You should bring the average of each condition across the two runs to
the group level. I thought the above was a group level analysis as
that would be where you want to test the main effects of emotion.
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 Fri, Sep 14, 2012 at 9:38 PM, Shan(Hannah) Luo <[log in to unmask]> wrote:
> Thank you so much Donald for your helpful response! I went ahead using
> second method you suggested! It gave me maps for F-tests on the first level.
> However when I run a second level analysis ( each person had two sessions,
> just a simple fixed-effect average across sessions), it only output maps for
> the contrasts I made on the first level analysis but no output for F-tests
> maps.
>
> How should I proceed if I want to get a subject-level F-tests maps? Thanks
> very much!
>
>
> On Friday, September 14, 2012, MCLAREN, Donald <[log in to unmask]>
> wrote:
>> Hannah,
>>
>> There are several issues with your design and analysis:
>> (1) This is a repeated measures and as such you need to have an extra
>> EV for each subject (see
>>
>> http://fsl.fmrib.ox.ac.uk/fsl/feat5/detail.html#ANOVA1factor4levelsRepeatedMeasures)
>> (2) If you treat this a one-way ANOVA - which you need to do in FSL --
>> you can keep the design matrix as is with modeling all conditions (or
>> you can follow the tutorial and leave one input out). I think for what
>> you are trying to do, you want model everything.
>> (3) Your contrasts will need to be:
>> 1 -1 0 0 0 0 0 ....
>> 0 1 -1 0 0 0 0 ....
>> 0 0 0 1 -1 0 0 ....
>> 0 0 0 0 1 -1 0 ...
>> 1/2 -1/2 0 1/2 -1/2 0 0 ...
>> 0 1/2 -1/2 0 1/2 -1/2 0 ...
>>
>> For the main effect of emotion during period 1, use the the first two
>> contrasts.
>> For the main effect of emotion during period 2, use the second two
>> contrasts.
>> For the main effect of emotion, use the last two contrasts.
>>
>> The better solution, would be to run two models --> one for period 1
>> and one for period 2.
>> Alternatively, you can build a 2x3 repeated measures ANOVA in SPM or
>> using GLM Flex
>> (http://nmr.mgh.harvard.edu/harvardagingbrain/People/AaronSchultz/GLM_Flex.html)
>> in MATLAB. GLM Flex uses partitioned variance to properly look at each
>> factor in you model (time and emotion).
>>
>> 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 Thu, Sep 13, 2012 at 7:00 PM, Shan(Hannah) Luo <[log in to unmask]> wrote:
>>> Dear FSL experts,
>>>
>>> I have a within-subject design,with three conditions, happy, fear and
>>> neutral emotional state. I want to do a F-test looking at the main effect
>>> of
>>> emotional state on BOLD data. For my GLM, I have 7 EVs, EV1 ( Happy state
>>> during period 1), EV2 ( Fear state during period 1), EV3 ( Neutral state
>>> during period 1), EV4 ( Happy state during period 2), EV5 (Fear during
>>> period 2), and EV 6 ( Neutral during period 2), and EV7 ( taking into
>>> account a secondary task performance). Period 1 and Period 2 were two
>>> different time periods I was interested in.
>>>
>>> I was planning to do two F-tests(main effect of state during period 1 and
>>> 2
>>> separately), as below,
>>>
>>> Ev1 Ev2 Ev3 Ev4 Ev5 Ev6 Ev7 F1 F2
>>> 1 0 0 0 0 0 0 1
>>> 0 1 0 0 0 0 0 1
>>> 0 0 1 0 0 0 0 1
>>> 0 0 0 1 0 0 0 1
>>> 0 0 0 0 1 0 0 1
>>> 0 0 0 0 0 1 0 1
>>>
>>> For F1, Ev4~Ev7 are irrelevant. What should I do with them? And the same
>>> for
>>> F2 ( Ev1~Ev3 and Ev7 are irrelevant). Any thoughts would be greatly
>>> appreciated. Thanks ahead of time.
>>
|