I'm sure to follow you here - how did you code for the task? if it is
ones everywhere then it simply removes the mean and 0 1 0 tells you how
much variance the second column explains
if you mean that for each scan you have a task and it is always the
same then the model is fine ..
hope this helps
> Thanks Cyril.
> So if I understand correctly, then it is not necessary to explicitly
> include the task column (the 3rd column with 1s everywhere) as an
> effect of interest?
> If the contrast should be 0 1 0 and the first 0 is controlling for age
> then isn't the second 0 somehow removing variance associated with the
> On Thu, Jul 23, 2009 at 4:38 AM, cyril pernet <[log in to unmask]
> <mailto:[log in to unmask]>> wrote:
> I am trying to figure out which set of contrast codes is
> correct to examine the effect illness while controlling for
> age. The order of the columns in the attached screen shots is
> age, illness and task.
> there are 3 columns only so suppose 1st is age, 2nd illness - the
> 3rd one has 1s everywhere, i.e. this is your constant (mean) -
> correct contrast is the 2nd one [0 1 0] showing the unique
> variance explained by illness, age effect being taken into account.
Dr Cyril Pernet,
fMRI Lead Researcher SINAPSE
SFC Brain Imaging Research Center
Division of Clinical Neurosciences
University of Edinburgh
Western General Hospital
[log in to unmask]
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.