Date:         Tue, 27 Sep 2005 17:17:05 +0200
Reply-To:     Marta García-Granero
              <biostatistics@terra.es>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Marta García-Granero
              <biostatistics@terra.es>
Organization: Asesoría Bioestadística
Subject:      Re: Techincal questions about SPSS
In-Reply-To:  <200509270554.j8R4c9ph007649@malibu.cc.uga.edu>
Content-Type: text/plain; charset=ISO-8859-1

Hi

Although I still use SPSS 12, to my knowledge, these questions are not
SPSS 13 specific.

aGC> I have some questions about SPPS 13.0, not using scripts but using the
aGC> interactive menu and panels.

You loose a lot if you are not willing to use some syntax (different
from scripting, BTW) now and then. I know it can look fearful at the
start, but it's worth the effort, I assure you.

aGC> I would like to know how to specify a design with a nested factor in the
aGC> UNIVARIATE ANOVA procedure. How to use the WITHIN option clicking the
aGC> mouse?

That's ONE of the things you miss if you don't use syntax. With
UNIANOVA, nesting is specified in a model only by syntax. Try MIXED
procedure instead, nesting is available thru point and click. As I see
we speak the same language (you are from Costa Rica and I'm from
Spain), I could send you a mail with some instructions using the
Spanish version of SPSS.

aGC> I need to specify coefficients for specific contrasts in the GLM
aGC> procedure as in the One-Way ANOVA. In GLM, I can't add any coefficient and
aGC> it is only possible to contrast differences among a reference category
aGC> (last or first).

Syntax again...

aGC> How can I specify the split plot design, making at the same time a post
aGC> hoc test (Tukey test for example), knowing that the errors are different.

Syntax again... I have an example I could send to you. Unfortunately,
using UNIANOVA in split-plot designs will lead to wrong results in
post-hoc tests (the residual error term is always used, even if the
post-hoc test is needed for the plot factor, not the sub-plot factor).
I haven't found yet a way of using MIXED for this complex experimental
design, perhaps I'll pest those poor guys from SPSS who still dare to
answer questions here ;)

aGC> If you can not answer me these questions, please to indicate me who could
aGC> answer them.

Mmmm... that was a bit harsh, don't U think so?

--
Saludos,
Marta García-Granero        mailto:biostatistics@terra.es
University of Navarra
SPAIN