Hi Michael:
Thanks for quick answer. But I used a fixed effects model in both attempts...
-Christina
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Harms, Michael
Sent: Thursday, February 27, 2014 4:15 PM
To: [log in to unmask]
Subject: Re: [FSL] Second-level analyses question
Hi Christina,
In your 2nd level analysis, use a Fixed-effects analysis with just the bold runs from that subject. Otherwise, your estimation of the variance will be mixing across subjects/sessions.
cheers,
-MH
--
Michael Harms, Ph.D.
-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders Washington University School of Medicine Department of Psychiatry, Box 8134
660 South Euclid Ave. Tel: 314-747-6173
St. Louis, MO 63110 Email: [log in to unmask]
On 2/27/14 3:08 PM, "Christina Fales" <[log in to unmask]> wrote:
>Sort of a beginner question here. The FSL course docs suggest in
>various places that for a straightforward fMRI study, the standard
>approach is to run a First-level analysis for each bold, and to average
>across each subject's processed bolds in a Second-level analysis. Eg,
>for 3 subjects with 2 bolds each, the 2nd-level design would includeone
>EV for each
>subject:
>Grp EV1 EV2 EV3
>1 1 0 0
>1 1 0 0
>1 0 1 0
>1 0 1 0
>1 0 0 1
>1 0 0 1
>
>C1 (Sub 1): 1 0 0
>C2 (Sub 2): 0 1 0
>C3 (Sub 3): 0 0 1
>C4 (GrpMean): 1 1 1
>
>My question: For purposes of inputing individual-subject data to a
>3rd-Level analysis, is there an important difference between doing
>2nd-level analyses as above (averaging across each subject en masse),
>versus running each subject individually? In other words, as you are
>collecting data, you might want to average across each subject's bolds
>individually, rather than re-running all subjects together each time.
>Doing all subjects together each time takes longer and longer with each
>new subject.
>
>Moreover, on a subject by subject basis, I would expect the outputs to
>be equivalent. However, if you compare the output for Subject 2 from
>the above design (ie, by looking at cope1.feat/stats/cope2.nii.gz) with
>the output of processing Subject 2 individually (by looking at his/her
>cope1.feat/stats/cope1.nii.gz), you don't get images that are exactly
>the same. Why not?
>
>Basically, I'm wondering if it is really necessary to re-run the whole
>group each time at the 2nd level... Thanks for any input.
________________________________
The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.
------------------------------------------------------------
This email message, including any attachments, is for the sole use of the intended recipient(s) and may contain information that is proprietary, confidential, and exempt from disclosure under applicable law. Any unauthorized review, use, disclosure, or distribution is prohibited. If you have received this email in error please notify the sender by return email and delete the original message. Please note, the recipient should check this email and any attachments for the presence of viruses. The organization accepts no liability for any damage caused by any virus transmitted by this email.
=================================
|