Continue with what you have done, do not model it as a continuous variable.
doug
On 06/02/2015 04:03 PM, Matthieu Vanhoutte wrote:
> Dear Experts,
>
> I have run GLM group analyses with Freesurfer tools according DODS
> model, with six classes ((Male,Female) ; (Control,Left,Right)) and one
> age
Dear Experts,
I have run GLM group analyses with Freesurfer tools according DODS
model, with six classes ((Male,Female) ; (Control,Left,Right)) and one
age variable :
/Class MaleControl
Class MaleLeft
Class MaleRight
Class FemaleControl
Class FemaleLeft
Class FemaleRight
Variables Age
/Howev
For the intercept, you are testing for a difference in the mean CBF
between the groups regressing out the effects of gender and age.
For the slope, you are testing for an interaction between group and age
regressing out the effects of gender.
doug
On 3/12/15 1:30 PM, Matthieu Vanhoutte wrote
Hello,
I have run GLM group analyses with Freesurfer tools according DODS model,
with six classes ((Male,Female) ; (Control,Left,Right)) and one age
variable :
*Class MaleControlClass MaleLeftClass MaleRightClass FemaleControlClass
FemaleLeftClass FemaleRightVariables Age *
I wanted to tes
the sig value is actually -log10(p)*sign(gamma) where gamma is the
contrast (gamma.mgh). The p-value is two-tailed, and the color indicates
what tail you are on.
doug
On 03/10/2015 01:21 PM, Matthieu Vanhoutte wrote:
> Dear FreeSurfer's experts,
>
> I have launched a GLM analyses following
