Hi Cornelius and Elena,
PPheights are not taken into account in randomise analyses of TBSS and VBM...
So Elena, the changes you've made are fine.
Cheers,
Gwenaelle
--- En date de : Mer 24.3.10, Cornelius Werner <[log in to unmask]> a écrit :
> De: Cornelius Werner <[log in to unmask]>
> Objet: Re: [FSL] Design matrix with covariat age
> À: [log in to unmask]
> Date: Mercredi 24 mars 2010, 11h48
> Hi,
>
> you probably need to adjust PPheights (peak-to-peak
> heights) to "1 1 6", with "6" being the distance between
> your smallest (-2) and your largest (4) value of the
> corresponding vector. Although I do not know to what extent
> this value influences subsequent statistics.
>
> Cheers
> Cornelius
>
> Am 24.03.2010 um 11:22 schrieb Elena Miramontes:
>
> > Hi,
> >
> > I have a further question concerning the design
> matrix, which I use for the voxelwise statistics in TBSS.
> > At the moment my design matrix has two EV, the patient
> group and the control group.
> > /Num Waves 2
> > /Num Points 31
> > /PPheights 1 1
> > /Matrix
> > 1 0
> > 1 0
> > 1 0
> > 0 1
> > 0 1
> > 0 1
> > ...
> >
> > Now, I want to add the age as a covariate. (I have
> demeaned the age). How do I have to change the values in
> /Num Waves, /Num Points and /PPheigts?
> >
> > My suggestion is:
> > /Num Waves 3
> > /Num Points 31
> > /PPheights 1 1
> > /Matrix
> > 1 0 -2
> > 1 0 1
> > 1 0 -1
> > 0 1 4
> > 0 1 2
> > 0 1 1
> > ...
> >
> > Is that right?
> > I changed the value in /Num Waves corresponding to the
> numbers of columns. But I didn't changed the value in
> /PPheights.
> >
> > Thank you for your help,
> > Elena
>
|