Thank you for your clarifications.
Following them I specify a flexible factorial but I get a strange design
matrix with beta not uniquely specified. I attached my design matrix. Could
you tell me what do you think about it?
Just another general question. I know that when I have between-subject
factors (normals vs. patients) I specify UNEQUAL for Variance and YES for
Indipendence while when I have within-subject facotrs (task1 vs. task2) I
specify EQUAL for Variance and NO for Indipendence. Is it better to set
UNEQUAL for variance also within-subject factors?
Best,
Federico.
-----Messaggio originale-----
Da: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] Per
conto di Volkmar Glauche
Inviato: luned́ 15 gennaio 2007 9.20
A: [log in to unmask]
Oggetto: Re: [SPM] Conditions in FLEXIBLE factorial
Dear Federico,
On Sat, 13 Jan 2007, Federico Tubaldi wrote:
> I'm new to spm5 and I try to set a flexible factorial design (I use
> Specify Subjects mode). I have a within subject design: A(a1,a2)xB(b1,b2).
>
> I define the first factor "subject", the second factor A, the third
> factor B. For all factors I set No for Indipendence and Unequal for
> Variance. After I specify four images (one for condition) obtained at
> the 1st level as scans for each subject.
It is a bad idea to specify unequal variance and depency between levels for
all your factors. In the best case you will get a very distorted model, in
the worst case, the REML estimation will fail to estimate error covariances
completely. The magic is to decide, which factor(s) should be treated as
having equal variance and to model dependent variance only for factors where
there is a reason to assume it.
In your case, if your subjects are all drawn from the same population, the
"subject" factor will have equal and independent variance. You don't need
any "repl" factor as long as you don't have repeated measurements for each
subject.
> Now my questions are:
> Shall I define also the factor "repl" before defining the factor "subject"
> or they are the same (I think that replications are my subjects)?
> What Do I type to define Conditions? I type 1 2 3 4 for each subject
> but when I run I get "Error running job: Index exceeds matrix dimensions".
You don't have 4 conditions, but 2x2, right? So your conditions need to be a
matrix containing the A and B levels for each scan like this:
[1 1 2 2;
1 2 1 2]
Volkmar
>
> Many thanks in advance.
>
> Federico.
>
>
--
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/
--
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.432 / Virus Database: 268.16.10/626 - Release Date: 14/01/2007
20.29
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.432 / Virus Database: 268.16.10/626 - Release Date: 14/01/2007
20.29
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.432 / Virus Database: 268.16.10/626 - Release Date: 14/01/2007
20.29
|