Hi Markus,
this time you would need a full/flexible factorial design as general
layout to define group/subject/other factors. You would then enter your
covariates in the "other" place not used in the regression model. For the
covariate you want to test separately per group you would enter an
interaction with your e.g. group factor. If you used a flexible factorial
design, you would also need to model group main effects or other-by-group
interactions. Note, that this model is more general than the multiple
regression model, because you will model group means separately for your
groups, whereas in regression, you did model only an overall mean.
Volkmar
On Thu, 22 Feb 2007, Markus Burgmer wrote:
> hi ged.
> thanks for your reply about the covariates. that was exactly what i thought
> to do next, but i was unsure if this was correct to do so. so you helped me
> a lot in this.
> may i ask you another question about this topic?
> what can i do, if i now want to see if the one group of my participants
> (patients) show other brain regions which correlate with the pain in
> contrast to the group of healthy controls, still controlling for the
> nuisance variable of anxiety (my first question was about the correlation
> in the complete group, now i want to include the factor "group").
> do i have to model this in the "full factorial" model of spm5? if so, do
> you know a source where a dummy like me can understand how to model this in
> spm5 (i couldn't find anything useful in the spm5 manual)?
>
> thanks a lot
> markus
>
>
--
Volkmar Glauche
-
Department of Neurology [log in to unmask]
Universitaetsklinikum Freiburg Phone 49(0)761-270-5331
Breisacher Str. 64 Fax 49(0)761-270-5416
79106 Freiburg http://fbi.uniklinik-freiburg.de/
|