Date:         Tue, 27 Sep 2005 09:14:27 +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: Help with nested design syntax MANCOVA
In-Reply-To:  <20050926.222208.632.6.babyblue325@juno.com>
Content-Type: text/plain; charset=ISO-8859-15

Hi Jordan

Although without a more complete explanation of your data design I
can't be sure, I see two errors in your model:

- A nested factor can NEVER interact with the nesting factor. That
  makes the terms: location*severity and treat_gp*location*severity
  wrong. Are you sure it is nested? Explain it with more detail.

- Also, you say you have 3 factors, and I see 5 (gender, treatment,
  group, location and severity, the last one being nested within
  location). Gender cannot be a covariate, as you have wrongly indicated
  in your model, because it's a qualitative variable (a factor), not a
  quantitative one.

Perhaps someone with more expertise in MANOVA models (I use don't use
it very often, nor I have read a lot about it) can help you more.

JH> I cannot figure out if my nested design sytax is working.
JH> Can you have a look and see if I've coded it correctly. I have 3 factors,
JH> treatment group, location and severity. Severity is nested within
JH> location.

JH> GLM
JH> recov_sp adh_scor BY treat_gp location severity WITH gender miqr_sum
JH> /METHOD = SSTYPE(3)
JH> /INTERCEPT = INCLUDE
JH> /CRITERIA = ALPHA(.05)
JH> /DESIGN = gender miqr_sum treat_gp location severity(location)
JH> treat_gp*location treat_gp*severity location*severity
JH> treat_gp*location*severity .

Regards,
 Marta                            mailto:biostatistics@terra.es