On Mon, Jul 23, 2012 at 10:16 PM, Rongxiang Tang
<[log in to unmask]> wrote:
> Hi Donald,
>
> Thanks for the reply. I do have a question about this passage:
>
> Any contrast within a single condition will require a new model.
>
> For C1 - you need a one-sample t-test.
> For C2 - you need a one-sample t-test.
> For C3 - you need a two-sample t-test.
>
> I assume that if I set all other EVs to 0, except EV1, it should give me the
> mean, right?
I meant you need separate models. For C1 (GROUP1 PRE-MEAN), you need
to run a one-sample t-test with only the images for group1 pre. For C2
(GROUP2 POST-MEAN), you need to run a one-sample t-test with only the
images for group2 post. For C3 (comparing 2 groups), you need to run a
two-sample t-test with the images from both groups for either the pre
or post. The important thing to realize is that C1-C3 were not valid
contrasts for a design with both between-subject effects and
within-subject effects. Thus, the between-subject contrasts require
new models.
> Do I have to do another three separate models of design?
Yes.
>
>
> Thanks again,
> Catherine
>
> ________________________________
> From: "MCLAREN, Donald" <[log in to unmask]>
> To: [log in to unmask]
> Sent: Sunday, July 22, 2012 3:23 PM
> Subject: Re: [FSL] GLM setup questions
>
> For the questions you are trying to answer, you will need several models.
>
> In the above setup, you can remove EV5 and EV6 since they are
> redundant; however, I would add EVs for group1 and group 2 as well as
> EVs for pre- and post-.
>
> The large model can be used to evaluate the pre-/post-differences
> across all subjects, within a single group, or the interaction between
> groups (e.g. C4)
>
> Any contrast within a single condition will require a new model.
>
> For C1 - you need a one-sample t-test.
> For C2 - you need a one-sample t-test.
> For C3 - you need a two-sample t-test.
>
>
>
> 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, Jul 19, 2012 at 7:22 PM, Rongxiang Tang <[log in to unmask]>
> wrote:
>> Dear All,
>>
>> I'm setting up Glm design and contrasts matrices for my ICA dual
>> regression
>> analysis, and would like to confirm couple things.
>> I have two groups (exp,contrl, each with a pre and post scan), and I want
>> to
>> test if there is a difference in each group before and after the training,
>> and if there is a difference between groups, (exp pre vs. contrl pre).
>> I run a concat-ica for all of my subjects(pre,post,exp,contrl, each group
>> 22
>> subjects, total 88 inputs), and now want to set up the glm for my dual
>> regression analysis.
>>
>> I was wondering how can I combine 1) unpaired t-test(within subjects), and
>> 2) paired t-test(between subjects) in the set up...
>>
>> let's say if I have 2 groups, 2 subjects in each group, pre&post
>> condition:
>> Group EV1 EV2 EV3 EV4 EV5
>> EV6 EV7 EV8 EV9 EV10
>> input1(sub1pre) 1 1 0 0 0
>> 1 0 1 0 0 0
>> input2(sub2pre) 1 1 0 0 0
>> 1 0 0 1 0 0
>> input3(sub1post) 1 0 0 1 0
>> -1 0 1 0 0 0
>> input4(sub2post) 1 0 0 1 0
>> -1 0 0 1 0 0
>> input5(sub1pre) 2 0 1 0 0
>> 0 1 0 0 1 0
>> input6(sub2pre) 2 0 1 0 0
>> 0 1 0 0 0 1
>> input7(sub1post) 2 0 0 0 1
>> 0 -1 0 0 1 0
>> input8(sub2post) 2 0 0 0 1
>> 0 -1 0 0 0 1
>>
>>
>> EV1 EV2
>> EV3 EV4 EV5 EV6 EV7 EV8 EV9 EV10
>> C1 GROUP1 PRE-MEAN 1 0 0 0
>> 0 0 0 ...............................
>> C2 GROUP2 POST-MEAN 0 0 0 1
>> 0 0 .........................
>> C3 GROU1P1PRE>GROUP2PRE 1 -1 0
>> .....................................................
>> C4 GROUP1PRE>GROUP1POST 0 0...................... 1
>> 0...........................................
>>
>> are the above setup correct?
>>
>> Thanks,
>> Catherine
>
>
|