|Date: ||Wed, 8 Nov 2000 14:43:27 -0600|
|Reply-To: ||"Nichols, David" <nichols@SPSS.COM>|
|Sender: ||"SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>|
|From: ||"Nichols, David" <nichols@SPSS.COM>|
|Subject: ||Re: profile analysis|
This is just a 2x11 or 11x2 within subjects or repeated measures design. If
the data are listed in the order given here, then the structure would be as
GLM a1 to b11
/WSFACTORS ab 2 time 11.
This is also easily specifiable through the dialog boxes for GLM - Repeated
Measures. In either case, you'll get the parallelism test (the interaction)
and the tests for levels across time and ab (the main effects). Since the
design is balanced, if parallelism holds, you can just look at the main
effects, as rerunning the model without the interaction is not necessary
(and wouldn't produce different numbers for the main effects).
Principal Support Statistician and
Manager of Statistical Support
> -----Original Message-----
> From: Stefano Perlusz [mailto:perlusz@IPROD.AUC.DK]
> Sent: Sunday, November 05, 2000 12:11 PM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: [SPSSX-L] profile analysis
> Dear colleagues,
> I have 11 couples of dependent variables: a1 - b1; a2 - b2;
> ... a11 - b11.
> I want to see if these 2 sets of 11 variables show different
> profiles. Do
> the 2 sets have the same highs and lows (are they parallel)?
> Do they have
> different overall levels?
> I understand that profile analysis (manova with dependent variables
> measured on a same scale) is what I need. But I can't create
> the syntax
> file, as I don't know how to tell that the 22 variables are
> actually 2 sets
> of 11.
> I would very much appreciate any help!