darling test, the 
>D'Agostino test is not an omnibus test for testing the
>hypothesis of normality. In fact, D'Agostino's D 
>is a suitable test statistic for testing the hypothesis of 
>uniformity. See 
>Baringhaus and Henze (1990). 
>A test for uniformity with unknown limits based on D'Agostino's D. 
>Statist. Probab. Lett. 9, 299-304 

So far, my understanding was the following that the DÂ’Agostino(-Pearson) 
omnibus test
analyzes the data to determine skewness and kurtosis. It then calculates 
how far each
of these values differs from the value expected with a Gaussian 
distribution, and
computes a P value from the sum of the squares of these discrepancies. 
Unlike the
Shapiro-Wilk test, this test is not affected if the data contains 
identical values.

This kind of test is implemented in several other software packages 
under the name
"d'Agostino Test".

May be we are talking about different things.

D. Wuertz

PS: Thanks for the reference, I will have a look on it to clarify my 

E.g., see also:
R.B. D'Agostino, "Tests for Normal Distribution" in /Goodness-Of-Fit 
E. Seier, "Comparisons of Tests for Univariate Normality"
J.D. McCauley, "Goodness of Fit Tests"
UITS, "Testing for Normality"

>L. Baringhaus
>Institut fuer Mathematische Stochastik
>Universitaet Hannover
>Welfengarten 1
>30167 Hannover
>R-help at stat.math.ethz.ch mailing list
>PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

More information about the R-help mailing list