Date:         Wed, 12 May 2004 11:38:14 -0400
Reply-To:     Maria Parmley <parmley@BRANDEIS.EDU>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Maria Parmley <parmley@BRANDEIS.EDU>
Subject:      ANCOVA with varying covariate

I am doing an ANCOVA.  I have a mixed design with 4 within subject
variables, and 2 between subject variables.  I also have a varying
covariate.  I've been told that you need to use teh MANOVA command in SPSS
if you have a covarying covariate.  I have used the following command
syntax:

MANOVA iden1 iden2 iden3 iden4 iden5 iden6 iden7 iden8 iden9 iden10
iden11 iden12 iden13 iden14 iden15 iden16 WITH idenx_1
idenx_2 idenx_3 idenx_4 idenx_5 idenx_6 idenx_7 idenx_8 idenx_9 idenx_10
 idenx_11 idenx_12 idenx_13 idenx_14 idenx_15 idenx_16
/wsfactors ssex(2) sage(2) sexpress(2) guess(2)
/wsdesign.

My question is how to include the between subjects variables in the syntax
(when I include the "BY" command I get an error message).  Also, how do I
test for homogeneity of regression in MANOVA?  Finally, in GLM I've used
the emmeans subcommand to analyze simple effects--is there a way to do
this with the MANOVA command?

Thanks,

Maria