You can't include a covariate like this.
The covariate is a linear combination of the subject terms -- so it
will have no effect on the data fitting process. However, I am
surprised the that the values are identical. The p-value and z-score
should change slightly since the df changes.
The covariates don't cancel out, but as for the reason above, they
will have no effect on the estimates of the task factor terms.
Let me know if you need further clarification.
Why do you want to include a covariate?
Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Postdoctoral Research Fellow, GRECC, Bedford VA
Research Fellow, Department of Neurology, Massachusetts General
Hospital and Harvard Medical School
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain
PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED
and which is intended only for the use of the individual or entity
named above. If the reader of the e-mail is not the intended recipient
or the employee or agent responsible for delivering it to the intended
recipient, you are hereby notified that you are in possession of
confidential and privileged information. Any unauthorized use,
disclosure, copying or the taking of any action in reliance on the
contents of this information is strictly prohibited and may be
unlawful. If you have received this e-mail unintentionally, please
immediately notify the sender via telephone at (773) 406-2464 or
email.
On Thu, Sep 2, 2010 at 11:43 AM, Lars Meyer <[log in to unmask]> wrote:
> dear all,
>
> i am running second-level analyses in spm 8, using the 'flexible factorial' design with the two built-in factors 'repl' and 'subject'. four conditions, as given by four images (four conditions) for each subject—that just works fine, so far, so good, nice results.
>
> i now want to include a covariate in the model, with a single number for each subject (=a digit span measure). i included my digit span vector (one value for each subject) as a covariate in the model, such that i repeat each entry of the vector four times (four conditions), using 'kron' or 'reshape' or whatever.
>
> the models runs through nicely, but my covariate does not have any effect, i.e. all values (p, z, etc.) stay identical—i wonder why this is the case, and i am very happy about any suggestions!
>
>
> thanks,
> l.
>
>
>
> (ps: is it so that the covariate (which is now identical for each condition, since i replicated the original vector) cancels itself out in any contrast i specify? e.g. if i go for '-1 1 -1 1', the associated covariates cancel each other out?)
>
>
>
>
>
> Lars Meyer | MSc
> Max Planck Institute for Human Cognitive & Brain Sciences
> Department of Neuropsychology
> Stephanstraße 1a
> 04103 Leipzig
>
> Office | +49 (0)341 99 40 22 66
> Fax | +49 (0)341 99 40 22 60
> Mobile | +49 (0)175 113 76 65
> Home | +49 (0)341 974 26 49
> Mail | [log in to unmask]
> Web | www.cbs.mpg.de/~lmeyer
>
|