Date:         Fri, 21 Dec 2007 11:45:09 -0800
Reply-To:     gigi1234 <ggramig@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         gigi1234 <ggramig@GMAIL.COM>
Organization: http://groups.google.com
Subject:      multivariate normality
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

Hi All and Happy Holidays!

Does anyone know an *easy* test for multivariate normality?

I assume it isn't sufficient to test for univariate normality.

I'd like to do some stepdisc and candisc procs but I am leary of the
assumptions.

Thanks, GG