The F-contrast is simply T^2 with DF(1,df of T) when then are only 2
levels being compared. For simplicity we GLM Flex only reports the T.
You could create the F by using imcalc can squaring the T-statistic
image.
If you use OrthoView to view the results, then you can see the +/-
values. When +- values are both shown, the test is treated as
two-tailed and will be identical to the F-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 Mon, Nov 26, 2012 at 6:25 PM, MCLAREN, Donald
<[log in to unmask]> wrote:
> Please see in line responses. Sorry for the delay.
>
> On Fri, Nov 2, 2012 at 8:59 AM, Anja Dietrich <[log in to unmask]> wrote:
>> Sorry for the empty message. Here is the question:
>>
>> Dear experts,
>>
>> since there is no opportunity to compare groups in a flexible factorial design in spm, I set up a mixed design ANOVA in GLM flex (see attachment for the design matrix).
>>
>> The model consists of 3 factors:
>> group (3 levels)
>> condition (2 levels)
>> the obligatory subject factor
>>
>> and a group/ condition interaction
>>
>> Because contrast definition differs from that in SPM, I would be very happy if one can help me to define the contrasts correctly.
>>
>> First of all I'd like to define a giant f contrast to check if any column of the design matrix differs from 0.
>
>>>>> This doesn't make any sense to do. I presume you want to test if any of your group/conditions are different than 0. Here is how to do it.
>
> Any group different than 0:
> I.Cons(x).Groups={columns representing factor 1} [1 2 3]
> I.Cons(x).Levs=0;
> I.Cons(x).ET=1;
>
> Any condition different than 0:
> I.Cons(x).Groups={columns representing factor 2} [4 5]
> I.Cons(x).Levs=0;
> I.Cons(x).ET=1;
>
> Any group/condition pair different than 0:
> I.Cons(x).Groups={columns representing the interaction} [6 7 8 9 10 11]
> I.Cons(x).Levs=0;
> I.Cons(x).ET=1;
>
>>
>> Further I'd like to create F and t contrasts like the following:
>> Group1Condition1 / Group2Condition1
>
> did you mean less than?? Group1Condition1 < Group2Condition1
> I.Cons(x).Groups={column for G2C1 then G1C1} [8 6]
> I.Cons(x).Levs=2;
> I.Cons(x).ET=1;
>
>> Group1Condition1 > Group2Condition1
>
> I.Cons(x).Groups={columns for G1C1 then G2C1} [6 8]
> I.Cons(x).Levs=2;
> I.Cons(x).ET=1;
>
> ET is 1 for both because these are between-subject comparisons.
> [ ] indicates what columns should be in the { }.
>
>
>>
>> Thank's a lot in advance!
>>
>> Anja
>>
>>
>>
>
> Hope this answers your questions.
|