> Do you mean that I have to set the contrast as 1 -1 1 0 0?
If you don't want to regress out effects due to which group the
subjects are in, then I would suggest not including the group
columns in the design matrix. You would then simply have a
design matrix containing va + block + global, for which you
could use contrasts [1 0 0 ...] and [-1 0 0 ...].
Best regards,
-John
> >===== Original Message From John Ashburner <[log in to unmask]> =====
> >
> >> I would like to use VBM to compare the gray matter concentration and
> >> volume in two groups: Group A and Group B, where Group B is actually a
> >> control group. In order to study the correlation between the gray
> >> matter changes in both groups and visual acuity (va), I used the
> >> following settings:
> >>
> >> design type: single-subject: conditions & covariates
> >> select images
> >> condition: A A A A A A A A A A B B B B B B B B B B
> >> # covariates: 1
> >> covariate: spm_load (select the ASCII file with the values for visual
> >> acuity)
> >> covariate name: va
> >> interaction: none
> >> covariate centre: around overall mean
> >> global normalization: AnCova
> >> grand mean scaling: scaling of overall grand mean
> >> scale overall grand mean to: 100
> >> threshold masking: absolute
> >> analysis threshold: 25
> >> global calculation: user specified (select the ASCII file with the
> >> global values)
> >>
> >> it came up with 5 columns: 2 conditions + va + block + global, and I set
> >> the contrast as 0 0 1 0 0.
> >>
> >> Am I using the correct way to study such correlation? Then how should I
> >> interpret the results on the glass brains?
> >
> >It depends what you want the statistics to relate to. The current model
> >will look for a significant relationship between GM and va that cannot be
> >explained by the subjects being in the patient group or not. Modelling
> >the control/patient effect as a confound in the model may be what you
> > intend.
--
Dr John Ashburner.
Functional Imaging Lab., 12 Queen Square, London WC1N 3BG, UK.
tel: +44 (0)20 78337491 or +44 (0)20 78373611 x4381
fax: +44 (0)20 78131420 http://www.fil.ion.ucl.ac.uk/~john
|