Devin,
The first thing that you need to do is to develop your hypothesis.
Once you have your hypothesis, then you can create your statistical
models.
For example, it seems like you want look at the change in fMRI
activity between two scans with covariates. This is most easily done
with a one-sample t-test. For the fMRI images, simple subtract one
scan from the other for each subject. For your covariates, you need to
decide what the relationship is that you are interested in testing.
Here are some examples:
(a) the change in fMRI is associated with the difference in number of
errors made --> covariate should be the difference in errors.
(b) the change in fMRI is associated with the change in hippocampal
volume --> covariate should be the change in hippocampal volume
(c) the change in fMRI is associated with hippocampal volume at time
1 --> covariate should be hippocampal volume at time 1.
Hope this helps.
Also, if your are considering hippocampal volume and or change in
hippocampal volume, you might want to normalize the volumes based on
ICV or TBV.
Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
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 Mon, Nov 5, 2012 at 12:35 PM, Devin Sodums
<[log in to unmask]> wrote:
> Dear SPM users,
>
> I am running the group level analyses for a project where we had subjects do a learning and memory task two times, each time with different but equivalent stimuli. So I am interested in comparing the activation of the group of subject's first scan to that of their second scan. In the past we have used three covariates for the second level model specs: the number of errors committed on each trial of the task, and then two for the ratio of left and right hippocampal volumes of each subject divided by total intracranial volume. We have used these covariates in the past for just looking at activations among subjects who did the test once, so I am not positive what the best way to use covariates when doing a test-retest analysis.
>
> My original idea for the errors, was to do # errors in scan 2- # errors in scan 1. Does this seem logical?
>
> And for the hippocampal volumes, would it be logical to do the same? I don't imagine that they have changed much, but I don't know quite yet as am waiting for the final subject's freesurfer to finish running. So if they haven't, I would imagine then I could simply take the volumes from the most recent scan and use that. But if they have changed, then would it make sense also do a difference between scans and use that?
>
> Any advice would be greatly appreciated!
>
> Many thanks,
> Devin Sodums
> Research assistant, Jones-Gotman Lab
> Montreal Neurological Institute
|