Since you are trying to test a between-subject effect (a single
regressor), you should use the between-subject error term, which is 1.
The between-subject error term is always the same.
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, Aug 9, 2012 at 11:17 AM, Johannes Keyser <[log in to unmask]> wrote:
> Dear people familiar with GLM flex,
>
> I am new to the GLM flex scripts (since yesterday), and so far everything worked great!
> My goal is to be able to specify the correct error term manually.
> Doing it manually is advised on the web-page (quoting from http://nmr.mgh.harvard.edu/harvardagingbrain/People/AaronSchultz/GLM_Flex_Intro_.html):
> "The Cons.ET field allows you to specify which error term to use, if you leave that blank [], the scripts will attempt to choose the correct error term for you, but I cannot guarantee that the scripts will guess correctly, so it is best practice to specify this yourself."
>
> However, I've not found any further information on what to base my judgment on, and in my case I have 4 error terms (ResMS_02%d.nii-files) to choose from!
> (My design includes 1 between factor, 2 within, each having 2 levels).
> In particular, I want to look at a single specific "column/level" of an interaction term.
>
> I shall give a simpler example to explain what I mean:
> Let's assume a design with two 2-level factors A and B, modeling both main effects (2 columns each) and its interaction (4 columns).
> What is the full contrast specification to inspect the AxB-interaction's level A1,B1?
> In particular, how can I determine the appropriate error term (field ET)?
>
> In GLM-flex parlance, assuming main effects sitting at columns 1,2 and 3,4 this is what I would do:
> I.Cons(1).name = 'A1,B1';
> I.Cons(1).Groups = {5}; % the entire interaction columns span 5,6,7,8
> I.Cons(1).Levs = 1;
> I.Cons(1).ET = []; %TODO: gain understanding, and then filling it in manually!
> I.Cons(1).mean = 0;
>
> Any help, both practical and theoretical, is very much appreciated!
>
> Johannes Keyser
> Institute of Cognitive Science
> University of Osnabrueck
|