Date:         Wed, 28 Oct 1998 20:07:54 +0000
Reply-To:     Edward Tverdek <tverdek@INTERACCESS.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         Edward Tverdek <tverdek@INTERACCESS.COM>
Organization: http://www.supernews.com,
              The World's Usenet: Discussions Start Here
Subject:      Re: canonical R
Content-Type: text/plain; charset=us-ascii

Michael,
 
The SPSS Advanced Statistics package installs an SPSS macro called
"Canonical Correlation.sps" to your SPSS directory.  It's an SPSS
command syntax file which, provided it is invoked along with two
variable lists, will provide canonical correlations.  If your GradPack
has the Advanced Stats package (I think they all do), you should be
able to run it (the macro invokes the MATRIX procedure available only
in this module).  You'll want to consult your SPSS Advanced Statistics
guide for details, however, and to confirm that it will run with
variable lists of different lenghts.
 
Regards,
 
Ed Tverdek
 
 
 
Michael Conley wrote:
>
> I am a new user of SPSS graduate pack.  I cannot see how to get a canonical r
> for two sets of variables with different number of variables in each set.
The
> GLM multivariate will run a one cell analysis with dependent variables and
> covariates but it does not provide a canonical r.  Is there a way to do
this??
> It does provide all of the multiple r's but I would like to get the canonical
> r between the canonical varibles.  Thanks for any advice.