Wendy wrote:
> Hi,
>
> I'm looking for a test statistic for trivariate normality. Does anybody know
> such a test-statistic, respectively a book/website where I can find one ?
>
> Thanks !
>
> Wendy
If you transform the components to independence (using the Choleski factor
of the covariance matrix, for instance) and scale each individually, the sum of
squares of the three N(0,1) components behave like a chi-squared with
3 degrees of freedom. Then perform a goodness of fit test.
--
gus gassmann
"When in doubt, travel"
======================================================================
To reply, remove SPAM BLOCK
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
