Date:         Tue, 22 Feb 2000 14:43:40 -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: follow up tests in ANOVA

I'm afraid that at this time LMATRIX doesn't offer the ability to assign
custom error terms. There are only two places where you can specify an error
term: POSTHOC and TEST (TEST is restricted to testing terms in the ANOVA
table, so it's not helpful for this situation).

David Nichols
Principal Support Statistician and
Manager of Statistical Support
SPSS Inc.

> -----Original Message-----
> From: Perkins, Nigel [mailto:N.R.Perkins@MASSEY.AC.NZ]
> Sent: Sunday, February 13, 2000 1:55 PM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: [SPSSX-L] follow up tests in ANOVA
>
>
> All,
> I am using SPSS 9.0 and am trying to figure out how to do contrasts in
> nested and mixed ANOVA models (using UNIVARIATE), which will use the
> appropriate MS term as an error term (denominator) in testing the
> significance of the contrast.
>
> I can use the POSTHOC command in syntax by adding "VS
> correct_MS_term" and
> this will allow me to perform follow up comparisons between
> simple main
> effects involving fixed effects terms. However, when I wish
> to define a
> particular contrast using an "LMATRIX" set up, I cannot seem
> to get the
> correct error term in to the syntax.
>
> At the moment, I am mostly identifying the correct MS term
> and then redoing
> the appropriate comparisons manually.
>
> Does any one have any advice for this?
> Thanks,
> Nigel
> Nigel Perkins, BVSc, MS, Dip ACT, FACVSc
> Senior Lecturer, Veterinary Epidemiology
> EpiCentre, IVABS
> Massey University
> Private Bag 11222
> Palmerston North
> New Zealand
> N.R.Perkins@massey.ac.nz
> http://epicentre.massey.ac.nz
>